← Back

Body map

Immune system

What research found here. What was shown to help comes first, then the risks reported, most common first.

Immune system

Strong several reviews agree · Moderate one review or consistent trials · Limited one or a few small studies · Untested no study in people, listed with what it rests on

Shown to help56

Antibiotics (common outpatient)Strong

Acute sinus infection in adults

Shown to cure slightly more adults with acute sinusitis than placebo, 5 to 11 more per 100, while 13 more per 100 had side effects.

Compared with placebo or no treatment [1]; placebo [2]

The numbers and sources
Effect
Without antibiotics, 46% were cured at 1 week and 64% at 14 days. With a clinical diagnosis, antibiotics raised cure (OR 1.25, 95% CI 1.02 to 1.54), a number needed to treat of 19, and 10 when the diagnosis was confirmed on imaging. Side effects were more frequent (OR 2.21), one extra person harmed for every 8 treated [1]. In individual patient data from 9 trials, 15 adults needed antibiotics for one more to be cured, and common signs and symptoms could not pick out who would benefit [2].
Population
3,057 adults in 15 trials in ambulatory care [1]; 2,547 adults in 9 trials [2].
Sources
[1], [2]
Caveats
Children, people with weakened immune systems and severe sinusitis were not studied [1]. The trials used a range of antibiotics.
Antibiotics (common outpatient)Strong

Doxycycline after sex to prevent sexually transmitted infections

Shown to cut chlamydia and syphilis by about three quarters with doxycycline after sex, mostly in men who have sex with men, with a smaller or no effect on gonorrhea.

Compared with standard care without doxycycline [15][16]; no prophylaxis [17]

The numbers and sources
Effect
In people on HIV prevention pills, a sexually transmitted infection was found at 10.7% of quarterly visits on doxycycline 200 mg within 72 hours of sex and 31.9% on standard care (RR 0.34), with similar results in people living with HIV (RR 0.38). Tetracycline-resistant gonorrhea occurred in 5 of 13 cases on doxycycline and 2 of 16 on standard care [15]. Pooled trials found lower chlamydia (RR 0.22) and syphilis (RR 0.24) and a smaller drop in gonorrhea (RR 0.79) [16]. Another meta-analysis found chlamydia (RR 0.35) and syphilis (RR 0.23) lower and no significant effect on gonorrhea (RR 0.90) [17].
Population
501 men who have sex with men and transgender women [15]; 1,278 participants in 3 trials [16]; 1,727 participants in 4 trials, 73% men who have sex with men [17].
Sources
[15], [16], [17]
Caveats
Open-label trials. Long-term effects on resistance remain uncertain [16][17].
High-sensitivity C-reactive protein (hs-CRP)StrongStudied in women

Polycystic ovary syndrome

Linked to higher CRP in women with polycystic ovary syndrome, about twice the level of women without it, in two meta-analyses that held after accounting for obesity.

The numbers and sources
Effect
In 31 studies, mean CRP was 96% higher in women with PCOS than in controls without it (95% CI 71% to 122%), with no sign of publication bias. The difference persisted after excluding five studies in which body mass or the frequency of obesity differed between the groups. The same review found no difference in interleukin-6 or tumor necrosis factor alpha, in 10 and 9 studies [21]. In 63 studies, CRP was higher in women with PCOS than in controls (standardized mean difference 1.26, 95% CI 0.99 to 1.53), and in 35 higher quality studies that included non-obese women the difference was 1.80 (1.36 to 2.25) [22].
Population
3,648 women, 2,359 with PCOS and 1,289 controls, in 31 studies [21]. 4,086 women with PCOS and 3,120 controls in 63 studies [22].
Sources
[21], [22]
Caveats
Observational comparisons of women with and without the condition, which cannot show which came first. Body weight raises CRP. The reviews dealt with it by setting aside studies with mismatched weight or by analyzing non-obese women, not by adjusting each result for body mass index [21][22].
Oral corticosteroidsStrong

Rheumatoid arthritis

Shown to ease joint tenderness and pain in rheumatoid arthritis more than placebo or NSAIDs, and to slow joint damage on X-ray over 1 to 2 years.

The numbers and sources
Effect
Low-dose prednisolone, at no more than 15 mg a day, improved joint tenderness against placebo by 12 tender joints (95% CI 6 to 18) and improved pain (SMD 1.75, 95% CI 0.87 to 2.64) and grip strength. Against NSAIDs it improved joint tenderness by 9 joints (95% CI 5 to 12) and pain (SMD 1.25, 95% CI 0.26 to 2.24) [4]. Across 15 studies, mostly in early disease and added to other disease-modifying drugs, glucocorticoids slowed X-ray progression (SMD 0.40, 95% CI 0.27 to 0.54), with a similar effect at 2 years (0.42, 95% CI 0.30 to 0.55) [5]. In a 2-year trial of prednisone 10 mg a day, clinical improvement over placebo was seen mainly in the first 6 months and X-ray progression was lower after month 6, with more osteoporotic fractures in the prednisone group [6].
Population
320 people with rheumatoid arthritis in 10 short-term studies [4]; 1,414 people in 15 studies [5]; people with early active rheumatoid arthritis not yet on disease-modifying drugs [6].
Sources
[4], [5], [6]
Caveats
The short-term review covers small trials [4]. The mean cumulative dose in the X-ray studies was 2,300 mg of prednisone equivalent in the first year [5].
Thunder god vineCautionStrong

Rheumatoid arthritis

Shown to improve joint swelling, stiffness and grip in adults with rheumatoid arthritis, matching or beating standard drugs, in trials mostly from China.

The numbers and sources
Effect
Across 14 RCTs, TwHF extracts improved grip strength, swollen joint count and morning stiffness against placebo (P < 0.001). Against DMARDs, grip strength, morning stiffness, CRP and tender joint count differed (P < 0.05), and ESR and swollen joint count did not [1]. Across 40 RCTs (3,092 patients), tripterygium glycosides added to DMARDs lowered morning stiffness, tender and swollen joint counts, pain scores, CRP, ESR and rheumatoid factor against DMARDs alone (all p < .001) [2]. A network meta-analysis of 21 trials (5,039 participants) ranked TwHF with methotrexate first and TwHF alone second for ACR20 response among the conventional DMARDs compared [3]. In 207 patients, ACR50 response at 24 weeks was 46.4% on methotrexate, 55.1% on TwHF (p 0.014) and 76.8% on both (p < 0.001) [4]. In 121 US patients, 65.0% on TwHF extract and 32.8% on sulfasalazine met ACR20 at 24 weeks (P 0.001) [5].
Population
Adults with active rheumatoid arthritis, mostly in China [1][2][3][4]. 121 patients at US centers [5].
Sources
[1], [2], [3], [4], [5]
Caveats
The methotrexate trial was open label, with blinded assessors [4]. In the sulfasalazine trial only 62 patients had outcome data at 24 weeks, and 62% and 41% stayed on treatment [5]. Most pooled trials were small and Chinese.
TurmericStrong

C-reactive protein and other inflammation markers

Shown to lower CRP and some other inflammation markers in adults with inflammatory conditions or diabetes. These are blood markers, not clinical outcomes.

The numbers and sources
Effect
In 9 studies, curcumin lowered CRP by 3.67 mg/L (95% CI -6.96 to -0.38). In 23 studies hs-CRP was lower than placebo, with stronger effects below 2,000 mg a day [14]. In 28 RCTs in prediabetes and type 2 diabetes, curcumin or turmeric lowered CRP (SMD -0.50), TNF-alpha (SMD -1.70) and IL-6 (SMD -2.97) [15]. In 6 RCTs (472 adults with type 2 diabetes), hs-CRP was 1.06 mg/L lower (95% CI -1.97 to -0.16, I² 17.3%) [16]. In 21 knee osteoarthritis studies (1,705 patients), CRP (SMD -0.906, 95% CI -1.543 to -0.269) and TNF-alpha were lower than placebo, while ESR, IL-1beta, IL-6 and PGE-2 did not differ [17]. With piperine added, 18 RCTs showed lower IL-6, while CRP and TNF-alpha did not change significantly [18].
Population
People with proinflammatory or autoinflammatory conditions [14]; adults with prediabetes or type 2 diabetes [15][16]; adults with knee osteoarthritis [17]; adults in curcumin with piperine trials [18].
Sources
[14], [15], [16], [17], [18]
Caveats
Blood markers, not clinical outcomes. Certainty was low for every marker in the 28-trial review, which also found publication bias for CRP [15]. The curcumin with piperine review found no significant CRP change [18].
AcupunctureModerate

Allergic rhinitis

Shown to lower nasal symptom scores in pooled trials. The largest trial found small gains over sham acupuncture that its authors said may not be clinically meaningful.

Compared with control groups the source abstract does not describe [13]; sham acupuncture [14]; rescue antihistamine alone [14]

The numbers and sources
Effect
Across 13 trials, nasal symptom scores fell more with acupuncture than with control (weighted mean difference -4.42, 95% CI -8.42 to -0.43) and medication scores fell too [13]. In 422 people with seasonal allergic rhinitis, 12 sessions improved rhinitis quality of life by 0.5 points more than sham acupuncture and 0.7 points more than antihistamine alone after 8 weeks. The differences were gone by week 16 [14].
Population
2,365 participants in 13 trials [13]. 422 adults sensitized to birch and grass pollen in Germany [14].
Sources
[13], [14]
Caveats
The meta-analysis abstract does not say what its control groups received [13]. Baseline scores in the largest trial were low [14].
Anti-inflammatory dietModerate

C-reactive protein

Linked to lower odds of elevated C-reactive protein, about 10% for each point lower on the Dietary Inflammatory Index.

Compared with people with higher (more pro-inflammatory) Dietary Inflammatory Index scores, per point [6]

The numbers and sources
Effect
In 14 studies (59,941 people), each one-unit rise in the index was linked to 10% higher odds of elevated C-reactive protein (OR 1.10, 95% CI 1.06 to 1.14, I² 0%) [6].
Population
Adults in observational studies [6].
Sources
[6]
Caveats
Observational. The abstract's confidence interval for the highest against lowest category (OR 1.39) does not contain the estimate, so only the per-unit figure is reported here [6].
Antibiotics (common outpatient)Moderate

Sore throat

Shown to reduce sore throat at day 3 against placebo, with fewer than 6 treated for one extra person without it, though 82% on placebo were symptom-free by one week.

Compared with placebo or no treatment [3]

The numbers and sources
Effect
Sore throat at day 3 was less frequent on antibiotics (RR 0.70, 95% CI 0.60 to 0.80), and at 1 week (RR 0.50), when the number needed to treat was 18. Fever at day 3 did not differ (RR 0.75, 95% CI 0.53 to 1.07). Antibiotics lowered ear infections within 14 days (Peto OR 0.21), quinsy (Peto OR 0.16) and acute rheumatic fever (Peto OR 0.36), but not sinusitis [3].
Population
15,337 cases of sore throat in 29 trials, mostly adults [3].
Sources
[3]
Caveats
Most trials were run in the 1950s, when complications such as rheumatic fever were far more frequent. Few trials separated bacterial from viral sore throat, and harms were poorly reported [3].
Antibiotics (common outpatient)ModerateStudied in children

Ear infections in children

Shown to slightly reduce children's ear pain at 2 to 3 days, 20 treated for one fewer in pain, but not at 24 hours, with extra vomiting, diarrhea or rash.

Compared with placebo [7]; waiting to see if antibiotics were needed [7]

The numbers and sources
Effect
Pain at 24 hours did not differ (RR 0.89, 95% CI 0.78 to 1.01), but was a third less frequent at 2 to 3 days (RR 0.71, number needed to treat 20). Abnormal tympanometry at 2 to 4 weeks (RR 0.83), eardrum perforation (RR 0.43) and infection in the other ear (RR 0.49) fell, and late recurrences did not change. Vomiting, diarrhea or rash rose (RR 1.38, one extra per 14 treated). Against waiting, immediate antibiotics did not lower pain at 3 to 7 days (RR 0.75) and raised these side effects (RR 1.87) [7].
Population
3,401 children in 13 placebo trials and 1,556 in 6 trials of immediate antibiotics against waiting, all in high-income countries [7].
Sources
[7]
Caveats
Serious complications were rare in both groups. Results may not apply where complications are more common [7].
AntihistaminesModerate

Hay fever and other allergic rhinitis

Shown to lower hay fever symptom scores more than placebo, but nasal steroid sprays relieved nasal symptoms more than oral antihistamines.

Compared with placebo [1][2]; nasal corticosteroid sprays [3]

The numbers and sources
Effect
In a perennial allergic rhinitis trial, levocetirizine 5 mg lowered the 4-symptom total score (0 to 12) by 1.22 points more than placebo (95% CI 0.76 to 1.69), with similar results at 2.5 and 10 mg [1]. A network meta-analysis found six of seven total nasal symptom estimates in seasonal rhinitis favored active treatment over placebo, but rated all comparisons between drugs very low certainty and found no reliable ranking [2]. Against oral antihistamines, nasal steroids gave more relief of total nasal symptoms (SMD -0.42), blockage (SMD -0.63) and sneezing, with no difference for eye symptoms [3].
Population
517 people aged 12 and older in one levocetirizine trial [1]; 34,831 participants in 61 reports [2]; randomized trials comparing nasal steroids with oral antihistamines [3].
Sources
[1], [2], [3]
Caveats
The network meta-analysis found the evidence mostly placebo-anchored and sparse [2]. The nasal steroid comparison dates from 1998 [3].
AntihistaminesModerate

Chronic hives

Shown to clear chronic hives completely more often than placebo, about 2.7 times as often with cetirizine 10 mg, in evidence drawn from few studies.

Compared with placebo [4][1]; other antihistamines [4]

The numbers and sources
Effect
Cetirizine 10 mg once daily led to complete suppression of urticaria more often than placebo (RR 2.72, 95% CI 1.51 to 4.91). Levocetirizine 5 mg beat placebo in the intermediate term but not the short term, and desloratadine results had very wide intervals. More people had complete suppression on cetirizine than on fexofenadine. Withdrawals for adverse events did not differ significantly [4]. In two 4-week label trials, levocetirizine 2.5, 5 and 10 mg lowered itch scores more than placebo, with fewer and smaller wheals [1].
Population
9,759 participants in 73 studies [4]; 423 adults aged 18 to 85 with chronic idiopathic urticaria [1].
Sources
[4], [1]
Caveats
Most comparisons rest on few studies or single-study estimates [4].
44 more
AstaxanthinModerate

Oxidative stress and inflammation

Shown to lower blood malondialdehyde, an oxidative stress marker, and IL-6 in type 2 diabetes. CRP and TNF-alpha did not change.

The numbers and sources
Effect
Astaxanthin lowered blood malondialdehyde (SMD -0.95, 95% CI -1.67 to -0.23). IL-6 fell in type 2 diabetes (WMD -0.70 pg/mL, 95% CI -1.29 to -0.11). CRP and TNF-α did not change [7].
Population
380 participants in 12 RCTs [7].
Sources
[7]
Caveats
Few trials measured each marker. These are biomarkers, not disease outcomes.
Beta-glucanModerate

Colds and upper respiratory infections (yeast beta-glucan)

Shown to lower the odds of colds and shorten them in healthy people taking yeast beta-glucan, across few trials with high heterogeneity.

The numbers and sources
Effect
In 13 trials in healthy people, yeast beta-glucan lowered the odds of an upper respiratory infection (OR 0.345, 95% CI 0.192 to 0.620), the number of episodes (SMD -0.315) and their duration (SMD -0.312) compared with placebo [9]. A review of 34 trials of fungal beta-glucans at 2.5 to 1,000 mg a day for up to 6.5 months reported fewer respiratory infection symptoms, but immune cell and molecular findings were inconsistent [10].
Population
Healthy subjects [9]. Healthy people and patients [10].
Sources
[9], [10]
Caveats
Heterogeneity was high and the trials were few [9]. A review of commercial beta-glucan products found an average of 33 participants per arm, and more than half of trials that reported funding were sponsored by beta-glucan producers [11].
Fermented foodsModerate

Respiratory infections

Shown to lower respiratory infections by about 19% with probiotic fermented dairy against placebo, in one meta-analysis of 22 trials across ages.

Compared with placebo [4]

The numbers and sources
Effect
Probiotic fermented dairy lowered respiratory tract infections overall (RR 0.81, 95% CI 0.74 to 0.89), in children (RR 0.82), adults (RR 0.81, 95% CI 0.66 to 1.00) and older people (RR 0.78). Upper respiratory infections fell (RR 0.83), the common cold fell (RR 0.68, 95% CI 0.49 to 0.96), and lower respiratory infections fell marginally (RR 0.78, 95% CI 0.60 to 1.01) [4].
Population
10,190 participants in 22 randomized trials [4].
Sources
[4]
Caveats
The products were dairy foods with added probiotic strains, so this may not apply to other fermented foods. The adult estimate's interval reaches 1.00.
Fish oilModerate

Lupus disease activity

Shown to lower lupus disease activity about 0.9 SLEDAI points at 3 g a day or less in adults with lupus, at low certainty.

The numbers and sources
Effect
Across 5 RCTs (274 patients), omega-3 at 3 g a day or less reduced lupus disease activity (SMD -0.33, 95% CI -0.57 to -0.09), about 0.9 SLEDAI points (95% CI -1.6 to -0.3) [25]. A systematic review of 13 trials found 8 of 10 that measured disease activity reported improvement. Kidney measures, C3, C4 and anti-dsDNA did not change [26].
Population
Adults with systemic lupus erythematosus, 80% or more women, over 12 to 52 weeks [25][26].
Sources
[25], [26]
Caveats
The meta-analysis rated certainty as low [25]. The systematic review describes the evidence as conflicting [26].
GarlicModerate

Inflammation markers

Shown to lower CRP by about 0.6 mg/L and lower TNF in adults. These are blood markers, and the reviews disagree on IL-6.

The numbers and sources
Effect
In 16 RCTs, garlic lowered CRP by 0.61 mg/L (95% CI -1.12 to -0.11, I² 76.9%, 13 trials), IL-6 by 0.73 ng/L (5 trials) and TNF by 0.26 ng/L (7 trials). Adiponectin and leptin did not change significantly [8].
Population
Adults in 16 RCTs lasting 2 to 52 weeks [8].
Sources
[4], [8]
Caveats
Blood markers, not clinical outcomes. Heterogeneity for CRP was high [8]. The 108-trial review found CRP 1.6 mg/L and TNF 3.94 pg/mL lower, effects several times larger than [8], and found no change in IL-6 [4].
GingerModerate

Inflammation markers

Shown to lower CRP, TNF-α and IL-6 in adults, more in people with health conditions. These are blood markers, not clinical outcomes.

The numbers and sources
Effect
In a 2025 GRADE-assessed meta-analysis of 29 RCTs, ginger lowered CRP by 0.86 mg/L (95% CI -1.10 to -0.62), TNF-α by 1.90 pg/mL (95% CI -2.61 to -1.18) and IL-6 by 1.15 pg/mL (95% CI -1.90 to -0.41). Malondialdehyde fell, and total antioxidant capacity and superoxide dismutase rose [8].
Population
Adults in 29 RCTs [8].
Sources
[8]
Caveats
Blood markers, not clinical outcomes. The authors report high heterogeneity among trials and larger effects in people with underlying health conditions [8].
MontelukastModerate

Hay fever (seasonal allergic rhinitis)

Shown to improve hay fever symptoms modestly against placebo, about 5% more, slightly less than antihistamines and less than nasal steroid sprays.

Compared with placebo [5]; antihistamines [5]; nasal corticosteroids [5]

The numbers and sources
Effect
Leukotriene receptor antagonists lowered daily rhinitis symptom scores 5% (95% CI 3% to 7%) more than placebo and improved rhinoconjunctivitis quality of life by 0.3 units. Antihistamines improved nasal symptoms 2% (95% CI 0% to 4%) more, and nasal corticosteroids 12% (95% CI 5% to 18%) more, than leukotriene receptor antagonists [5].
Population
8 trials with 3,924 people with seasonal allergic rhinitis [5].
Sources
[5]
Caveats
No randomized trials of perennial rhinitis were found in this review [5]. The label reserves rhinitis use for people who did not respond to or could not take other treatments [1].
Oral corticosteroidsModerate

Severe COVID-19 and pneumonia in hospital

Shown to lower 28-day deaths in hospitalized COVID-19 patients, 22.9% against 25.7%, with the benefit limited to those on oxygen or a ventilator.

The numbers and sources
Effect
Dexamethasone 6 mg a day for up to 10 days, by mouth or vein, lowered 28-day deaths from 25.7% on usual care to 22.9% (rate ratio 0.83, 95% CI 0.75 to 0.93). Deaths were 29.3% against 41.4% in those on a ventilator and 23.3% against 26.2% in those on oxygen. In those on no respiratory support they were 17.8% against 14.0% (rate ratio 1.19, 95% CI 0.92 to 1.55) [13]. In severe community-acquired pneumonia, added corticosteroids lowered deaths (RR 0.47, 95% CI 0.3 to 0.74) and C-reactive protein on day 8 (SMD -0.8, 95% CI -1.11 to -0.5), with no difference in gastrointestinal bleeding or high blood sugar [14].
Population
6,425 people hospitalized with COVID-19, 2,104 on dexamethasone and 4,321 on usual care [13]; 665 people with severe community-acquired pneumonia in 10 trials [14].
Sources
[13], [14]
Caveats
The COVID-19 trial was open label [13]. The pneumonia trials were small and used several corticosteroids and dosing schedules, so they do not test oral treatment alone [14].
Paleo dietModerate

Inflammation markers

Shown in two reviews to lower CRP about 0.4 to 0.8 mg/L more than comparison diets in adults, with one result sensitive to single trials.

Compared with control diets the source does not describe [1]; a control group the source does not describe [5]

The numbers and sources
Effect
CRP fell 0.41 mg/L more than on control diets (95% CI 0.008 to 0.81) [1] and 0.84 mg/L more in people with metabolic disorders (95% CI 0.06 to 1.62) [5].
Population
Adults in 8 RCTs [1]; adults with metabolic disorders in 10 RCTs [5].
Sources
[1], [5]
Caveats
Confidence intervals reach close to zero. In [1], the CRP result changed when single studies were removed. CRP is a blood marker, not a disease outcome.
Pet ownershipModerateStudied in children

Eczema in children

Linked to about 28% less eczema in children exposed to dogs in one meta-analysis of birth cohorts. Cat exposure was not linked.

Compared with children not exposed to dogs, cats or pets [12]

The numbers and sources
Effect
Exposure to dogs (RR 0.72, 95% CI 0.61 to 0.85) and to pets overall (RR 0.75, 95% CI 0.67 to 0.85) was linked to less atopic dermatitis. Exposure to cats was not (RR 0.94, 95% CI 0.76 to 1.16) [12].
Population
Children in 21 birth cohorts, 26 publications, exposed during pregnancy, infancy or childhood [12].
Sources
[12]
Caveats
Observational. Families with allergies may avoid getting pets, which can make pets look protective. Heterogeneity was moderate (I² 46% for dogs) [12].
ProbioticsModerateStudied in children

Late-onset sepsis in preterm infants

Shown to lower late-onset sepsis confirmed by culture about 17% in preterm infants in hospital, with shorter stays.

The numbers and sources
Effect
In 31 studies (8,040 preterm infants of 36 weeks or less), probiotics lowered culture-proven late-onset sepsis (RR 0.83, 95% CI 0.72 to 0.95). Hospital stays were shorter, and mortality trended lower without reaching significance [42].
Population
Preterm infants in hospital [42].
Sources
[42]
Caveats
One meta-analysis. Probiotic sepsis and the FDA warning on preterm products are listed under Risks.
Thyroid peroxidase antibodies (TPO antibodies)Moderate

Later hypothyroidism

Linked to later hypothyroidism: over 20 years, women with thyroid antibodies alone had 8 times the odds, and 38 times with a raised TSH as well.

The numbers and sources
Effect
In a 20-year British follow-up, positive thyroid antibodies alone carried odds ratios for developing hypothyroidism of 8 in women (95% CI 5 to 15) and 25 in men (10 to 63). With a raised TSH as well, the odds ratios were 38 (22 to 65) and 173 (81 to 370) [7]. In an Australian cohort, hypothyroidism 13 years later was present in 12.0% of antibody-positive women whose baseline TSH was 2.5 mU/L or less, 55.2% at 2.5 to 4.0 mU/L and 85.7% above 4.0 mU/L [5]. In a 5-year Chinese survey, 14.44% of people positive for TPO or thyroglobulin antibodies developed thyroid dysfunction, against 3.31% of those without antibodies [8]. Among 82 women with subclinical hypothyroidism, overt hypothyroidism developed in 58.5% of those with microsomal antibodies and 23.2% of those without [9]. A US guideline puts progression from subclinical to overt hypothyroidism at 4.3% a year with TPO antibodies and 2.6% a year without [4].
Population
2,779 adults in the Whickham Survey [7]. 1,184 adults in the Busselton Health Surveys [5]. 3,018 people in three Chinese regions with different iodine intakes [8]. 82 women followed a mean 9.2 years [9].
Sources
[7], [5], [8], [9], [4]
Caveats
Observational. The studies used different antibody assays and cutoffs, and earlier ones measured antibodies by semiquantitative methods, not automated immunoassay [5]. The US guideline also reports that about 75% of people with chronic autoimmune thyroiditis have raised antibody levels, many with normal thyroid function at the time [4].
TurmericModerate

COVID-19 in hospital

Shown to lower death, ventilation and worsening in patients with COVID-19 when added to standard care, about 61% fewer deaths, in small trials.

The numbers and sources
Effect
Across 23 RCTs (1,407 patients), turmeric added to standard care lowered all-cause death (RR 0.39, 95% CI 0.23 to 0.67, 8 RCTs, moderate certainty), need for intubation or mechanical ventilation (RR 0.35, 95% CI 0.17 to 0.72) and clinical deterioration (RR 0.36, 95% CI 0.22 to 0.59) [40].
Population
Patients with mild to severe COVID-19, hospitalized in 17 of 23 trials. Nano-curcumin at 160 to 240 mg a day was the most common form [40].
Sources
[40]
Caveats
Nine trials were at high risk of bias. Results held when those were excluded [40]. The trials were small.
Whole-body cryotherapyModerate

Inflammatory markers

Shown to lower interleukin-1 beta and raise interleukin-10 in a meta-analysis of 11 small trials. Single sessions after exercise did not change interleukin-6 or C-reactive protein.

The numbers and sources
Effect
Across 11 RCTs (274 participants), serum interleukin-1 beta was lower with cryotherapy than control (SMD -2.08) and interleukin-10 was higher (SMD 0.78) [9]. In 11 endurance-trained men, one 3-minute session at -110°C after interval running did not change interleukin-6, interleukin-10, C-reactive protein, cortisol or testosterone against rest [10]. In two placebo-controlled trials, neither cryotherapy nor cold water changed inflammation or muscle damage markers against placebo, apart from C-reactive protein at 24 hours [4][5].
Population
274 people in 11 RCTs, including athletes and people with obesity [9]. 11 trained men in a crossover [10]. 31 and 24 trained men [4][5].
Sources
[9], [10], [4], [5]
Caveats
The pooled trials are small and differ in population and exposure. Blood cytokines are lab markers, and none of these studies tied the change to a health outcome.
ZincModerate

Length of a cold, zinc started at onset

Shown to shorten colds about 2.4 days when started at onset, in low-certainty evidence from children and adults.

The numbers and sources
Effect
In treatment trials, zinc shortened colds by a mean 2.37 days (95% CI -4.21 to -0.53, I² 97%, 8 studies, 972 participants, low certainty) [1]. In 7 lozenge trials using more than 75 mg a day, colds were 33% shorter (95% CI 21% to 45%). Acetate and gluconate lozenges did not differ significantly, and 192 to 207 mg a day did not outperform 80 to 92 mg a day [2].
Population
34 trials with 8,526 children and adults [1]; 575 adults with natural colds [2].
Sources
[1], [2]
Caveats
Heterogeneity was very high and most trials were at unclear or high risk of bias [1]. The single-author lozenge analysis [2] selected trials by dose.
Absolute neutrophil count (ANC)Limited

Viral infections, blood cancers and death after a low count on a routine test

Limited evidence: one Danish primary care cohort linked low counts to more viral infections, blood cancers and deaths, more strongly the lower the count.

The numbers and sources
Effect
Neutropenia, defined as below 1,500 cells/µL, was seen in about 1% of people. Odds ratios for viral infections were 2.32 (95% CI 1.84 to 2.91) at 1,500 to 1,800, 2.80 (2.20 to 3.57) at 1,000 to 1,500 and 4.77 (3.22 to 7.07) below 1,000. For blood cancers they were 3.23 (2.35 to 4.45), 8.69 (6.58 to 11.47) and 46.03 (33.98 to 62.35). Neutropenia was particularly associated with HIV, acute leukemias and myelodysplastic syndromes, and not with autoimmune disorders or solid cancers. For severe neutropenia the authors report absolute risks of 40% for blood cancers and above 50% for death from any cause over 4 years [14].
Population
More than 370,000 people with a routine blood count in Danish primary care, linked to national health registers and followed for 4 years [14].
Sources
[14]
Caveats
Observational, in people whose doctor had ordered a blood count. A low count can be the first sign of an illness already present, so the count marks the risk and need not cause it [14].
AstragalusLimitedStudied in children

Colds

Limited evidence: in children with nephrotic syndrome, astragalus granules were tied to fewer upper respiratory infections. A Cochrane review found no placebo-controlled trial for preventing them in children.

The numbers and sources
Effect
A Cochrane review screened 120 full-text papers on oral astragalus to prevent frequent respiratory infections in children and found no randomized trial comparing astragalus alone with placebo, so it could not assess benefit or safety. 44 studies were excluded for having no placebo [10]. In two studies pooled for children with nephrotic syndrome, upper respiratory tract infections occurred in 23.9% given astragalus granules with prednisone against 42.9% on prednisone alone (RR 0.56, 95% CI 0.33 to 0.93), with evidence quality rated low [11].
Population
Children up to 14 years in community settings [10]. Children with nephrotic syndrome taking prednisone [11].
Sources
[10], [11]
Caveats
Evidence quality was low, and the comparison was against prednisone alone with no placebo [11]. No trial in adults with colds was located.
Black seed oilLimitedStudied in women

Inflammation markers

One small trial found 1,000 mg a day lowered IL-6 in 60 women with PCOS. Trials of other inflammation markers conflicted.

The numbers and sources
Effect
In 60 women with polycystic ovary syndrome, 1,000 mg a day of oil for 12 weeks lowered serum IL-6 against placebo (adjusted mean difference -4.21, 95% CI -6.34 to -2.08) and fasting glucose by 8.16 mg/dL. Fasting insulin and HOMA-IR did not differ [9]. A review of 4 human trials found conflicting results for the adhesion molecules VCAM-1 and ICAM-1 [10].
Population
60 women with PCOS [9]. Patients with coronary artery disease, type 2 diabetes, Hashimoto's thyroiditis or cardiovascular risk factors [10].
Sources
[9], [10]
Caveats
One small trial for IL-6 [9]. The adhesion molecule review also included 15 animal and in vitro studies, which are not counted here [10].
BoswelliaLimited

Inflammation markers

One trial found a boswellic acid syrup lowered several inflammation markers more than placebo in mild to moderate COVID-19.

The numbers and sources
Effect
In patients with mild to moderate COVID-19, a boswellic acid syrup (Inflawell) for 14 days lowered serum IL-1α, IL-8, TNF-α and MCP-1 more than placebo. The abstract reports p values only, not the size of the changes [7].
Population
Patients with mild to moderate COVID-19. The abstract does not give the number randomized [7].
Sources
[7]
Caveats
One trial of one product during an acute infection. Blood markers, not symptoms or disease outcomes [7].
Cold water immersionLimited

Sickness absence and immune markers

One trial of 3,018 adults found 29% less sickness absence with a daily cold shower finish for 30 days. Days of illness did not differ.

The numbers and sources
Effect
Adults randomized to end a daily shower with 30, 60 or 90 seconds of cold water had 29% less sickness absence from work than controls (incident rate ratio 0.71, P = 0.003). Illness days showed no significant group effect [17]. In 60 adults randomized to daily cold or hot showers for 90 days, the cold group showed higher immunoglobulin levels and higher interleukin-2 and interleukin-4 at 90 days, with no difference in interferon-gamma or TNF-alpha [18]. Pooled trials found no significant effect on immune function immediately or 1 hour after immersion [14].
Population
3,018 Dutch adults aged 18 to 65 without severe comorbidity [17]. 60 healthy Egyptian adults [18]. Healthy adults in 11 trials [14].
Sources
[17], [18], [14]
Caveats
Sickness absence was self-reported by people who knew their group, and people felt ill as often in both groups [17]. Blood immune markers are not infections avoided [18].
Why limited
This rests on one trial, of 3,018 people. Strength counts independent studies that agree, so one trial is limited at any size.
CordycepsLimitedStudied in men

Immune markers

One trial found 1.5 g a day raised natural killer cell activity and other immune markers in 79 healthy men. Infections were not measured.

The numbers and sources
Effect
In 79 healthy Korean men, 1.5 g a day of Cordyceps militaris for 4 weeks raised natural killer cell activity, lymphocyte proliferation, IL-2 and interferon gamma more than placebo [5].
Population
79 healthy men [5].
Sources
[5]
Caveats
Blood markers, not infection rates. One trial.
ElderberryLimited

Cold and upper respiratory symptoms

Small trials found fewer cold days in air travelers, 57 against 117 on placebo. A review rated the evidence uncertain.

The numbers and sources
Effect
A meta-analysis of 180 participants reported a large reduction in upper respiratory symptoms with elderberry [1]. A systematic review of 5 randomized trials found elderberry may reduce cold duration and severity, with uncertain evidence [2]. In 312 air travelers, cold days totaled 57 on elderberry and 117 on placebo, and summed symptom scores 247 against 583 [3].
Population
180 participants [1], 5 RCTs [2], 312 economy-class passengers flying from Australia [3].
Sources
[1], [2], [3]
Caveats
The meta-analysis abstract gives no effect size or trial count. It was funded by the Franklin Institute of Wellness [1]. The air travel trial was funded by Iprona AG, which supplied the capsules and was partly involved in design [3].
Erythrocyte sedimentation rate (ESR)Limited

Diagnoses found at 100 mm/hr or above

Limited evidence: among 4,807 clinic patients with an ESR of 100 or above, infection was found in 40%, autoimmune disease in 38% and malignancy in 36%.

The numbers and sources
Effect
The leading diagnosis associated with an ESR of 100 mm/hr or above was infection (1,932 patients, 40%), followed by autoimmune disease (1,839, 38%) and malignancy (1,736, 36%). The authors report that almost all patients had an identifiable cause. Values were slightly higher in men than in women, and did not correlate with age, race or C-reactive protein [10].
Population
4,807 patients with an ESR of 100 mm/hr or above examined at Mayo Clinic in Rochester, Minnesota, from 2002 to 2011 [10].
Sources
[10]
Caveats
Retrospective, at one referral center, with no comparison group below 100 mm/hr. The diagnosis groups overlap [10].
Fisetin and quercetin (senolytics)LimitedStudied in men

Inflammation and metabolic markers (fisetin alone)

Small trials found lower inflammation and insulin resistance markers with 200 mg a day of fisetin in men with obesity, most with training.

The numbers and sources
Effect
Over 12 weeks, IL-6 and TNF-α fell in the fisetin-only group, and fasting glucose, insulin and HOMA-IR fell in all active groups, most in training plus fisetin [11]. A second report from the same research group found weight loss in the fisetin-only group compared with placebo, and the largest drops in asprosin and MCP-1 with training plus fisetin [12].
Population
44 men with obesity at 200 mg a day [11]; 60 sedentary men with obesity at 200 mg a day [12].
Sources
[11], [12]
Caveats
Both reports come from one research group with similar four-arm designs. Neither measured senescent cells.
GarlicLimited

Common cold

One trial found 24 colds on garlic against 65 on placebo over 12 weeks. Days to recovery were similar.

The numbers and sources
Effect
A Cochrane review found only one eligible trial. Its 146 participants on a daily garlic supplement for 12 weeks reported 24 colds against 65 on placebo, and 111 against 366 days of illness. Days to recovery were similar, 4.63 against 5.63 [9].
Population
146 adults in one trial [9].
Sources
[9]
Caveats
One trial with self-reported colds. The review authors found the evidence insufficient [9].
GinsengLimited

Colds

One trial found fewer colds and fewer cold days over 4 months on a North American ginseng extract in adults prone to colds.

The numbers and sources
Effect
In 279 adults with at least 2 colds in the prior year, an extract of North American ginseng taken for 4 months lowered colds per person (0.68 vs 0.93), the share with 2 or more colds (10.0% vs 22.8%) and days with cold symptoms (10.8 vs 16.5) [8].
Population
Adults aged 18 to 65 in Edmonton, Canada [8].
Sources
[8]
Caveats
One trial of one polysaccharide-rich extract. 44 people who never started treatment were excluded from analysis [8].
Greens powdersLimited

Energy and immunity

One trial found a fruit and vegetable juice powder cut days with moderate or severe cold symptoms by 20% in hospital staff. No trial measuring energy was located.

The numbers and sources
Effect
In 529 healthcare workers given encapsulated fruit and vegetable juice powder (Juice Plus+) or placebo daily for 8 months, days with moderate or severe cold symptoms over 6 months averaged 7.6 against 9.5 (p = 0.023). Total days with any cold symptom did not differ, 29.4 against 30.7 (p = 0.616) [11].
Population
529 hospital staff in Berlin aged 18 to 65, mostly nurses, 80% women [11].
Sources
[11]
Caveats
One trial of one encapsulated fruit and vegetable product, not a greens drink powder. Symptoms were self-reported in diaries, and the effect was limited to more severe symptom days [11]. No trial of AG1 or other greens blends measured energy, fatigue or infections.
H2 blockersLimited

Hives, added to an antihistamine

Small trials found adding ranitidine or cimetidine to diphenhydramine cleared hives more often than diphenhydramine alone. The evidence was old and at high or unclear risk of bias.

Compared with diphenhydramine alone [8]

The numbers and sources
Effect
Ranitidine with diphenhydramine resolved urticaria more often than diphenhydramine alone (RR 1.59, 95% CI 1.07 to 2.36). Cimetidine alone did not improve overall symptoms more than diphenhydramine, but the combination did (RR 2.02, 95% CI 1.03 to 3.94). Sedation from baseline did not differ with famotidine or diphenhydramine [8].
Population
144 participants in 4 small studies [8].
Sources
[8]
Caveats
The review authors called the evidence very limited and rated the studies at high to unclear risk of bias. The direct evidence is for ranitidine and cimetidine, not famotidine [8].
High-sensitivity C-reactive protein (hs-CRP)LimitedStudied in women

Premenstrual symptoms and period cramps

One cohort of midlife women linked hs-CRP above 3 mg/L to more premenstrual mood symptoms, cramps or back pain, bloating and breast pain, but not headaches.

The numbers and sources
Effect
Against hs-CRP of 3 mg/L or below, a level above 3 mg/L had adjusted odds ratios of 1.27 for premenstrual mood symptoms (95% CI 1.02 to 1.58), 1.40 for abdominal cramps or back pain (1.09 to 1.80), 1.41 for appetite cravings, weight gain or bloating (1.04 to 1.89) and 1.26 for breast pain (1.02 to 1.55). It was not associated with premenstrual headaches or with reporting three or more symptoms [25].
Population
2,939 women aged 42 to 52, before or early in the menopausal transition, at the start of the Study of Women's Health Across the Nation, from several racial and ethnic groups [25].
Sources
[25]
Caveats
Cross-sectional, so it cannot show which came first. Covariates, body mass index among those tested, stayed in each model only if they were significant [25]. The women were in midlife, so the result may not describe younger women with period pain.
HydroxyzineLimited

Itching

Limited evidence: hydroxyzine matched loratadine for clearing chronic hives in one comparison, and itch scores fell in a small dialysis trial with no placebo group.

The numbers and sources
Effect
In a Cochrane review of antihistamines for chronic hives, complete suppression over the short term did not differ between hydroxyzine 25 mg and loratadine 10 mg (RR 1.00, 95% CI 0.32 to 3.10) [8]. In a crossover trial in people on dialysis, itch severity on a 10-point scale fell from 6.83 to 2.86 over 6 weeks on hydroxyzine 25 mg a day and from 7.1 to 2.17 on gabapentin, with no significant difference between the two [9]. A Cochrane review found no randomized trial of an oral antihistamine used alone for eczema [10].
Population
73 studies of antihistamines with 9,759 participants, of which one comparison involved hydroxyzine [8]; 32 patients on dialysis [9].
Sources
[8], [9], [10]
Caveats
An approved use that rests on older evidence. Neither comparison here had a placebo arm for hydroxyzine, and the hives comparison had a wide confidence interval.
Juice cleanseLimitedStudied in women

Inflammation

One trial found the inflammation marker hs-CRP fell on a 7-day lemon detox but not on a calorie-matched placebo juice, in overweight Korean women.

Compared with before the diet, in the same people (overweight Korean women), with a calorie-matched placebo juice group measured the same way [1]

The numbers and sources
Effect
Serum hs-CRP fell only in the lemon detox group. Hemoglobin and hematocrit stayed stable on the lemon detox and fell in the other two groups [1].
Population
84 premenopausal overweight Korean women [1].
Sources
[1]
Caveats
One short trial, with a blood marker rather than a disease outcome. Each group was measured against its own starting level, and no test between the groups is reported for hs-CRP [1].
MeditationLimited

Telomerase and cell aging

Small trials found telomerase activity in blood immune cells rose with mindfulness meditation, pooled across 4 trials and 190 people.

Compared with control groups in 4 randomized trials the source abstract does not describe [20]

The numbers and sources
Effect
A meta-analysis of 4 randomized trials found increased telomerase activity in peripheral blood mononuclear cells (d = 0.46) [20].
Population
190 participants in 4 trials [20].
Sources
[20]
Caveats
Telomerase activity is an enzyme measure, not telomere length, lifespan or a health outcome. Four trials is a small base.
MontelukastLimitedStudied in children

Sleep apnea and snoring in children

Small trials found montelukast improved sleep study measures, snoring and mouth breathing in children with mild to moderate obstructive sleep apnea over short periods.

Compared with placebo [7]; nasal mometasone alone [7]

The numbers and sources
Effect
Against placebo, oral montelukast improved polysomnography measures, snoring, mouth breathing and adenoid size. Added to mometasone nasal spray, it improved sleep study measures and snoring more than the spray alone [7].
Population
305 children with mild to moderate obstructive sleep apnea in 4 randomized trials [7].
Sources
[7]
Caveats
Four small short-term trials. The abstract gives no effect sizes [7].
MSMLimited

Seasonal allergies

Small uncontrolled trials found fewer nasal allergy symptoms after 2 to 4 weeks of MSM, measured against each person's own baseline.

Compared with before MSM, in the same people [8][9]

The numbers and sources
Effect
In an open-label trial, 50 people with seasonal allergic rhinitis took 2,600 mg a day for 30 days, and upper and total respiratory symptoms fell significantly from baseline by day 7 [8]. In a randomized, double-blind study comparing doses after a standard allergen challenge, 1, 3 or 6 g a day for 14 days lowered average nasal symptom scores 53.72% from baseline (P = .001) and raised nasal airflow 17.32% (P = .02). No adverse events occurred [9].
Population
50 adults with seasonal allergic rhinitis, 30 days [8]. Adults with a history of allergic nasal congestion, 14 days [9].
Sources
[8], [9]
Caveats
Neither study had a placebo group. The second study randomized people between doses, not to placebo, and its abstract does not give the number enrolled [9]. The first trial's abstract does not give the size of the fall [8].
NutsLimitedStudied in children

Peanut allergy in high-risk infants

One trial in infants at high risk found peanut allergy at age 5 in 1.9% of those fed peanut and 13.7% of those who avoided it.

Compared with avoiding peanut until 60 months of age [14]

The numbers and sources
Effect
In infants with a negative skin test at the start, peanut allergy at 60 months was 13.7% with avoidance and 1.9% with regular peanut consumption (P < 0.001). In infants with a small positive skin test, it was 35.3% and 10.6% (P = 0.004). Serious adverse events did not differ significantly between groups [14]. The FDA lists a qualified health claim for ground peanuts and reduced risk of developing peanut allergy [2].
Population
640 infants aged 4 to 11 months with severe eczema, egg allergy or both [14].
Sources
[14], [2]
Caveats
One trial in infants at high risk. It does not test nuts in older children or adults.
RapamycinLimited

Immune response to vaccines and infections in older adults

Small trials of rapamycin analogs, not sirolimus, found a better flu vaccine response and fewer reported infections in older adults.

The numbers and sources
Effect
The rapamycin analog everolimus (RAD001) raised the response to influenza vaccine by about 20% in older volunteers [4]. A low-dose combination of everolimus and BEZ235 taken for 6 weeks was associated with fewer reported infections over the following year (P = 0.001) and a better influenza vaccine response [5].
Population
Older volunteers [4]; 264 older adults in a phase 2a trial [5].
Sources
[4], [5]
Caveats
These trials tested rapamycin analogs, not sirolimus itself. Infections were self-reported [5]. [5] lists Novartis funding.
SaunaLimitedStudied in men

Pneumonia and respiratory disease

Linked to fewer hospital diagnoses of pneumonia and other respiratory disease in Finnish men who used a sauna 2 or more times a week.

The numbers and sources
Effect
Over 25.6 years, 379 of 1,935 men had a hospital diagnosis of chronic obstructive pulmonary disease, asthma or pneumonia. Against one session a week or fewer, hazard ratios were 0.73 (95% CI 0.58 to 0.92) at 2 to 3 sessions and 0.59 (95% CI 0.37 to 0.94) at 4 or more [17]. For pneumonia alone, in 2,210 men with 375 cases, the fully adjusted hazard ratios were 0.72 (95% CI 0.57 to 0.90) and 0.63 (95% CI 0.39 to 1.00) [18].
Population
Finnish men aged 42 to 61 [17][18].
Sources
[17], [18]
Caveats
Observational, and both analyses use the same cohort. Only men were studied.
SaunaLimited

Common colds

One trial found fewer colds over 6 months in 25 people who took regular saunas than in 25 who did not. Duration and severity did not differ.

The numbers and sources
Effect
The sauna group had significantly fewer episodes of common cold over 6 months. In the last 3 months the incidence was roughly half that of controls. Mean duration and severity of colds did not differ between groups [19].
Population
50 volunteers, 25 taking regular saunas and 25 abstaining, followed for 6 months [19].
Sources
[19]
Caveats
One small trial from 1990. The abstract gives no exact counts and does not say that allocation was randomized or that anyone was blinded.
SeleniumLimited

Lymphedema

Small trials found less lymphedema and fewer skin infections in people after cancer treatment, two of them giving selenium by vein.

The numbers and sources
Effect
A Cochrane review found one RCT of selenium for secondary lymphedema, with 60 participants at the start. It reported fewer recurrent erysipelas infections on selenium than on placebo [9]. In 26 women with stage II to III lymphedema after breast cancer, five intravenous doses of 500 µg sodium selenite within 2 weeks shifted 83.3% from stage III to II at follow-up, against 10.0% on saline (p 0.002), and lowered the extracellular water ratio [10]. In 20 patients after oral cancer surgery, 1,000 µg sodium selenite a day by mouth or vein for 3 weeks reduced postoperative lymphedema [11]. A 2022 systematic review of lymphedema drug trials found one small RCT suggesting selenium might help [12].
Population
Cancer patients with secondary lymphedema [9]. 26 women with breast cancer-related lymphedema [10]. 20 patients after oral tumor surgery [11].
Sources
[9], [10], [11], [12]
Caveats
The Cochrane authors note drawbacks in study quality and reporting, and that the result cannot be generalized [9]. All trials are small, and two gave sodium selenite intravenously [10][11].
Sleep durationLimited

Colds and vaccine response

Limited evidence: sleeping under 6 hours was linked to more colds after viral exposure in one study and to weaker vaccine antibody responses when sleep was measured objectively.

Compared with people sleeping more than 7 hours a night [16]; people not sleeping under 6 hours a night [17]

The numbers and sources
Effect
After rhinovirus nasal drops, adults whose wrist-measured sleep was under 5 hours had higher odds of a clinical cold (OR 4.50, 95% CI 1.08 to 18.69), as did those sleeping 5 to 6 hours (OR 4.24). Those sleeping 6.01 to 7 hours did not (OR 1.66, 95% CI 0.40 to 6.95) [16]. Objectively measured short sleep was linked to a lower antibody response to vaccination (effect size 0.79, 95% CI 0.40 to 1.18). Self-reported short sleep was not (0.29, 95% CI -0.04 to 0.63). The link was significant in men and not in women [17].
Population
164 healthy adults aged 18 to 55, quarantined [16]. 304 adults with objective sleep measures and 504 with self-reported sleep [17].
Sources
[16], [17]
Caveats
Sleep was observed, not assigned, so other differences between short and long sleepers may explain the results. Intervals were wide [16]. Antibody levels are not infections prevented.
Slow breathing and breathworkLimited

Hyperventilation breathwork and inflammation

Small trials of a program combining cyclic hyperventilation, cold exposure and meditation found a blunted inflammatory response to injected endotoxin, and a fall in ESR in spondyloarthritis.

Compared with untrained volunteers [17]; a control period without training [18]

The numbers and sources
Effect
After 10 days of training in meditation, cyclic hyperventilation with breath retention, and cold immersion, 12 volunteers given intravenous endotoxin had higher plasma adrenaline and anti-inflammatory IL-10, lower TNF-alpha, IL-6 and IL-8, and fewer flu-like symptoms than 12 untrained controls [17]. In 24 patients with axial spondyloarthritis, ESR fell over 8 weeks of the program from a median of 16 to 9 mm/hr (p = 0.040) with no change in controls, and hs-CRP did not change [18]. A systematic review of 8 trials of the Wim Hof Method concluded it may reduce inflammation and found mixed effects on exercise performance [19].
Population
24 healthy volunteers [17]. 24 patients with moderately active axial spondyloarthritis and raised hs-CRP [18]. 8 trials [19].
Sources
[17], [18], [19]
Caveats
The breathing was never tested apart from the cold and meditation in these trials. The spondyloarthritis trial was open label [18]. Endotoxin injection is a laboratory model of inflammation, not a disease [17].
Thyroid peroxidase antibodies (TPO antibodies)Limited

Thyroid inflammation after childbirth

Limited evidence: a guideline review reports postpartum thyroiditis in 33 to 50% of women who were thyroid antibody positive in the first trimester.

The numbers and sources
Effect
The American Thyroid Association reports that women who are thyroid antibody positive in the first trimester have a 33 to 50% risk of postpartum thyroiditis, and that women with the highest antibody levels have the highest risk. It reports that 10 to 20% of cases result in permanent hypothyroidism [15].
Population
A guideline summary of earlier studies, whose sizes it does not give [15].
Sources
[15]
Caveats
Secondhand figures from guideline text, not a primary study.
Vitamin DLimited

Preventing autoimmune disease

One trial found about 22% less autoimmune disease over 5 years in older adults, and the difference was gone 2 years after the trial ended.

The numbers and sources
Effect
In VITAL, 2,000 IU a day for a median 5.3 years led to confirmed autoimmune disease in 123 people against 155 on placebo (HR 0.78, 95% CI 0.61 to 0.99, P 0.05). The outcome counted rheumatoid arthritis, polymyalgia rheumatica, autoimmune thyroid disease, psoriasis and others [48]. Two years after the trial ended, the difference was gone (HR 0.98, 95% CI 0.83 to 1.17 at 7 years) [49].
Population
25,871 US men 50 and older and women 55 and older, mean age 67.1 [48]. 21,592 followed for 2 more years [49].
Sources
[48], [49]
Caveats
Both sources are the same trial. Adding 65 newly confirmed cases from the randomized period changed the HR to 0.85 (95% CI 0.70 to 1.04) [49]. Participants were not selected for deficiency.
Why limited
This rests on one trial, of 25,871 people. Strength counts independent studies that agree, so one trial is limited at any size.
Zinc levelLimited

Pneumonia in older adults in nursing homes

Limited evidence: one study of Boston nursing home residents linked serum zinc of 70 µg/dL or above to fewer and shorter pneumonia episodes.

The numbers and sources
Effect
Compared with residents whose serum zinc was below 70 µg/dL, those with final concentrations of 70 µg/dL or above had a lower incidence of pneumonia, almost 50% fewer new antibiotic prescriptions, a shorter duration of pneumonia and fewer days of antibiotic use (2.6 against 3.9 days), with P at or below 0.004 for all. Baseline serum zinc of 70 µg/dL or above was associated with lower all-cause mortality (P = 0.049) [6].
Population
Residents of 33 nursing homes in Boston who took part in a 1-year vitamin E trial, 578 with baseline and 420 with final serum zinc. All received daily essential vitamins and minerals, including zinc, at 50% of the US dietary allowance (RDA) [6].
Sources
[6]
Caveats
An observational analysis within a trial of another nutrient. Infection itself lowers serum zinc [1], so low zinc may mark illness and need not cause it.

Reported risks78

Oral corticosteroidsModerateVery common

Infections

Shown to raise infections, in 12.7% on systemic corticosteroids against 8.0% of controls across 71 trials, with no rise below 10 mg a day of prednisone.

The numbers and sources
How common
Very common: more than 1 in 10 people.
Effect
Pooling 71 controlled trials, infectious complications occurred in 12.7% of people assigned systemic corticosteroids and 8.0% of controls (RR 1.6, 95% CI 1.3 to 1.9). The rate was not raised at a daily dose under 10 mg or a cumulative dose under 700 mg of prednisone [19]. The prednisone label states that live or live attenuated vaccines are contraindicated at immunosuppressive doses, and warns people on corticosteroids to avoid exposure to chickenpox and measles [20].
Population
2,111 people assigned systemic corticosteroids and 2,087 controls in 71 trials [19].
Sources
[19], [20]
Caveats
Published in 1989. Infection rates also rose with dose among people given placebo, which the authors attribute to the underlying illness [19].
TesamorelinModerateVery common

Injection site reactions and allergic reactions

Shown to raise injection site reactions, 25% against 14% on placebo, in adults with HIV. Allergic reactions such as rash and hives occurred in 4%.

The numbers and sources
How common
Very common: more than 1 in 10 people.
Effect
Injection site reactions occurred in 25% on tesamorelin and 14% on placebo during the first 26 weeks. Hypersensitivity reactions, including itching, redness, flushing, hives and rash, occurred in 4% of tesamorelin-treated patients [6]. More people on tesamorelin than placebo withdrew because of an adverse event [1]. Pooled across 4 RCTs, discontinuation rates were higher on tesamorelin (RR 2.25, 95% CI 0.98 to 5.17, p = 0.06) [4].
Population
543 adults on tesamorelin and 263 on placebo in the 26-week phase [6]; 412 adults [1]; 909 people in 4 RCTs [4].
Sources
[6], [1], [4]
Caveats
In the liver trial, local injection site complaints were more common on tesamorelin, and none were judged serious [8].
Direct oral anticoagulants (DOACs)LimitedVery common

More clots than warfarin in antiphospholipid syndrome

One small trial found strokes or heart attacks in 12% on rivaroxaban and none on warfarin in high-risk antiphospholipid syndrome, and was stopped early.

The numbers and sources
How common
Very common: more than 1 in 10 people.
Effect
In people with antiphospholipid syndrome positive on all three antibody tests, thromboembolic events occurred in 7 of 59 on rivaroxaban (12%), 4 ischemic strokes and 3 heart attacks, against none of 61 on warfarin. Major bleeding occurred in 7% against 3% [26]. The rivaroxaban label states that DOACs have been associated with more recurrent thrombotic events than vitamin K antagonists in this group [23].
Population
120 high-risk patients followed a mean 569 days [26].
Sources
[26], [23]
Caveats
A small open-label trial in one high-risk group. The evidence is for rivaroxaban.
HomeopathyLimitedVery common

Temporary worsening of symptoms

One trial found initial worsening of rhinitis in 30% of people on homeopathy against 7% on placebo. Homeopaths call this an aggravation.

The numbers and sources
How common
Very common: more than 1 in 10 people.
Effect
In 51 people with perennial allergic rhinitis, initial aggravations of symptoms occurred in 7 (30%) given homeopathy and 2 (7%) given placebo (P = 0.04) [12]. Five of 41 trials (12%) reported homeopathic aggravations, 98% of them graded mild [21]. A review of anxiety studies found "remedy reactions" of temporary worsening and the return of old symptoms [16].
Population
51 adults [12]. 41 trials [21]. Anxiety studies [16].
Sources
[12], [21], [16]
Caveats
One small trial gives the only rate per person.
BupropionModerateCommon

Rash and allergic reactions

Shown to raise rash, in 4% to 5% against 1% on placebo in depression trials. Severe allergic reactions were reported during trials and after approval.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
In depression trials, rash occurred in 5% at 300 mg a day and 4% at 400 mg a day against 1% on placebo, and hives in 2% and 1% against 0% [15]. The label states that anaphylactoid reactions with itching, hives, swelling and shortness of breath occurred during clinical trials, and that Stevens-Johnson syndrome and anaphylactic shock have been reported rarely after approval [8].
Population
490 adults on sustained-release bupropion and 385 on placebo [15]; trial and postmarketing reports [8].
Sources
[15], [8]
Caveats
The label gives no rate for the severe reactions.
SulfonylureasModerateCommon

Allergic skin reactions

Shown to cause allergic skin reactions such as itching and hives in about 1.5% on glyburide and under 1% on glimepiride in trials.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
Allergic skin reactions, including itching, redness, hives and measles-like rashes, occurred in 1.5% of people on glyburide in trials [14], in under 1.5% on extended-release glipizide [13] and in under 1% on glimepiride. Postmarketing reports for glimepiride include anaphylaxis, angioedema and Stevens-Johnson syndrome [3].
Population
Participants in the labels' clinical trials [14][13][3].
Sources
[14], [13], [3]
Caveats
The band above is the glyburide rate. The labels give no placebo rate for skin reactions.
CGRP migraine drugsLimitedCommon

Allergic reactions

Limited evidence: allergic reactions occurred in 1% to 2% on eptinezumab against none on placebo. Labels report rash, swelling and anaphylaxis after approval.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
Hypersensitivity reactions, such as itching and flushing on the day of infusion, occurred in 1% on eptinezumab 100 mg and 2% on 300 mg against 0% [14]. The erenumab label reports rash, angioedema and anaphylaxis after approval, some more than a week after a dose [12]. The ubrogepant label lists anaphylaxis and facial or throat swelling and contraindicates it after a serious reaction [15]. One anaphylactic reaction occurred on atogepant in a trial [7].
Population
1,153 on eptinezumab and 588 on placebo [14]; postmarketing reports [12][15]; 273 on atogepant [7].
Sources
[14], [12], [15], [7]
Caveats
The band above is the eptinezumab trial rate.
DairyLimitedCommon

Cow's milk allergy

A national survey estimated convincing milk allergy in 1.9% of US adults. Self-reported milk allergy ranged from 1.2% to 17% across studies.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
Among 40,443 US adults, convincing milk allergy was estimated at 1.9% (95% CI 1.8 to 2.1) [19]. A meta-analysis found self-reported milk allergy from 1.2% to 17%, far higher than objective measures [20]. Milk is one of the nine major food allergens under US law [2].
Population
40,443 US adults surveyed [19]. 51 studies of infants, children and adults [20].
Sources
[19], [20], [2]
Caveats
Survey estimates rest on reported symptoms [19]. Heterogeneity between studies was marked [20].
Fish and seafoodLimitedCommon

Shellfish and fish allergy

A national survey estimated convincing shellfish allergy in 2.9% and fin fish allergy in 0.9% of US adults. Studies using food challenges found lower figures.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
Among 40,443 US adults, convincing shellfish allergy was estimated at 2.9% (95% CI 2.7 to 3.1) and fin fish allergy at 0.9% (95% CI 0.8 to 1.0) [22]. A systematic review of 61 studies found fish allergy prevalence from 0% to 7% and shellfish allergy from 0% to 10.3%, depending on how it was diagnosed, and 0% to 0.3% and 0% to 0.9% where food challenges were used [23]. Fish and crustacean shellfish are among the nine major food allergens under US law [3].
Population
40,443 US adults surveyed in 2015 and 2016 [22]. 61 population-based studies worldwide [23].
Sources
[22], [23], [3]
Caveats
Surveys give higher figures than food challenge testing [23]. The survey estimate rests on reported symptoms [22].
NutsLimitedCommon

Peanut and tree nut allergy

A national survey estimated convincing peanut allergy in 1.8% and tree nut allergy in 1.2% of US adults. The FDA lists both among nine major food allergens.

The numbers and sources
How common
Common: between 1 in 100 and 1 in 10 people.
Effect
Among 40,443 US adults surveyed, convincing peanut allergy was estimated at 1.8% (95% CI 1.7 to 1.9) and tree nut allergy at 1.2% (95% CI 1.1 to 1.3). Among adults with any food allergy, 51.1% had experienced a severe reaction [15]. Peanuts and tree nuts are among the nine major food allergens identified by US law [3].
Population
40,443 US adults, mean age 46.6 [15].
Sources
[15], [3]
Caveats
Self-reported symptoms, judged convincing by symptom type, without allergy testing [15].
ACE inhibitorsModerateUncommon

Angioedema

Shown to raise angioedema, swelling of the face, tongue or throat, to 0.30% against 0.07% on placebo. The label reports a higher rate in Black patients.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Across 26 trials, the weighted incidence of angioedema was 0.30% on ACE inhibitors (95% CI 0.28 to 0.32), 0.11% on ARBs and 0.07% on placebo. In 7 head-to-head trials the risk was 2.2 times that on ARBs (95% CI 1.5 to 3.3). Incidence was higher in heart failure trials [25]. In ONTARGET, angioedema occurred in 0.3% on ramipril and 0.1% on telmisartan [15]. The label states that fatal cases have occurred, that it can happen at any time during treatment, that the rate is higher in Black than in non-Black patients, and that people taking an mTOR inhibitor such as sirolimus or everolimus, or a neprilysin inhibitor, may be at increased risk. It also describes angioedema of the intestine [3].
Population
74,857 people on ACE inhibitors in 26 trials with 232,523 person-years of follow-up [25]; 17,118 people in ONTARGET [15].
Sources
[25], [15], [3]
Caveats
Trials of at least 100 people and 8 weeks. The label gives no rate by race [3].
LamotrigineModerateUncommon

FDA boxed warning: serious skin rashes

FDA states lamotrigine has caused life-threatening rashes, including Stevens-Johnson syndrome, in about 0.3% of adults in epilepsy trials and 0.08% to 0.13% in bipolar trials.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
The label's boxed warning states that life-threatening serious rashes, including Stevens-Johnson syndrome and toxic epidermal necrolysis, and rash-related deaths have been caused by lamotrigine. Rash needing hospitalization occurred in 0.3% (11 of 3,348) of adults in epilepsy trials, 0.08% of adults starting it alone and 0.13% adding it to other drugs in mood disorder trials, and about 0.3% to 0.8% of children aged 2 to 17. Taking it with valproate, exceeding the starting dose and escalating faster than labeled raise the risk [4]. Across 122 randomized trials of lamotrigine alone, Stevens-Johnson syndrome or toxic epidermal necrolysis occurred in 0.04% [13].
Population
Adults and children in the manufacturer's clinical trials, as reported in the label [4]; 18,698 patients in 122 randomized trials [13].
Sources
[4], [13]
Caveats
The label states rash-related deaths in postmarketing use are too few to estimate a rate, and that a benign rash cannot be told apart early from a serious one [4].
66 more
Whole grainsModerateUncommon

Celiac disease

A global meta-analysis found 1.5% of the general population tested positive for celiac antibodies and 0.8% had celiac disease confirmed by biopsy.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Global seroprevalence was 1.5% (95% CI 1.3 to 1.8%) and biopsy-confirmed prevalence 0.8% (95% CI 0.6 to 1.0%). Prevalence was 50% higher in females and nearly double in children, and 41.4% of biopsy-confirmed patients had no symptoms [16].
Population
394,274 participants in 87 general population studies [16].
Sources
[16]
Caveats
Heterogeneity was very high (I² 97.1%), and 36.6% of people testing positive did not have a biopsy [16]. Celiac disease is triggered by gluten in wheat, barley and rye.
DulaglutideLimitedUncommon

Allergic and injection site reactions

The label reports systemic allergic reactions in 0.5% of adults in trials and injection site reactions in 0.5% against 0% on placebo, with postmarketing anaphylaxis.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Systemic hypersensitivity reactions, such as severe hives, widespread rash and facial or lip swelling, occurred in 0.5% of adults on dulaglutide. Injection site reactions occurred in 0.5% against 0.0% on placebo. Anaphylaxis and angioedema have been reported after marketing [1].
Population
Adults in clinical trials and postmarketing reports [1].
Sources
[1]
Caveats
No placebo rate is given for systemic reactions.
GardeningLimitedUncommon

Allergic reaction in the garden

One trial reported one allergic event in the garden among 145 people given a community garden plot.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
One participant in the gardening group (under 1%) had an adverse allergic event in the garden. No other adverse events were reported in the abstract [7].
Population
145 adults given a community garden plot [7].
Sources
[7]
Caveats
One trial, and the waiting-list group was not described as having events to compare [7].
Modafinil and armodafinilLimitedUncommon

Serious rash, including Stevens-Johnson syndrome

The label warns of serious rash, including Stevens-Johnson syndrome. In children's trials 0.8% stopped for rash, against none on placebo. Modafinil is not approved for children.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
In modafinil trials in patients under 17, rash leading to discontinuation occurred in about 0.8% (13 of 1,585), including 1 possible case of Stevens-Johnson syndrome and 1 multi-organ hypersensitivity reaction, against none of 380 on placebo. Median time to rash was 13 days. The label states the postmarketing reporting rate of toxic epidermal necrolysis and Stevens-Johnson syndrome exceeds the background rate of 1 to 2 cases per million person-years, and reports multi-organ hypersensitivity with at least one death [3]. Angioedema and allergic reactions were seen in armodafinil trials [3].
Population
1,585 pediatric patients in trials, and postmarketing reports in adults and children [3].
Sources
[3]
Caveats
The trial rate comes from children. The adult rate is not known, since postmarketing reports have no denominator.
Soy foodsLimitedUncommon

Soy allergy

A national survey estimated convincing soy allergy in 0.6% of US adults.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Convincing soy allergy was estimated at 0.6% (95% CI 0.5 to 0.7) of US adults, about 1.5 million people [21]. Soybeans are one of the nine major food allergens under US law [3].
Population
40,443 US adults surveyed [21].
Sources
[21], [3]
Caveats
The estimate rests on reported symptoms, not food challenge testing [21].
Whole grainsLimitedUncommon

Wheat allergy

A national survey estimated convincing wheat allergy in 0.8% of US adults.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Convincing wheat allergy was estimated at 0.8% (95% CI 0.7 to 0.9) of US adults, about 2.0 million people [17]. Wheat is one of the nine major food allergens under US law [2].
Population
40,443 US adults surveyed [17].
Sources
[17], [2]
Caveats
The estimate rests on reported symptoms consistent with allergy, not food challenge testing [17].
AllopurinolModerateRare

Severe skin reactions and hypersensitivity

The label warns of serious and sometimes fatal skin reactions in about 5 in 10,000 people, with higher risk in carriers of the HLA-B*58:01 gene.

The numbers and sources
How common
Rare: fewer than 1 in 1,000 people.
Effect
The label states that toxic epidermal necrolysis, Stevens-Johnson syndrome and drug reaction with eosinophilia and systemic symptoms occur in about 5 in 10,000 (0.05%) people taking allopurinol, mostly in the first few months, and that the HLA-B58:01 allele, more frequent in people of African, Asian and Native Hawaiian or Pacific Islander ancestry, marks higher risk [5]. In Taiwan, hypersensitivity occurred in 4.68 per 1,000 new users a year, related hospitalization in 2.02 and related death in 0.39, with higher risk in women, people 60 or older, starting doses above 100 mg a day, kidney or cardiovascular disease, and use for high uric acid without symptoms [16]. In Han Chinese, all 51 people with a severe reaction carried HLA-B5801, against 15% of 135 who tolerated the drug [17]. Pooled across 4 studies the odds ratio for Stevens-Johnson syndrome or toxic epidermal necrolysis in carriers was 96.60 (95% CI 24.49 to 381.00) [18]. In a case-control study, 91% of people with hypersensitivity syndrome and 36% of controls had started at 1.5 mg or more per unit of estimated GFR [19].
Population
495,863 new users in Taiwan's national insurance database [16]; 51 cases and 228 controls [17]; 55 cases and 678 matched controls in 4 studies [18]; 54 cases and 157 controls [19].
Sources
[5], [16], [17], [18], [19]
Caveats
The cohort rate covers all hypersensitivity, a wider group than the severe skin reactions in the label's figure, and comes from a population where the allele is frequent [16]. The band here is taken from the label's 0.05%. The label notes reactions also occur in people without the allele [5].
H2 blockersLimitedRare

Low white cells and platelets (cimetidine)

The cimetidine label reports low white cell counts in about 1 per 100,000 patients and agranulocytosis in about 3 per million, mostly in seriously ill people on other drugs.

The numbers and sources
How common
Rare: fewer than 1 in 1,000 people.
Effect
Decreased white blood cell counts occurred in about 1 per 100,000 patients, agranulocytosis in about 3 per million and low platelets in about 3 per million, with very rare aplastic anemia. Most reports were in patients with serious illnesses taking drugs known to lower blood counts [14].
Population
Postmarketing reports summarized in labeling [14].
Sources
[14]
Caveats
Rates from spontaneous reports.
LamotrigineLimitedRare

Drug reaction with organ involvement (DRESS) and liver failure

The label reports multiorgan hypersensitivity reactions, some fatal, and deaths from organ or liver failure in 2 of 3,796 adults and 4 of 2,435 children in trials.

The numbers and sources
How common
Rare: fewer than 1 in 1,000 people.
Effect
Multiorgan hypersensitivity (DRESS) has occurred, presenting with fever, rash or swollen lymph nodes alongside hepatitis, kidney inflammation, blood abnormalities or heart or muscle inflammation. Deaths from acute multiorgan failure and liver failure occurred in 2 of 3,796 adults and 4 of 2,435 children in epilepsy trials. Isolated liver failure without rash has also been reported [4].
Population
Adults and children in epilepsy trials, and postmarketing reports [4].
Sources
[4]
Caveats
Trial deaths are counted without a placebo comparison.
MirtazapineLimitedRare

Very low white blood cell count

The label reports agranulocytosis in 2 of 2,796 people in premarketing trials, and severe neutropenia in a third. All three recovered after stopping.

The numbers and sources
How common
Rare: fewer than 1 in 1,000 people.
Effect
In premarketing trials, 2 of 2,796 people treated with mirtazapine developed agranulocytosis with signs such as fever or infection, one of whom had Sjögren's syndrome, and a third developed severe neutropenia without symptoms. Onset was on days 61, 9 and 14 of treatment, and all three recovered after the drug was stopped [7].
Population
2,796 people in premarketing clinical trials [7].
Sources
[7]
Caveats
Three cases with no placebo comparison. The band above is for the two cases of agranulocytosis.
Antibiotics (common outpatient)ModerateHow often is not known

Resistance in the person treated

Linked to resistant bacteria in the person treated, about 2.5 times the odds within 2 months for urinary bacteria, lasting up to 12 months.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
After a primary care antibiotic, resistance in urinary bacteria rose (OR 2.5, 95% CI 2.1 to 2.9 within 2 months, 1.33 within 12 months), and in respiratory bacteria (OR 2.4 at both times). Longer and repeated courses were linked to more resistance [25]. Tetracycline-resistant gonorrhea occurred in 5 of 13 cases on doxycycline after sex and 2 of 16 on standard care [15], and resistance rose in some long-term COPD trials [14].
Population
14,348 participants in 5 urinary and 2,605 in 7 respiratory studies [25]; trial populations [15][14].
Sources
[25], [15], [14]
Caveats
Mostly observational [25].
Antibiotics (common outpatient)ModerateHow often is not known

Allergic reactions and penicillin to cephalosporin cross-reactions

The labels warn of serious, occasionally fatal allergic reactions. In proven penicillin allergy, cross-reaction to cephalosporins ranged from 2% to 16%, highest with a shared side chain.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The amoxicillin label reports serious and occasionally fatal anaphylaxis with penicillins, and the cephalexin label reports rash, hives, anaphylaxis and severe skin reactions [22][19]. In people with confirmed penicillin allergy, cross-reactivity was 16.45% for aminocephalosporins, which share an identical side chain with a penicillin, 5.60% for intermediate similarity and 2.11% for low similarity, and 0.87% for carbapenems [26].
Population
Label statements [22][19]; 1,269 penicillin-allergic patients in 21 studies [26].
Sources
[22], [19], [26]
Caveats
Cross-reactivity rates apply to people with proven allergy, not to everyone with a penicillin allergy label [26].
CLAModerateHow often is not known

Inflammation and oxidative stress markers

Shown to raise CRP in adults across 14 studies. Sources disagree on oxidative stress, with one marker rising and another falling.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Across 14 studies, CLA raised serum CRP by 0.63 mg/dL, the unit as the review's abstract prints it (95% CI 0.13 to 1.13, I² 52.3%), independent of dose or duration [7]. In the 12-week trial in obese men, trans-10 cis-12 CLA raised the lipid peroxidation marker 8-iso-PGF2α by 578% and CRP by 110% compared with placebo [8]. A separate pooled analysis of 22 RCTs found malondialdehyde, another oxidative stress marker, lower with CLA [9].
Population
Adults in prospective studies [7]; 60 men with metabolic syndrome randomized to trans-10 cis-12 CLA, a CLA mixture or placebo, the same design as [6] [8]; adults in RCTs [9].
Sources
[7], [8], [9]
Caveats
The size of the CRP rise is uncertain. The abstract of [7] gives it in mg/dL, and 0.63 mg/dL is 6.3 mg/L, a large change for a pooled result. The full text could not be read to confirm the unit, so the direction is reported here and the size is not relied on. Sources disagree on oxidative stress. The isoprostane rise [8] and the malondialdehyde fall [9] used different markers and mostly different products.
GarlicModerateHow often is not known

Odor, allergy and skin burns

Garlic breath and body odor were the most common complaints in trials. Case reports describe allergic reactions and skin burns from fresh garlic.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Garlic breath and body odor are the most common complaints in trials. Case reports describe allergic reactions including contact dermatitis, urticaria and anaphylaxis, and burns where fresh garlic was applied to skin under dressings [11]. The cold trial reported rash and odor [9].
Population
Trial participants and case reports [9][11].
Sources
[9], [11]
Caveats
Case reports cannot give a rate.
IronModerateStudied in childrenHow often is not known

Malaria in children in endemic areas

Clinical malaria was about as common as without iron in children in endemic areas. Where malaria services were lacking, it may rise slightly.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Across 40 trials (33,785 children), oral iron made little or no difference to clinical malaria (RR 0.93, 95% CI 0.87 to 1.00, high certainty), slightly lowered severe malaria (RR 0.90, 95% CI 0.81 to 0.98) and may not change deaths (RR 1.15, 95% CI 0.76 to 1.74, low certainty). Where malaria prevention and treatment services were lacking, iron with or without folic acid may raise clinical malaria slightly (RR 1.15, 95% CI 1.02 to 1.30, 7 trials, 19,754 children). With services in place, the RR was 0.91 (95% CI 0.84 to 0.97) [13].
Population
Children under 18 in areas of intense, year-round malaria transmission [13].
Sources
[13]
Caveats
The results split by malaria services are low certainty [13]. They do not apply to areas without malaria.
LamotrigineModerateHow often is not known

Serious rash and the HLA-B*1502 gene variant

Linked to about 2.9 times the odds of Stevens-Johnson syndrome or toxic epidermal necrolysis on lamotrigine in people carrying HLA-B*1502, mostly studied in Asian populations.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A meta-analysis found a pooled odds ratio of 2.88 (95% CI 1.60 to 5.17) for HLA-B*1502 and lamotrigine-induced Stevens-Johnson syndrome or toxic epidermal necrolysis [15]. The label states that case-control studies in people of certain Asian ancestry, such as Han Chinese and Thai, found a risk about 2 to 3 times higher, and that these reactions also occur in people without the variant [4].
Population
Case-control studies of people with lamotrigine-induced severe skin reactions [15][4].
Sources
[15], [4]
Caveats
Only HLA-B*1502 data could be pooled. The label states that many carriers never develop the reaction [4].
Pet ownershipModerateStudied in childrenHow often is not known

Childhood asthma

Mixed results: one meta-analysis linked cats and dogs to slightly more childhood asthma, another of 77,434 children found no link, and a review found less allergy.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Cat ownership (OR 1.18, 95% CI 1.05 to 1.33) and dog ownership (OR 1.12, 95% CI 1.00 to 1.24) were linked to childhood asthma [14]. In nine European birth cohorts, early-life cat (OR 0.97, 95% CI 0.87 to 1.09) and dog ownership (OR 0.92, 95% CI 0.85 to 1.01) were not linked to school-age asthma, but children already sensitized to cats or dogs had much more asthma, and ownership may add to that risk [13]. In 9 urban cohorts, 6 found less allergic disease with perinatal dog or cat exposure, and 2 found more risk only in high-risk groups [15].
Population
1,871,295 children in 60 studies, 42 of them case-control [14]. 77,434 mother and child pairs in 9 birth cohorts, children aged 5 to 11 [13]. 6,498 participants in 9 cohorts [15].
Sources
[14], [13], [15]
Caveats
Observational. Families with allergic parents often avoid pets, which confounds the link, and case-control studies can be biased by recall [15].
St. John's WortModerateHow often is not known

Cyclosporine and transplant rejection

Shown to cut cyclosporine levels about half in kidney transplant patients with a high-hyperforin extract. Two heart transplant patients had acute rejection from the interaction.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Two heart transplant patients had acute rejection from an interaction between St. John's wort and cyclosporine [9]. In 10 kidney transplant patients, a high-hyperforin extract at 900 mg a day for 14 days cut cyclosporine exposure by 52% and trough levels by 55%, and cyclosporine doses had to rise 65%. A low-hyperforin extract had no significant effect [8].
Population
2 heart transplant patients [9]. 10 kidney transplant patients in a randomized crossover [8].
Sources
[9], [8]
Caveats
The size of the interaction varied with hyperforin content [8].
Time in natureModerateHow often is not known

Lyme disease in outdoor workers

Linked to about twice the odds of Lyme disease antibodies in outdoor workers, and 2.4 times in forestry and farm workers, in one meta-analysis.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Outdoor workers had a 20.5% rate of antibodies to Borrelia burgdorferi. Outdoor work raised the odds of antibodies against controls (OR 1.93, 95% CI 1.15 to 3.23), and forestry and agricultural work more so (OR 2.36). In studies after 2010 the raised odds were not significant (OR 1.08, 95% CI 0.63 to 1.85) [15].
Population
35 studies, 15 pooled with 3,932 people [15].
Sources
[15]
Caveats
Antibodies show past infection, not illness. The studies are of people working outdoors in endemic areas, not people taking walks. Risk depends on region and season.
5-HTPLimitedHow often is not known

Eosinophilia-myalgia syndrome and peak X contaminants

Limited evidence: EMS-like symptoms have been linked to 5-HTP, and contaminants called peak X were found in tested products. Sources disagree on whether peak X matters.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Eosinophilia-myalgia syndrome (EMS) was traced in 1989 to contaminated L-tryptophan from one manufacturer [9]. EMS-like symptoms have been associated with 5-HTP, and analysis of EMS-implicated 5-HTP found "peak X," a family of contaminants. All 8 commercial 5-HTP samples tested contained three or more peak X contaminants [10]. A 2004 review reported no definitive toxicity cases in 20 years of use, apart from one unresolved case, and described peak X as a chromatographic artifact at very small concentrations [9].
Population
Product analyses and case reports [9][10].
Sources
[9], [10]
Caveats
The sources disagree on peak X. The Cochrane review states the possible link between these substances and EMS has not been elucidated [1].
AcetaminophenLimitedHow often is not known

Serious skin reactions

The label warns that acetaminophen may rarely cause serious skin reactions, including Stevens-Johnson syndrome and toxic epidermal necrolysis, which can be fatal.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The prescription combination label states that acetaminophen may rarely cause serious skin reactions such as acute generalized exanthematous pustulosis, Stevens-Johnson syndrome and toxic epidermal necrolysis, which can be fatal. It also describes postmarketing reports of hypersensitivity and anaphylaxis, with swelling of the face, mouth and throat, breathing difficulty, hives and rash [22]. The over-the-counter label carries an allergy alert for severe skin reactions, with skin reddening, blisters and rash as symptoms [19].
Population
People taking acetaminophen, as described in the labels [22][19].
Sources
[22], [19]
Caveats
The labels give no rate and cite no study.
AcupunctureLimitedHow often is not known

Infections

Case reports: infections were the most common complication in 294 published acupuncture adverse event cases from 2000 to 2011, mostly bacterial.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
From 2000 to 2011, 117 reports described 294 adverse event cases after acupuncture in 25 countries and regions. Infections, mostly mycobacterial and staphylococcal, were the main complication. An earlier review by the same team had found hepatitis from reusable needles, and none was found in this period [24]. NCCIH notes complications from nonsterile needles and improper delivery, including infections, punctured organs and nervous system injury [1].
Population
294 published cases [24].
Sources
[24], [1]
Caveats
Case reports give no rate. Only English-language reports were included [24].
AllopurinolLimitedHow often is not known

With azathioprine and mercaptopurine

The label warns that allopurinol raises exposure to azathioprine and mercaptopurine, which can suppress the bone marrow.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label states that allopurinol blocks the breakdown of mercaptopurine and azathioprine, raising exposure to either and the risk of bone marrow suppression, and that a reduced dose of those drugs is required. It also reports anemia, low white cells and low platelets on allopurinol itself, from 6 weeks to 6 years after starting, and higher cyclosporine levels when the two are combined [5].
Population
People taking these drugs together, as described in the label [5].
Sources
[5]
Caveats
The label gives no rate.
Alpha-lipoic acidLimitedHow often is not known

Low blood sugar from insulin autoantibodies (insulin autoimmune syndrome)

Case reports: six adults had episodes of low blood sugar 30 to 120 days after starting 600 mg a day, which eased after stopping.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Six patients developed episodes of fasting and post-meal hypoglycemia 30 to 120 days after starting 600 mg a day. Hypoglycemic episodes fell after stopping lipoic acid. Five carried the HLA-DRB1*04:03 allele [11].
Population
Case series of 6 European adults, median age 63 [11].
Sources
[11]
Caveats
Case reports without controls. The authors note the syndrome is linked to sulfhydryl compounds, and in Japanese patients to the HLA-DRB1*04:06 allele [11].
Antiplatelets: clopidogrel and low-dose aspirinLimitedHow often is not known

Aspirin-sensitive asthma and allergic reactions

The label warns that aspirin may cause severe hives, swelling or bronchospasm in people with the syndrome of asthma, rhinitis and nasal polyps.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The prescription aspirin label lists hypersensitivity to NSAIDs and the syndrome of asthma, rhinitis and nasal polyps as contraindications, stating that aspirin may cause severe urticaria, angioedema or bronchospasm [23].
Population
A contraindication in labeling, with no study counts [23].
Sources
[23]
Caveats
The label gives no rate.
BromelainLimitedHow often is not known

Allergy

Case reports: four laboratory workers developed allergy to bromelain, mostly from breathing it in. No count exists for people taking it by mouth.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Four workers in a blood grouping laboratory developed occupational allergy to bromelain, with mostly respiratory symptoms. Sensitization was more common among exposed workers than unexposed people and usually came from breathing it in, not swallowing it. The authors report that skin testing with bromelain can trigger systemic reactions, even at very high dilutions [10].
Population
Laboratory workers and unexposed controls [10].
Sources
[10]
Caveats
Workplace inhalation, not supplement use. No count of allergy among people taking bromelain by mouth was located.
ClonidineLimitedHow often is not known

Allergic skin reactions

The labels report that people sensitized to the clonidine patch may develop a widespread rash, hives or swelling when switched to clonidine tablets. They give no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In people with localized contact sensitization to the clonidine patch, continuing the patch or switching to oral clonidine may produce a generalized rash, urticaria or angioedema [6].
Population
Labeling statement [6].
Sources
[6]
Caveats
No rate is given.
Colon hydrotherapyLimitedHow often is not known

Infection spread by irrigation equipment

An outbreak investigation linked at least 36 cases of amebiasis to colonic irrigation at one clinic, where the machine stayed contaminated after routine cleaning.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
From 1978 to 1980, at least 36 people who had colonic irrigation at a chiropractic clinic in Colorado developed amebiasis, and 10 needed colectomy. Bloody diarrhea occurred in 21% of the irrigation group and 1% of other clinic patients, and evidence of amebic infection in 37% and 2.4%. The risk was highest for people irrigated right after someone with bloody diarrhea, and the machine was heavily contaminated with fecal bacteria after routine cleaning [14].
Population
176 clinic patients in the last 4 months of 1980, and outbreak cases from 1978 to 1980 [14].
Sources
[14]
Caveats
One outbreak at one clinic, more than 40 years ago.
CuppingLimitedHow often is not known

Burns, scars, infection and skin damage

Case reports: burns, blisters, lasting skin discoloration, scars and infections after cupping, and rare bleeding inside the skull after cupping on the scalp.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
NCCIH lists persistent skin discoloration, scars, burns and infections, worsening of eczema or psoriasis, and rare bleeding inside the skull after scalp cupping, and notes that shared equipment used without sterilization can spread hepatitis B and C [1]. A review of case reports from 2000 to 2011 found 10 adverse events tied to cupping [15]. Published cases include burns from cupping [16].
Population
Case reports [15][16].
Sources
[1], [15], [16]
Caveats
Case reports give no rate.
EchinaceaLimitedHow often is not known

Allergic reactions

Case reports: rare allergic reactions, which may be severe. Gastrointestinal upset and rash were the most frequent adverse events reported.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A systematic review of trials, spontaneous reporting programs, WHO data and manufacturer files found gastrointestinal upset and rash to be the most frequent adverse events. In rare cases, echinacea was associated with allergic reactions that may be severe [9].
Population
Clinical studies and spontaneous reports [9].
Sources
[9]
Caveats
Spontaneous reports cannot give a rate. The review found little data on use in pregnancy and lactation [9].
EggsLimitedHow often is not known

Egg allergy

Self-reported egg allergy ranged from 0.2% to 7% across studies, in a meta-analysis that found objective tests gave far lower figures than self-report.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Self-reported prevalence of egg allergy varied from 0.2% to 7%. Heterogeneity between studies was marked, and self-reported food allergy was very high against objective measures [12].
Population
51 studies of infants, school children and adults [12].
Sources
[12]
Caveats
The range spans study methods and populations. The authors recommend food challenge testing for prevalence estimates [12].
ElderberryLimitedHow often is not known

Worse flu course without oseltamivir

One trial found that, in a post hoc subgroup, flu symptoms were 2 days worse on elderberry alone than on placebo alone.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In a post hoc analysis of the 2020 trial, symptom outcomes for elderberry taken alone, without oseltamivir, were 2 days worse than for placebo taken alone [5].
Population
Subgroup of 87 emergency room flu patients [5].
Sources
[5]
Caveats
Post hoc subgroup of a small trial.
FenugreekLimitedHow often is not known

Allergy

Case reports: two people had severe immediate allergy to fenugreek with wheezing, and both also reacted to chickpea.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Two patients had severe immediate allergy to fenugreek, one with wheezing and fainting after inhaling the powder and one with facial angioedema and wheezing after scalp application. Both had a peak flow drop of more than 20% on double-blind challenge with fenugreek and with chickpea [12].
Population
Two cases from a food allergy survey [12].
Sources
[12]
Caveats
Case reports. Cross-reaction with chickpea was seen in both [12].
Fiber supplementsLimitedHow often is not known

Allergy and anaphylaxis to psyllium

Case reports: a woman died of anaphylaxis after taking a psyllium product. Workers who handle psyllium can become sensitized.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A 42-year-old woman died of anaphylaxis after ingesting a psyllium product. The literature review describes sensitization in health care workers and pharmaceutical plant workers who handle psyllium, and a link between atopy and psyllium allergy [11].
Population
One fatal case and a literature review [11].
Sources
[11]
Caveats
The review does not give a rate in people taking psyllium.
Fisetin and quercetin (senolytics)LimitedHow often is not known

Fisetin or quercetin alone at senolytic doses

One trial found quercetin shifted inflammatory proteins in a pro-inflammatory direction in women after bypass surgery, and the opposite in men.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In the coronary bypass trial, quercetin at 500 mg twice a day shifted circulating inflammatory proteins in a pro-inflammatory direction in women and increased inflammaging signatures in their artery fibroblasts, while men showed the opposite [4].
Population
97 patients, 19 of them women, treated from 2 days before surgery to hospital discharge [4].
Sources
[4]
Caveats
Exploratory analyses with few women. No trial of fisetin alone at senolytic doses was located.
GabapentinoidsLimitedHow often is not known

Severe allergic reactions

The label warns of angioedema and anaphylaxis with both drugs, and of a multiorgan hypersensitivity reaction with gabapentin that has been fatal. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The gabapentin label states that anaphylaxis and angioedema can occur after the first dose or at any time, and that drug reaction with eosinophilia and systemic symptoms, also called multiorgan hypersensitivity, has occurred, in some cases fatal or life-threatening [19]. The pregabalin label reports postmarketing cases of angioedema with swelling of the face, mouth and throat, some life-threatening [7].
Population
Postmarketing reports summarized in labeling [19][7].
Sources
[19], [7]
Caveats
Reports with no denominator.
Glucosamine and chondroitinLimitedHow often is not known

Allergic skin reactions in adverse event reports

Case reports: glucosamine products drew the most supplement reports in Singapore, mainly rash, itching, hives and swelling.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Of 727 adverse event reports for complementary health products sent to Singapore's Health Sciences Authority from 2017 to 2023, health supplements made up 68%, and glucosamine-containing products accounted for the most reports, mainly skin and hypersensitivity reactions such as rash, itching, hives and swelling [8].
Population
National spontaneous reporting data from Singapore [8].
Sources
[8]
Caveats
Spontaneous reports cannot give a rate, and wide use raises report counts. The authors note glucosamine is typically derived from shellfish [8].
GrapefruitLimitedHow often is not known

Higher blood levels of transplant drugs

The FDA states grapefruit juice can interact with cyclosporine. Small trials found higher cyclosporine exposure in volunteers and higher tacrolimus levels in liver transplant patients.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The FDA names cyclosporine (Neoral, Sandimmune) among drugs grapefruit juice can interact with [1]. With 240 mL of grapefruit juice, median dose-corrected cyclosporine AUC was 15.6 against 11.3 with orange juice (P ≤ 0.007), and juice free of furanocoumarins showed no such rise [11]. In liver transplant patients, tacrolimus blood levels rose by 1.56 and 10.33 ng/mL after a week of two kinds of grapefruit juice, and one group needed 2.3 mg a day less tacrolimus [21].
Population
18 healthy volunteers [11]. 30 liver transplant patients in 3 groups [21].
Sources
[1], [11], [21]
Caveats
Small studies. The tacrolimus trial used juice on purpose to raise levels under monitoring [21].
GrapefruitLimitedHow often is not known

Lower blood levels of fexofenadine

The FDA states grapefruit juice can cause less fexofenadine to enter the blood. In one trial, 300 mL cut its exposure to 58% of that with water.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The FDA states grapefruit juice can lower the amount of fexofenadine (Allegra) entering the blood, by acting on drug transporters, and that its label says not to take it with fruit juices [1]. With 300 mL of grapefruit juice, fexofenadine AUC fell to 58% and peak concentration to 53% of the values with water. With 1,200 mL, they fell to 36% and 33% [17].
Population
12 healthy volunteers [17].
Sources
[1], [17]
Caveats
One small crossover study measured blood levels, not allergy symptoms.
HomeopathyLimitedHow often is not known

Harms in case reports, including delayed care

Case reports: 38 reports described harms in 1,159 people, including four deaths, from remedies themselves or from homeopathy replacing conventional care.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Of 38 reports, 30 described direct harms from homeopathic remedies and 8 described harm when homeopathy replaced conventional medicine. Harms ranged from mild to severe and included four deaths. Allergic reactions and poisoning were the most common, and Rhus toxicodendron was the remedy most often implicated [22]. FDA lists products for serious or life-threatening conditions as higher risk because they may lead people to delay or stop treatments shown to work [2].
Population
1,159 people in 38 case reports and case series [22].
Sources
[22], [2]
Caveats
Case reports give no rate and cannot show how often harm occurs.
HydroxyzineLimitedHow often is not known

Serious skin reaction

The label warns hydroxyzine may rarely cause acute generalized exanthematous pustulosis, a serious rash with fever and pustules. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label describes acute generalized exanthematous pustulosis and fixed drug eruptions in postmarketing reports, and warns of cross-sensitivity with the related antihistamines cetirizine and levocetirizine [5].
Population
Postmarketing reports summarized in labeling [5].
Sources
[5]
Caveats
The label says "rarely" without a frequency, so no band can be set.
InsulinLimitedHow often is not known

Allergic reactions

The label reports severe, life-threatening generalized allergy including anaphylaxis with insulin, and redness, itching and swelling at injection sites. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The glargine label states that severe, life-threatening, generalized allergy, including anaphylaxis, generalized skin reactions, angioedema, bronchospasm, hypotension and shock, has occurred with insulin. Local reactions included redness, pain, itching, hives, edema and inflammation [5]. The insulin aspart label describes local reactions as usually self-limiting [19].
Population
Trial and postmarketing reports summarized in labeling [5][19].
Sources
[5], [19]
Caveats
No denominator is given for generalized allergy.
IsotretinoinLimitedHow often is not known

Serious skin reactions

Case reports: the label describes Stevens-Johnson syndrome and toxic epidermal necrolysis after approval, some fatal. One case led to withdrawal in trials.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label reports postmarketing erythema multiforme, Stevens-Johnson syndrome and toxic epidermal necrolysis, some resulting in death or disability, and allergic vasculitis [3]. Stevens-Johnson syndrome was the one serious adverse effect in the antibiotic comparison trials [4].
Population
Postmarketing reports [3]; trial participants [4].
Sources
[3], [4]
Caveats
Reports have no denominator.
IV vitamin therapyLimitedHow often is not known

Bloodstream infection from contaminated infusions

Case reports: previously healthy women developed bloodstream infections with environmental bacteria after IV vitamin infusions given outside hospitals, one of them nearly fatal.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A previously healthy woman developed Rahnella aquatilis bacteremia right after an unlicensed IV vitamin infusion presumed to be contaminated [10]. A 52-year-old woman with no medical history had nausea, vomiting and diarrhea and a nearly fatal Pseudomonas fluorescens bacteremia after a naturopathic IV vitamin infusion given at home [11].
Population
Two women [10][11].
Sources
[10], [11]
Caveats
Case reports give no rate.
LamotrigineLimitedHow often is not known

Hemophagocytic lymphohistiocytosis

The label reports a life-threatening immune overactivation syndrome in children and adults, starting 8 to 24 days after lamotrigine was begun. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label reports hemophagocytic lymphohistiocytosis, a syndrome of extreme systemic inflammation with high mortality if not recognized, with fever, rash, enlarged liver and spleen, organ dysfunction and blood cell abnormalities, reported 8 to 24 days after starting treatment [4].
Population
Postmarketing reports in children and adults [4].
Sources
[4]
Caveats
Reports, with no denominator.
LamotrigineLimitedHow often is not known

Blood cell disorders

The label reports low white cells, low platelets, anemia and, rarely, aplastic anemia, with or without a hypersensitivity reaction. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Reports include neutropenia, leukopenia, anemia, thrombocytopenia, pancytopenia and, rarely, aplastic anemia and pure red cell aplasia [4].
Population
Reports summarized in the label [4].
Sources
[4]
Caveats
No rate or comparison group.
LamotrigineLimitedHow often is not known

Aseptic meningitis

The label states lamotrigine raises the risk of aseptic meningitis, with symptoms starting 1 day to 6 weeks after it was begun and returning quickly on restarting.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Postmarketing cases presented with headache, fever, nausea, vomiting and neck stiffness. Most resolved after stopping. Re-exposure brought symptoms back within 30 minutes to 1 day, often more severe. Some affected people had lupus or other autoimmune diseases [4].
Population
Postmarketing reports in children and adults [4].
Sources
[4]
Caveats
Reports, with no denominator.
LiraglutideLimitedHow often is not known

Skin reactions

Limited evidence: injection-site redness, itching and bleeding were reported with GLP-1 drugs, rising with dose. Liraglutide-specific rates were not given.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A systematic review of 101 studies found injection-site reactions, including redness, itching and bleeding, most often with tirzepatide (2 to 7%), then liraglutide, then semaglutide, rising with dose. Uncommon reactions reported across these drugs included angioedema, hypersensitivity reactions, autoimmune blistering diseases, drug eruptions and vasculitis [14].
Population
32 RCTs, 6 cohort or cross-sectional studies, 20 pharmacovigilance studies and 43 case reports or series of semaglutide, liraglutide and tirzepatide [14].
Sources
[14]
Caveats
The abstract does not give liraglutide-specific rates. Case reports cannot give a rate.
Loop diureticsLimitedHow often is not known

Allergy in people with a sulfa allergy

Mixed results: the labels warn that people allergic to sulfonamides may react, while a large cohort found reactions to nonantibiotic sulfonamide drugs reflected general allergy tendency, not cross-reactivity.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The furosemide and bumetanide labels state that patients allergic to sulfonamides may also be allergic to these drugs [10][15]. In a UK cohort, 9.9% of people who had reacted to a sulfonamide antibiotic reacted to a later sulfonamide nonantibiotic, against 1.6% of people who had not (adjusted OR 2.8), but their risk after a penicillin was higher still (adjusted OR 3.9), and the authors attributed the link to a predisposition to allergy [19].
Population
969 people with and 19,257 without a prior sulfonamide antibiotic reaction [19].
Sources
[10], [15], [19]
Caveats
The cohort grouped all sulfonamide nonantibiotics and did not report loop diuretics separately [19].
MirtazapineLimitedHow often is not known

Severe skin reactions

The label reports severe skin reactions after mirtazapine, including drug reaction with eosinophilia and systemic symptoms, Stevens-Johnson syndrome and toxic epidermal necrolysis. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label states that drug reaction with eosinophilia and systemic symptoms, Stevens-Johnson syndrome, bullous dermatitis, erythema multiforme and toxic epidermal necrolysis have been reported following use [7].
Population
Reports summarized in labeling [7].
Sources
[7]
Caveats
Reports with no denominator.
MontelukastLimitedHow often is not known

Eosinophilia and Churg-Strauss vasculitis

The label reports systemic eosinophilia, sometimes with vasculitis consistent with Churg-Strauss syndrome, often as oral steroids were reduced. A causal link has not been established.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label states people with asthma on montelukast may present with systemic eosinophilia, sometimes with vasculitis consistent with Churg-Strauss syndrome, sometimes associated with reducing oral corticosteroids. It states a causal association has not been established [1].
Population
Postmarketing reports summarized in labeling [1].
Sources
[1]
Caveats
Reducing oral steroids may unmask the underlying condition [1].
N-acetylcysteineLimitedHow often is not known

Reactions to the drug form

Limited evidence: at antidote doses for acetaminophen poisoning, acetylcysteine has caused reactions from nausea to death, and intravenous dosing anaphylaxis-like reactions.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
As an antidote for acetaminophen poisoning, acetylcysteine has caused reactions from nausea to death, most deaths from incorrect dosing. Intravenous dosing causes anaphylaxis-like reactions with rash, itching, swelling and bronchospasm. Women and people with asthma or atopy were more susceptible [13].
Population
Patients treated for acetaminophen poisoning [13].
Sources
[13]
Caveats
Antidote doses are far higher than supplement doses. The review reports reaction frequency with oral dosing as at least as high as with intravenous dosing [13].
NiacinLimitedHow often is not known

Serious side effects, infection and bleeding

One trial found more serious gut, muscle, skin, infection and bleeding events with niacin plus laropiprant. EMA suspended these products in 2013.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Against placebo, niacin with laropiprant raised serious adverse events of the gut (1.0 percentage point), muscles and joints (0.7), skin (0.3), infection (1.4) and bleeding (0.7), all P of .003 or less [2]. EMA suspended these products across the EU in 2013, stating their benefits no longer outweighed their risks [10].
Population
25,673 adults with vascular disease, median 3.9 years [2].
Sources
[2], [10]
Caveats
One trial of a combination with laropiprant [2].
Why limited
This rests on one trial, of 25,673 people. Strength counts independent studies that agree, so one trial is limited at any size.
NSAIDsLimitedHow often is not known

Allergic acute coronary syndrome (Kounis syndrome) with diclofenac

Case reports: 27 people had allergic acute coronary events after diclofenac, mostly coronary spasm, and all survived to discharge.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A review of 27 published cases described acute coronary events with allergic features after diclofenac, most often coronary spasm without underlying coronary disease. Hypotension occurred in 70.4%, hives in 66.7% and chest pain in 85.2%, and all patients survived to discharge. The FDA adverse event database showed a positive reporting signal from 137 reports [22].
Population
27 case patients, 81.5% male; FAERS reports [22].
Sources
[22]
Caveats
Case reports and reporting signals cannot give a rate.
OndansetronLimitedHow often is not known

Allergic reactions

The label reports anaphylaxis and bronchospasm, including in people allergic to other drugs of the same class. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Hypersensitivity reactions, including anaphylaxis and bronchospasm, have been reported, including in people who had reacted to other 5-HT3 antagonists [3].
Population
Reports summarized in the label [3].
Sources
[3]
Caveats
Reports, with no denominator.
Oral corticosteroidsLimitedHow often is not known

Sepsis and blood clots after a short course

Linked to more sepsis and blood clots in the 30 days after starting a short course, with rate ratios of 5.30 and 3.33.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Within 30 days of starting a short course, rates rose for sepsis (incidence rate ratio 5.30, 95% CI 3.80 to 7.41) and venous thromboembolism (3.33, 95% CI 2.78 to 3.99), and fell back over the next 31 to 90 days. The rise held below 20 mg a day of prednisone equivalent (4.02 and 3.61). One in five adults in the insurance plan received a short course over 3 years [22].
Population
1,548,945 privately insured US adults, of whom 327,452 received a short course of oral corticosteroids [22].
Sources
[22]
Caveats
Observational. The illness that prompted the prescription, most often a respiratory infection, spinal condition or allergy, may account for part of the rise. The study gives rate ratios and no share of users affected.
Oral corticosteroidsLimitedHow often is not known

Raised white blood cell count

One study of 80 patients found prednisone raised the white blood cell count, in some above 20,000 per cubic millimeter, mainly through neutrophils.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In 80 patients, the white blood cell count rose on prednisone, passing 20,000 per cubic millimeter as early as the first day in some, peaking within 2 weeks in most and then falling without returning to pretreatment levels. The rise was mainly in neutrophils, with fewer eosinophils and a variable fall in lymphocytes. More than 6% band forms was rare, which the authors offer as a way to tell the change from infection [30].
Population
80 patients treated with prednisone [30].
Sources
[30]
Caveats
One uncontrolled study from 1981. The abstract gives no mean change.
Ozone therapyLimitedHow often is not known

Spinal infection after disc injection

Case reports: eight cases of infection of the disc and spine after oxygen-ozone therapy, one of them fatal.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A 55-year-old woman developed L4-L5 spondylodiscitis three months after intradiscal oxygen-ozone therapy and needed surgery and IV antibiotics. A literature review identified 8 cases in all, caused by Staphylococcus aureus, streptococci, E. coli, mycobacteria and other organisms. Outcomes were favorable in all but one, who died [14].
Population
8 published cases [14].
Sources
[14]
Caveats
Case reports give no rate.
ProbioticsLimitedHow often is not known

Bloodstream infection from the probiotic strain

Case reports: bloodstream infections matching the product strain occurred in ICU patients and 32 preterm infants, two of whom died.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
ICU patients given probiotics had a markedly higher risk of Lactobacillus bacteremia than those not given them, and whole-genome sequencing found blood isolates inseparable from the product strain, one with an antibiotic resistance mutation [48]. A review of 16 reports found 32 preterm infants with probiotic sepsis, mostly Bifidobacterium (19) and Lactobacillus (10). 25 were confirmed as the administered strain, and two infants died [49].
Population
ICU patients at one center [48]; preterm infants under 37 weeks [49].
Sources
[48], [49]
Caveats
The authors of [49] describe probiotic sepsis as uncommon in preterm infants. The abstract of [48] does not give a rate.
Proton pump inhibitorsLimitedHow often is not known

Severe skin reactions and lupus

The label warns of severe skin reactions and of cutaneous and systemic lupus reported with proton pump inhibitors, mostly cutaneous lupus. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label lists Stevens-Johnson syndrome, toxic epidermal necrolysis, DRESS and other severe skin reactions as reported. Cutaneous lupus has been reported more often than systemic lupus, as new onset or a flare, within weeks to years of continuous use [3].
Population
Postmarketing reports summarized in labeling [3].
Sources
[3]
Caveats
Reports with no denominator.
RapamycinLimitedHow often is not known

Infection and lymphoma (boxed warning)

The label warns immunosuppression may increase infection and the possible development of lymphoma and other malignancies, at transplant doses.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label's boxed warning states that immunosuppression may increase susceptibility to infection and the possible development of lymphoma and other malignancies. It reports excess mortality, graft loss and hepatic artery thrombosis in liver transplant and bronchial anastomotic dehiscence in lung transplant. The label lists latent viral infections, interstitial lung disease and impaired wound healing among its warnings [8]. In the exercise trial, 85% in each arm reported an adverse event, with 99 events on sirolimus against 63 on placebo, and one possibly drug-related pneumonia [3].
Population
Transplant recipients, summarized in the label [8]; 40 older adults [3].
Sources
[8], [3]
Caveats
The label describes transplant doses with other immunosuppressants. Aging trials were too small to measure rare serious infections.
StatinsLimitedHow often is not known

Autoimmune hepatitis

Limited evidence: statins appeared in 9% of FDA autoimmune hepatitis reports, strongest for fluvastatin. Spontaneous reports cannot give a rate or prove cause.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
In the FDA adverse event database from 2004 to 2024, 337 of 3,581 autoimmune hepatitis reports (9.41%) involved statins. Fluvastatin had the strongest signal (reporting odds ratio 54.85, 95% CI 32.32 to 93.10), and signals were stronger in people 65 and older and in women. The review also summarized 30 published cases [10].
Population
Spontaneous reports and 30 published cases [10].
Sources
[10]
Caveats
Spontaneous reports cannot give a rate or prove cause.
TopiramateLimitedHow often is not known

Serious skin and allergic reactions

The label reports Stevens-Johnson syndrome, toxic epidermal necrolysis, DRESS, anaphylaxis and angioedema with topiramate. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Serious skin reactions, multiorgan hypersensitivity (DRESS), anaphylaxis and angioedema have been reported in people taking topiramate [3].
Population
Postmarketing reports [3].
Sources
[3]
Caveats
Reports, with no denominator.
Tranexamic acidLimitedHow often is not known

Severe allergic reaction

Case reports: one severe allergic reaction occurred in the tablet trials, and anaphylactic shock has been reported after an intravenous dose.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
One woman in the trials had shortness of breath, throat tightening and facial flushing that needed emergency care. A case of anaphylactic shock after an intravenous bolus is reported in the literature [4].
Population
Trial participants and published case reports summarized in the label [4].
Sources
[4]
Caveats
Single cases. No rate is available.
Tretinoin and topical retinoidsLimitedHow often is not known

Allergic reactions

Case reports: the adapalene label lists anaphylaxis, swelling of the face, eyelids and lips, and itching after approval. The tretinoin label calls true contact allergy rare.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
Postmarketing reports with adapalene include anaphylaxis, angioedema, face and eyelid swelling, lip swelling and itching, some needing medical treatment [2]. The tretinoin label states true contact allergy is rarely encountered [13]. The tazarotene label lists hives among hypersensitivity reactions [3].
Population
Postmarketing reports [2][3]; labeling statement [13].
Sources
[2], [13], [3]
Caveats
Reports have no denominator.
ZolpidemLimitedHow often is not known

Severe allergic reactions

The label warns of angioedema of the tongue, glottis or larynx after first or later doses of sedative hypnotics including zolpidem. It gives no rate.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
The label reports cases of angioedema involving the tongue, glottis or larynx, some with shortness of breath, throat closing or nausea and vomiting suggesting anaphylaxis, and some needing emergency treatment. It states that airway obstruction may be fatal [13].
Population
Case reports summarized in labeling [13].
Sources
[13]
Caveats
Reports with no denominator. The trial database lists anaphylactic shock as rare, fewer than 1 in 1,000 [13].

Mixed results41

Absolute neutrophil count (ANC)Moderate

Variation from one draw to the next in one person

Shown to vary about 15% from one draw to the next within the same person, in a pooled estimate from a European laboratory federation.

The numbers and sources
Compared with
The same person's results on repeated venous draws
Effect
The European Federation of Clinical Chemistry and Laboratory Medicine database gives a pooled within-person variation of 14.6% for the neutrophil count, with lower and upper limits of 6.3% and 26.0%, from 9 estimates in 7 studies. Variation between people is 24.8% [16]. At a count of 3,000 cells/µL, 14.6% is about 440 cells/µL.
Population
Adults without known illness in the studies the database pools [16].
Sources
[16]
Caveats
A summary across studies of volunteers. The database states no threshold for agreement, so the result is given as numbers only [16]. A result near the edge of a range may fall on the other side of it at the next draw.
Chiropractic and spinal manipulationModerate

Immune markers in blood

Mixed results: in 6 trials, manipulation did not change lymphocyte levels against sham, while some immune markers changed briefly in people without symptoms.

Compared with sham manipulation [13]; a lecture series [13]; blood draw only [13]

The numbers and sources
Effect
Spinal manipulation was not associated with changes in lymphocyte levels or physiological markers in people with low back pain or without symptoms. It was associated with short-term changes in selected immune biomarkers in people without symptoms. The reviewers found no clinical evidence that manipulation changes immune system outcomes [13].
Population
529 participants in 6 high- and acceptable-quality trials [13].
Sources
[13]
Caveats
The reviewers call the clinical relevance of the short-term biomarker changes unknown [13].
EchinaceaModerate

Preventing colds and respiratory infections

Mixed results: a Cochrane review found no single prevention comparison significant, while pooled analyses suggested about 10% to 35% fewer infections.

The numbers and sources
Effect
A Cochrane review of 24 trials (4,631 participants) found none of 12 prevention comparisons reached significance alone, and a post hoc pooling suggested a 10% to 20% relative risk reduction [1]. A 2019 meta-analysis found a risk ratio of 0.78 (95% CI 0.68 to 0.88) for infection [2]. A third meta-analysis of 6 studies (2,458 participants) reported a risk ratio of 0.649 (95% CI 0.545 to 0.774) for recurrent infections and 0.503 for complications such as pneumonia and ear infections [3].
Population
Otherwise healthy children and adults [1][2][3].
Sources
[1], [2], [3]
Caveats
Products differed widely in species, plant part and extraction [1]. The Cochrane authors call potential effects of questionable clinical relevance [1]. Review [2] lists selective reporting and protocol deviations. Review [3] used a fixed-effect model and its abstract does not report funding.
EchinaceaModerateStudied in children

Respiratory infections in children

Mixed results: a meta-analysis found fewer infections and less antibiotic use in children, while the largest single trial found no difference in duration or severity.

The numbers and sources
Effect
A 2025 meta-analysis of 9 RCTs (1,518 children on Echinacea purpurea, 1,651 on placebo) found shorter treatment duration for upper respiratory infections (SMD -0.19, 95% CI -0.30 to -0.09), fewer episodes (RR 0.81, 95% CI 0.75 to 0.87) and less antibiotic use (RR 0.18, 95% CI 0.13 to 0.25). Ear infection episodes fell (RR 0.56, 95% CI 0.44 to 0.73), and ear infection treatment duration did not change (p 0.10) [5]. A 2025 systematic review of 5 RCTs rated the evidence high certainty for shorter symptoms and fewer antibiotics and moderate certainty for preventing infection [6]. In one trial of 407 children aged 2 to 11 with 707 upper respiratory infections, duration, severity, peak severity and days of fever did not differ between Echinacea purpurea and placebo [7].
Population
Children in 9 RCTs [5] and 5 RCTs [6]; 407 healthy children in one trial [7].
Sources
[5], [6], [7]
Caveats
The duration effect is small (SMD -0.19). Products varied in plant part and dosage form, and the funnel plot showed bias linked to heterogeneity [5]. The 5-trial review found some concerns of bias in randomization, allocation concealment, blinding and selective reporting, and did not pool results [6]. The largest single trial found no difference [7].
Erythrocyte sedimentation rate (ESR)Moderate

Detecting giant cell arteritis

Shown in a meta-analysis of 68 studies: an ESR above 100 raised the likelihood of giant cell arteritis about 3 times, and 40 or below lowered it.

The numbers and sources
Compared with
Temporal artery biopsy, imaging or clinical diagnosis
Effect
In patients suspected of giant cell arteritis, an ESR above 100 mm/hr had a positive likelihood ratio of 3.11 (95% CI 1.43 to 6.78). The absence of an ESR above 40 mm/hr had a negative likelihood ratio of 0.18 (0.08 to 0.44). The authors conclude that no single feature was strong enough to confirm or refute the diagnosis taken alone [15]. The 2022 classification criteria give 3 points for an ESR of 50 mm/hr or above or a raised C-reactive protein, out of 6 needed [9].
Population
14,037 patients with suspected giant cell arteritis in 68 studies [15].
Sources
[15], [9]
Caveats
The studies used different reference standards, and the patients were already suspected of the disease, so the figures do not apply to an ESR found on a routine test [15].
Fermented foodsModerate

Inflammation and C-reactive protein

Mixed results: one meta-analysis found fermented dairy lowered C-reactive protein slightly, while two meta-analyses of kefir trials found no change.

Compared with control groups the abstract does not describe [5]; people who did not drink kefir [6]; control groups [7]; a high-fiber diet, with changes measured over time in each group [8]

The numbers and sources
Effect
Across 14 trials of yogurt, fermented milk and kefir, C-reactive protein fell (SMD -0.21, 95% CI -0.40 to -0.02), mainly in people with metabolic disease, and TNF-alpha and interleukin-6 did not change [5]. Kefir did not change C-reactive protein (WMD -0.17 mg/L, 95% CI -0.84 to 0.49), though it fell in trials of 8 weeks or longer [6]. A second kefir review found no significant effect on inflammatory markers [7]. In a 17-week trial, a high fermented food diet steadily raised gut microbiome diversity and lowered inflammatory markers [8].
Population
14 trials [5]. 385 adults in 7 trials [6]. 24 studies [7]. 36 healthy adults, 18 in each diet group [8].
Sources
[5], [6], [7], [8]
Caveats
Few small trials with different foods. The abstract of [8] does not name C-reactive protein among the markers that fell.
High-sensitivity C-reactive protein (hs-CRP)Moderate

Repeat testing in one person

Mixed results: one study found 63% of repeat hs-CRP results stayed in the same quartile, and another found 69% of results above 3 mg/L later fell to a lower band.

The numbers and sources
Compared with
A repeat measurement in the same person
Effect
In adults measured five times at quarterly intervals, 63% of first and second hs-CRP results fell in the same quartile and 90% were within one quartile, similar to total cholesterol at 60%. The authors concluded hs-CRP is about as stable as total cholesterol [17]. In a US cohort, the intraclass correlation of CRP was 0.62 (95% CI 0.55 to 0.68), lower than 0.75 for total cholesterol. Half of participants in the highest third at baseline were in a different third at follow-up, and among those above 3 mg/L at baseline, 69% had a later result in a lower risk category [18]. FDA guidance, citing the AHA/CDC statement, calls for the average of two results about two weeks apart [2].
Population
113 healthy adults over 1 year [17]. 760 participants in the Multi-Ethnic Study of Atherosclerosis, 255 with serial values, after excluding those with conditions or drugs known to affect CRP or CRP of 10 mg/L or above [18].
Sources
[17], [18], [2]
Caveats
The two studies reach different conclusions about stability. The intervals between measurements were months to years, not the two weeks used in practice [17][18].
HomeopathyModerate

Flu and flu-like illness, Oscillococcinum

Mixed results: Oscillococcinum did not prevent flu-like illness more than placebo, and two low-quality trials found slightly more symptom relief at 48 hours.

Compared with placebo [10]

The numbers and sources
Effect
For prevention, flu-like illness did not differ from placebo (RR 0.48, 95% CI 0.17 to 1.34, 2 trials). For treatment, symptom relief at 48 hours was 7.7 percentage points more common than with placebo (RR 1.86, 95% CI 1.27 to 2.73), smaller at 3 days (RR 1.27), and not different at 4 days (RR 1.11) or 5 days (RR 1.06). The reviewers found insufficient good evidence for robust conclusions [10].
Population
6 trials: 327 young to middle-aged adults in two prevention trials in Russia, and 1,196 teenagers and adults in four treatment trials in France and Germany [10].
Sources
[10]
Caveats
Trial reporting was poor and many methods had unclear risk of bias. Only 2 treatment trials could be fully analyzed [10].
Keto dietModerate

Inflammation markers

Mixed results: CRP fell 0.62 mg/dL in one review of short and long trials, and a review of trials lasting a year or more found no CRP difference.

Compared with diets that did not restrict carbohydrates to the level of ketosis

The numbers and sources
Effect
In 5 RCTs pooled, CRP fell 0.62 mg/dL (95% CI 0.40 to 0.84) and IL-6 did not change significantly (-1.31 pg/mL, 95% CI -2.86 to 0.25) [11]. In 13 trials of 12 months or more, CRP was measured and was not among the outcomes that differed from low-fat diets [1].
Population
218 adults with overweight or obesity in 7 RCTs of 8 weeks to 2 years [11]; adults in trials against low-fat diets [1].
Sources
[1], [11]
Caveats
Few small trials. The abstract of [11] does not say whether the CRP change was measured against the control diet or from baseline. CRP is a blood marker, not a disease outcome.
LactoferrinModerateStudied in children

Sepsis in preterm infants

Mixed results: pooled trials found less late-onset sepsis in preterm infants, while the largest trial, of 2,203 infants, found no difference.

The numbers and sources
Effect
In 12 trials (5,425 infants), lactoferrin added to feeds lowered late-onset sepsis (RR 0.82, 95% CI 0.74 to 0.91, number needed to treat 25, low certainty) and shortened hospital stay by 2.38 days [1]. A 2025 meta-analysis of 10 trials found fewer infants with culture-proven or probable late-onset sepsis (OR 0.60, 95% CI 0.42 to 0.86) [2]. In the largest trial, 2,203 very preterm infants in 37 UK hospitals, late-onset infection occurred in 29% on lactoferrin and 31% on sucrose (adjusted RR 0.95, 95% CI 0.86 to 1.04) [3].
Population
Preterm infants in hospital [1][2]. Infants born before 32 weeks, given 150 mg/kg a day [3].
Sources
[1], [2], [3]
Caveats
The pooled analyses and the largest trial disagree. The Cochrane authors note publication bias and small studies of poor method that may inflate the effect [1]. Both were funded by the UK National Institute for Health Research [1][3].
LactoferrinModerate

Respiratory infections

Mixed results: one review found fewer respiratory infections in children but not adults, and another found no difference in children.

The numbers and sources
Effect
Across 25 studies, respiratory infection incidence fell in infants and children (OR 0.78, 95% CI 0.61 to 0.98) but not in adults (OR 1.00, 95% CI 0.76 to 1.32). In adults, IL-6 fell by 24.9 pg/mL, while CRP and NK cell activity did not change [4]. A 2025 meta-analysis of 8 trials in people under 18 found no significant difference in upper respiratory infections (OR 0.61, 95% CI 0.27 to 1.40) [2].
Population
20 studies in adults and 5 in children [4]. 8 trials in children [2].
Sources
[4], [2]
Caveats
The two reviews differ on children [4][2]. The authors of the first describe the clinical studies as limited. The review was supported by an Australian government grant in partnership with Freedom Foods Group Nutritionals [4].
Lymphatic drainage massageModerate

Lymphedema after breast cancer, compared with no drainage

Mixed results: in one review, drainage added 7% more swelling reduction to compression bandaging, while two later meta-analyses found no significant reduction in arm volume.

Compared with compression bandaging alone [1]; standard physiotherapy alone [1]; control groups without drainage [3][4]

The numbers and sources
Effect
Added to compression bandaging, drainage gave an additional 7.11% reduction in swelling (95% CI 1.75% to 12.47%, 2 trials, 83 participants), and added to standard physiotherapy it did not differ. Pain and heaviness eased in both groups with no difference between them [1]. A later review found no significant effect on volume or quality of life, and lower pain (SMD -0.72, 95% CI -1.34 to -0.09) [3]. Another found no significant reduction in lymphedema (SMD -0.09, 95% CI -0.85 to 0.67, 8 trials, 338 patients) [4].
Population
6 trials of 24 to 45 women each [1]. 1,564 patients in 11 trials [3]. 338 patients in 8 trials [4].
Sources
[1], [3], [4]
Caveats
Trials were small, and in four of the Cochrane trials the person measuring swelling knew the treatment [1]. None of these comparisons used placebo drainage.
29 more
Lymphatic drainage massageModerate

Preventing lymphedema after breast cancer surgery

Mixed results: the largest trial found arm lymphedema in 24% with drainage and 19% without at one year, while pooled analyses disagreed about prevention.

Compared with the same guidelines and exercise therapy without drainage [5]; control groups without drainage [3][4]

The numbers and sources
Effect
After axillary lymph node removal, 12-month incidence of arm lymphedema was 24% with drainage and 19% without (OR 1.3, 95% CI 0.6 to 2.9), and time to lymphedema did not differ [5]. Pooling 4 trials with 1,364 patients, drainage did not significantly lower the risk (RR 0.61, 95% CI 0.29 to 1.26) [4]. Another review found a lower incidence (RR 0.58, 95% CI 0.37 to 0.93) [3].
Population
160 patients after breast cancer surgery with axillary node dissection in Belgium [5]. 1,364 patients in 4 trials [4]. Patients in 11 trials [3].
Sources
[5], [4], [3]
Caveats
The two pooled analyses overlap in their trials and disagree. The largest single trial was designed to detect a large effect [5].
Massage therapyModerate

Stress hormones, blood pressure and immune cells after one session

Mixed results: a single session lowered blood pressure and heart rate in pooled trials, and cortisol fell in one small trial but not in pooled trials.

Compared with control conditions in randomized studies the source abstract does not describe [12]; light touch [18]

The numbers and sources
Effect
Single sessions reduced state anxiety, blood pressure and heart rate, and did not reduce cortisol [12]. In healthy adults, 45 minutes of Swedish massage produced a large drop in arginine vasopressin and a small drop in blood cortisol against light touch. Circulating lymphocytes rose (effect sizes 0.14 to 0.43) and stimulated cytokine production fell [18].
Population
37 randomized studies [12]. Healthy adults aged 18 to 45 [18].
Sources
[12], [18]
Caveats
The trial calls its data preliminary, and these were changes over one session [18]. Lymphocyte counts in blood do not show fewer infections.
Olive oilModerate

Inflammation

Mixed results: one pooled analysis found C-reactive protein fell 0.64 mg/L with olive oil, while a later one found no change in C-reactive protein or interleukin-6.

Compared with control groups, with control diets that varied [12]; trials comparing olive oil or flaxseed oil with controls [13]

The numbers and sources
Effect
Olive oil lowered C-reactive protein (mean difference -0.64 mg/L, 95% CI -0.96 to -0.31, 15 trials) and interleukin-6 (-0.29) and raised flow-mediated dilation by 0.76% against controls [12]. A pooling of trials in adults with normal or mildly raised blood pressure found no effect on C-reactive protein (8 trials) or interleukin-6 (3 trials) [13].
Population
3,106 participants in 30 randomized trials [12]. 17 randomized trials, 14 of olive oil [13].
Sources
[12], [13]
Caveats
The first review calls for conservative reading because olive oil was given as a supplement or within a diet and control diets varied [12].
ProbioticsModerate

Rheumatoid arthritis

Mixed results: reviews agree probiotics lower CRP or IL-6 in rheumatoid arthritis, but disagree on disease activity.

The numbers and sources
Effect
Across 9 studies (361 patients), probiotics lowered IL-6 (SMD -0.708) and did not change disease activity scores (MD 0.023, 95% CI -0.584 to 0.631) [8]. Across 4 RCTs (153 participants), DAS fell (SMD -0.58, 95% CI -0.97 to -0.19), CRP fell borderline (SMD -0.32, 95% CI -0.65 to 0.00), and ESR and cytokines did not change [9]. A review of 8 RA trials (344 patients) found CRP 3.04 mg/L lower (95% CI -4.47 to -1.62). With high risk-of-bias trials excluded, DAS28 did not differ from placebo (SMD -0.54, 95% CI -1.94 to 0.85) [10].
Population
Adults with rheumatoid arthritis, mostly women, over 8 weeks to 1 year [8][9][10].
Sources
[8], [9], [10]
Caveats
The reviews agree on lower CRP or IL-6 and disagree on disease activity. Strains, doses and durations varied [10].
ProbioticsModerate

Upper respiratory infections

Mixed results: one review found fewer respiratory infections and shorter episodes, while a later review found no significant drop in infections.

The numbers and sources
Effect
Across 23 trials (6,950 children, adults and older people), probiotics reduced the number with at least one infection (RR 0.76, 95% CI 0.67 to 0.87, low certainty) and at least three infections (RR 0.59, moderate certainty), shortened episodes by 1.22 days (95% CI -2.12 to -0.33), and reduced antibiotic use (RR 0.58) [15]. A 2026 review of 32 RCTs found no significant pooled drop in infection incidence (risk difference -0.07, 95% CI -0.23 to 0.09), while some strains shortened symptoms and cut antibiotic use [16]. Across 13 RCTs of Lacticaseibacillus rhamnosus GG, respiratory episodes fell (MD -0.14, 95% CI -0.27 to -0.01) [17].
Population
People in community, care facilities, schools and hospitals [15]. RCTs of oral or topical probiotics [16]. People given LGG for prevention or treatment [17].
Sources
[15], [16], [17]
Caveats
Most outcomes were low certainty [15]. Heterogeneity in [16] was very high (I² 93%) and certainty very low. The reviews disagree on incidence.
ResveratrolModerate

Inflammation and blood lipids

Mixed results: reviews disagree on whether CRP falls, and most found no change in LDL or triglycerides.

The numbers and sources
Effect
One meta-analysis found no change in CRP (WMD -0.144 mg/L, 95% CI -0.968 to 0.680), total cholesterol, LDL, triglycerides or glucose. HDL fell slightly (WMD -4.18 mg/dL, 95% CI -6.54 to -1.82) [4]. A later meta-analysis found lower TNF-α (WMD -0.44, 95% CI -0.71 to -0.16) and hs-CRP (WMD -0.27, 95% CI -0.5 to -0.02), with no change in IL-6 [6]. In 25 RCTs in metabolic disorders, CRP fell (SMD -0.18, 95% CI -0.33 to -0.03), and IL-6 and TNF-α did not change significantly [7]. In 6 RCTs in type 2 diabetes (533 participants), CRP fell (SMD -1.40, 95% CI -2.60 to -0.21) [8]. Apolipoprotein B (WMD 2.11 mg/dL, 95% CI -1.53 to 5.75) and apolipoprotein A-I did not change in 7 RCTs [9]. An umbrella review rated as high certainty a 0.19 mmol/L drop in total cholesterol in adults with overweight (95% CI 0.06 to 0.32) [5].
Population
10 RCTs [4]; 17 RCTs with 736 adults [6]; adults with metabolic disorders [7]; adults with type 2 diabetes [8]; 7 RCTs [9]; meta-analyses of RCTs [5].
Sources
[4], [6], [7], [8], [9], [5]
Caveats
The reviews disagree on CRP, and the CRP effect in [8] is several times the size of that in [7]. Heterogeneity in [6] was about 50%. Evidence in [8] was low certainty. CRP is a marker, not a clinical outcome [7].
SeleniumModerate

Hashimoto's thyroiditis

Mixed results: thyroid antibodies fell in people with Hashimoto's in both reviews, while reviews disagree on whether TSH changed.

The numbers and sources
Effect
Thyroid peroxidase antibodies fell in both reviews. In people on levothyroxine, the weighted mean difference after 3 months was -271 (95% CI -366 to -175) [5]. Across 29 cohorts, the SMD was -0.96 (95% CI -1.36 to -0.56) [6]. On TSH, sources disagree. In people not on thyroid hormone, a 2017 review found no change in TSH, quality of life or ultrasound findings [7]. A 2024 review found TSH lower (SMD -0.21, 95% CI -0.43 to -0.02, 869 people) [6].
Population
Adults with chronic autoimmune thyroiditis, 16 trials [5], 9 trials [7] and 35 studies [6].
Sources
[5], [6], [7]
Caveats
Antibody levels are a lab marker. The 2016 review rated evidence quality low [5], and the 2024 review moderate, with I² of 90% for antibodies [6].
Vitamin CModerate

Catching colds and how long they last

Not shown to prevent colds in the general population, though colds were 8% shorter in adults. Risk halved in people under heavy physical stress.

The numbers and sources
Effect
In general community trials, regular vitamin C did not change the risk of catching a cold (RR 0.97, 95% CI 0.94 to 1.00). In marathon runners, skiers and soldiers on subarctic exercises, the risk halved (RR 0.48, 95% CI 0.35 to 0.64). Regular supplementation shortened colds by 8% in adults (95% CI 3% to 12%) and 14% in children. Taken only once a cold had started, vitamin C showed no consistent effect on duration or severity [1].
Population
29 trial comparisons with 11,306 participants for incidence, 10,708 of them in community trials and 598 under heavy physical stress. 31 comparisons covering 9,745 cold episodes for duration [1].
Sources
[1]
Caveats
Trials used at least 0.2 g a day. The therapeutic trials were few, 7 comparisons [1].
Vitamin DModerate

Lupus disease activity

Mixed results: an earlier review found no significant change in lupus activity, while later reviews found modestly lower activity in adults with lupus.

The numbers and sources
Effect
Across 5 RCTs (490 participants), SLEDAI disease activity scores did not differ significantly from placebo (SMD -0.507, 95% CI -1.055 to 0.041), and fatigue scores fell in 2 trials (SMD -1.179) [40]. Later meta-analyses found lower disease activity: SMD -0.54 across 10 RCTs (847 participants) [41], and SLEDAI MD -1 (95% CI -2 to -0.43) in a review of 21 studies [42].
Population
Adults with systemic lupus erythematosus [40][41][42].
Sources
[40], [41], [42]
Caveats
The earlier and later reviews disagree.
White blood cell count (WBC)Moderate

Variation from one draw to the next in one person

Shown to vary about 11% from one draw to the next within the same person, in a pooled estimate from a European laboratory federation.

The numbers and sources
Compared with
The same person's results on repeated venous draws
Effect
The European Federation of Clinical Chemistry and Laboratory Medicine database gives a pooled within-person variation of 11.1% for the white cell count, with lower and upper limits of 8.9% and 15.9%, from 9 included estimates. Variation between people is 17.3%. For the neutrophil count the within-person figure is 14.6% [12]. At a count of 7.0 thousand/µL, 11.1% is about 0.8 thousand/µL.
Population
Healthy adults in the studies the database pools [12].
Sources
[12]
Caveats
A summary across studies in healthy volunteers. The database states no threshold for agreement, so the result is given as numbers only [12]. A result near the edge of a range may fall on the other side of it at the next draw.
Whole grainsModerate

C-reactive protein and inflammation

Mixed results: two pooled analyses of trials found whole grains lowered C-reactive protein slightly, while a third found no significant change.

Compared with refined grains [11]; refined grains or placebo [9]; control diets low in or without whole grains [10]

The numbers and sources
Effect
Whole grains lowered high-sensitivity C-reactive protein (Hedges' g -0.22, 95% CI -0.45 to 0.00) and interleukin-6, with no change in TNF-alpha, and high heterogeneity (I² 69%) [11]. Another review found lower C-reactive protein (SMD -0.22, 95% CI -0.44 to 0.00) [9]. A third found no significant change in C-reactive protein (-0.29 mg/L, 95% CI -1.10 to 0.52), interleukin-6 or TNF-alpha, with effects in people with raised C-reactive protein and in trials with matched calories [10].
Population
466 participants in 13 trials [11]. 25 trials [9]. 1,238 adults in 14 trials [10].
Sources
[11], [9], [10]
Caveats
Heterogeneity was high across reviews [10][11]. The intervals of the two reviews that found a fall reach zero.
Zinc levelModerate

Effect of infection and inflammation on the result

Mixed results: in pooled surveys, low zinc became more common as inflammation markers rose in half of the preschool child surveys, and the relation in women was weak and inconsistent.

The numbers and sources
Compared with
Zinc in people in the same surveys with lower inflammation markers
Effect
A technical document reports that serum zinc is reduced during acute infection and inflammation, likely because zinc moves from plasma to the liver, and that low albumin also lowers it [1]. In pooled survey data, the estimated prevalence of zinc deficiency rose with increasing C-reactive protein in 6 of 12 surveys of preschool children. In women of reproductive age the association with inflammation markers was weak and inconsistent. Where an adjustment was applied in children, it lowered the estimated prevalence of deficiency by a median of 11 percentage points (range 4 to 18) [11]. In a US survey, low albumin raised the odds of falling below the serum zinc cut-off (OR 11.2, 99% CI 3.4 to 37.3) [9].
Population
18,859 preschool children and 22,695 non-pregnant women aged 15 to 49 in 13 nationally representative surveys [11]. 4,347 people in a US survey [9].
Sources
[1], [11], [9]
Caveats
The survey authors found the relations inconsistent and advise checking for an association before any adjustment. They found too little evidence to warrant adjusting results in women [11].
BenzopyronesLimited

Lymphedema of the arms and legs

Small trials found opposite results: coumarin reduced excess limb fluid in one crossover, while a trial in 140 women after breast cancer found no change in arm volume.

The numbers and sources
Effect
A Cochrane review included 15 trials, 8 of them of coumarin. Meta-analysis was not done because of poor trial quality, and the authors could not draw conclusions about limb volume, pain or discomfort [1]. In a crossover trial of 400 mg a day against placebo for six months each, coumarin reduced excess arm fluid from 46% to 26% above normal and excess leg fluid from 25% to 17% (p < 0.001), and 93% of patients preferred it [2]. In 140 women with arm lymphedema after breast cancer, 200 mg twice a day against placebo for six months each did not change arm volume. The affected arm grew by 21 mL during placebo and 58 mL during coumarin (P 0.80), and 15% on coumarin against 10% on placebo said the medication helped a moderate or large amount (P 0.19) [3]. A 2022 review of lymphedema drug trials noted that earlier meta-analyses had cast doubt on the positive benzopyrone results [4].
Population
31 people with arm lymphedema after mastectomy and 21 with leg lymphedema in Australia [2]. 140 women with arm lymphedema after breast cancer treatment in the US [3]. Trials in the Cochrane review included five in India or China that added anti-filarial drugs [1].
Sources
[1], [2], [3], [4]
Caveats
The two most detailed trials reached opposite results [2][3]. Three trials in the Cochrane review involved the same researcher [1]. The Australian trial authors note this coumarin is not an anticoagulant [2].
ElderberryLimited

Flu duration

Small trials found mixed results: one saw flu symptoms ease about 4 days sooner, and a later trial of 87 patients saw no difference.

The numbers and sources
Effect
In 60 Norwegian adults with flu symptoms for 48 hours or less, 15 mL of elderberry syrup four times a day for 5 days relieved symptoms about 4 days sooner than placebo, with less rescue medicine [4]. In a 2020 trial of 87 emergency room patients with PCR-confirmed flu, days until symptoms were none or mild were 5.3 on elderberry and 4.9 on placebo (p 0.57), and days to full resolution were 8.6 and 8.7 [5]. The systematic review rated the flu duration evidence uncertain [2].
Population
60 adults aged 18 to 54 [4]. 87 patients aged 5 and older [5].
Sources
[4], [2], [5]
Caveats
The two trials disagree. The 2020 trial was powered above 0.90 to detect a 2-day difference and was funded by the Wendel Family Foundation and Cleveland Clinic [5]. The abstract of [4] does not report funding.
Fish oilLimited

Inflammation markers (CRP)

Limited evidence: CRP fell in hemodialysis patients but not with high-dose EPA, in knee osteoarthritis, or in most trials of trained people.

The numbers and sources
Effect
In 13 hemodialysis trials (678 patients), omega-3 in triglyceride form lowered CRP (SMD -0.62, 95% CI -1.22 to -0.03, I² 74%), and so did doses under 2,000 mg a day (SMD -0.32) [32]. In 3 trials of trained people, only 1 found a significant drop in CRP after exercise [33]. High-dose EPA did not change hs-CRP across 6 outcome trials [9], and krill oil did not change CRP in knee osteoarthritis [29].
Population
Hemodialysis patients [32]; trained individuals [33]; cardiovascular and osteoarthritis trials [9][29].
Sources
[32], [33], [9], [29]
Caveats
The reviews disagree by population. In rheumatoid arthritis, only leukotriene B4 of five inflammatory markers fell [11].
Fish oilLimited

Preventing autoimmune disease

One trial found autoimmune disease was not significantly lower on 1,000 mg a day over 5.3 years. Two more years of follow-up found 17% fewer cases.

The numbers and sources
Effect
In VITAL, 1,000 mg a day of marine omega-3 for a median 5.3 years led to confirmed autoimmune disease in 130 people against 148 on placebo (HR 0.85, 95% CI 0.67 to 1.08, P 0.19) [36]. With 2 more years of follow-up after the trial, the HR was 0.83 (95% CI 0.70 to 0.99), with 234 cases against 280 [37].
Population
25,871 US men 50 and older and women 55 and older [36]. 21,592 followed for 2 more years [37].
Sources
[36], [37]
Caveats
Both sources are the same trial. The randomized result did not reach significance, and the later result includes cases after the trial ended [36][37].
Why limited
This rests on one trial, of 25,871 people. Strength counts independent studies that agree, so one trial is limited at any size.
Grape seed extractLimited

Inflammation

Limited evidence: a 50-trial review found lower CRP, while an earlier review found no effect.

The numbers and sources
Effect
Across 50 trials, C-reactive protein fell (WMD -0.81, 95% CI -1.25 to -0.38) [5]. An earlier meta-analysis found no significant effect on C-reactive protein [2].
Population
Adults in RCTs [5][2].
Sources
[5], [2]
Caveats
C-reactive protein is a blood marker, not a symptom. The two reviews disagree.
GroundingLimited

Inflammation and white blood cells

Limited evidence: small company-funded trials found white blood cell counts shifted after damaging exercise with grounding. No study measured C-reactive protein or an inflammatory disease.

Compared with sham grounding [4][5]

The numbers and sources
Effect
On day 3 after exercise, neutrophils were higher in grounded than sham-grounded men (P = 0.05), and platelets rose significantly in the grounded group on days 2 to 4 [4]. White blood cell counts were among the measures that differed by 10% or more in the 8-person pilot [5]. A review by researchers holding shares in the sponsoring company states that grounding alters circulating neutrophils, lymphocytes and inflammatory factors [6].
Population
32 healthy young men [4]. 8 healthy adults [5]. Studies summarized in a review [6].
Sources
[4], [5], [6]
Caveats
A higher neutrophil count is not a lower level of inflammation, and the trials did not say whether these shifts help or harm. The review's authors disclose contracts with and shares in the company sponsoring grounding research [6].
H2 blockersLimited

COVID-19

Limited evidence: famotidine was not linked to fewer deaths in hospitalized COVID-19 patients, and one small outpatient trial found faster symptom improvement but no difference in time to resolution.

Compared with patients not given famotidine [9]; placebo [10]

The numbers and sources
Effect
In hospitalized patients, famotidine use was not associated with the composite of death, intubation or intensive care (HR 0.63, 95% CI 0.35 to 1.16) or with death (HR 0.67, 95% CI 0.26 to 1.73) [9]. In outpatients with mild to moderate illness, time to symptom resolution did not differ, but a 50% drop in symptom scores came at 8.2 days on famotidine and 11.4 days on placebo [10].
Population
4 observational studies with 46,435 patients, 3,110 on famotidine [9]; 55 non-hospitalized adults, median age 35, randomized [10].
Sources
[9], [10]
Caveats
The pooled studies were observational with heterogeneity of 84% to 88% [9]. The trial was a small phase 2 study [10].
High-sensitivity C-reactive protein (hs-CRP)LimitedStudied in women

Endometriosis

Mixed results: six studies in a Cochrane review found no CRP difference with endometriosis, while a US survey linked endometriosis to 60% higher odds of CRP above 3 mg/L.

The numbers and sources
Effect
A Cochrane review of blood tests for endometriosis found six studies that showed no difference in CRP or hs-CRP between women with endometriosis confirmed at surgery and women without it, across stages I to IV. One used an hs-CRP assay in 834 participants and the others a standard CRP assay in 499. The reviewers concluded that CRP measured either way is not a reliable marker of endometriosis [23]. In a national US survey, women who reported a diagnosis of endometriosis had greater odds of CRP above 3.0 mg/L than women without it (adjusted odds ratio 1.60, 95% CI 1.05 to 2.45) [24].
Population
1,333 participants in six studies, within a review whose studies enrolled women having surgery for suspected endometriosis, infertility or an ovarian mass [23]. 2,342 women in the National Health and Nutrition Examination Survey, 1999 to 2002 [24].
Sources
[23], [24]
Caveats
Observational. In the survey, endometriosis was self-reported, and the abstract does not list what the model adjusted for, so body weight may not be accounted for [24]. The Cochrane review rated all its included studies as of poor methodological quality [23].
HomeopathyLimited

Hay fever, allergies and allergic asthma

Mixed results: early small trials found less nasal blockage than placebo, and later trials in allergic asthma and hay fever found no difference.

Compared with placebo [5][12][13][14]

The numbers and sources
Effect
In an early pooled analysis, four trials of one remedy for seasonal allergies favored homeopathy for eye symptoms at 4 weeks (OR 2.03, 95% CI 1.51 to 2.74) [5]. In 51 people with perennial allergic rhinitis, a 30c dilution of their main allergen improved nasal airflow by 19.8 L/min more than placebo (95% CI 10.4 to 29.1), and symptom scores did not differ [12]. In 242 people with asthma allergic to house dust mite, homeopathic immunotherapy did not differ from placebo on most outcomes over 16 weeks [13]. In 62 people with seasonal allergic rhinitis, neither a standard nor an individualized remedy improved quality of life more than placebo [14].
Population
4 trials [5]. 51 adults in Scotland [12]. 242 adults in 38 English general practices [13]. 62 adults in Germany [14].
Sources
[5], [12], [13], [14]
Caveats
The 2026 trial reached about a quarter of its planned size [14]. The 51-person trial was run by the group whose three earlier trials it added to [12].
Hyperbaric oxygen therapyLimited

Telomere length and senescent cells

Limited evidence: blood cell telomeres lengthened by over 20% after 60 sessions in 35 older adults, in a study with no control group and no health outcome.

The numbers and sources
Effect
Telomere length in T helper, T cytotoxic, natural killer and B cells rose by over 20% after 60 daily sessions, with B cells up 37.63% ± 52.73 after treatment. Senescent T helper cells fell by 37.30% ± 33.04 and senescent T cytotoxic cells by 10.96% ± 12.59 [16].
Population
35 healthy, independently living adults aged 64 and older [16].
Sources
[16]
Caveats
No control group. The standard deviations are as large as the changes. Telomere length in blood cells is a lab marker, and the study did not measure lifespan, disease or function. Blood was last sampled 1 to 2 weeks after the final session.
MeditationLimited

Inflammation and C-reactive protein

Mixed results: one meta-analysis found C-reactive protein fell against active controls, while a review of trials found a fall in three of six that measured it.

Compared with active control groups [17]; waiting list, usual care and active control groups [18]

The numbers and sources
Effect
Across 45 trials against active controls, meditation reduced cortisol, C-reactive protein, blood pressure, heart rate, triglycerides and tumor necrosis factor alpha [17]. In a systematic review of 20 randomized trials, six measured C-reactive protein. One found a reduction and two found a trend toward reduction. The reviewers called the immune findings tentative and heterogeneous [18].
Population
Participants in 45 randomized trials [17]. Participants in 20 randomized trials, including people with cancer, HIV and ulcerative colitis [18].
Sources
[17], [18]
Caveats
The meta-analysis abstract gives no effect sizes [17]. Assay methods and populations varied widely [18].
MontelukastLimited

Chronic hives

Limited evidence: most trials found leukotriene blockers alone no better than placebo or antihistamines for chronic hives, while adding one to an antihistamine appeared to help more.

Compared with placebo or antihistamines [6]

The numbers and sources
Effect
Across 10 randomized trials, most found leukotriene receptor antagonists not superior to placebo or antihistamines as monotherapy, while the combination with antihistamines appeared to improve hives more than antihistamines alone. Heterogeneity was too high for a pooled analysis [6].
Population
10 randomized or crossover trials in chronic urticaria [6].
Sources
[6]
Caveats
No pooled estimate. Trials differed in design and outcomes [6].
Osteopathy and osteopathic manipulative treatmentLimitedStudied in children

Ear infections in children

Mixed results: one trial found fewer ear infections and surgeries when treatment was added to routine care, and a trial against sham treatment found no difference.

Compared with routine pediatric care [12]; sham osteopathic treatment [13]

The numbers and sources
Effect
In 57 children with recurrent acute ear infections, adding osteopathic treatment to routine care led to fewer episodes (mean difference -0.14 a month, 95% CI -0.27 to 0.00), fewer surgical procedures (1 against 8) and more normal tympanograms [12]. In 90 otitis-prone children, up to five osteopathic treatments did not significantly change the risk of an ear infection over 6 months compared with sham (44% against 61%, RR 0.72, 95% CI 0.48 to 1.10) [13].
Population
57 children aged 6 months to 6 years [12]. 90 children aged 12 to 60 months in Arizona [13].
Sources
[12], [13]
Caveats
Two small trials. In the first, the pediatrician was blinded but the families were not [12].
PalmitoylethanolamideLimited

Inflammation markers (CRP, IL-6)

Small trials found lower IL-6 and CRP in diabetic nerve pain, while a COVID-19 trial lowered other markers but not those two.

The numbers and sources
Effect
In 70 people with diabetic nerve pain, IL-6 and raised CRP fell more on 600 mg a day than on placebo (p 0.05), and fibrinogen did not differ [5]. In 60 unvaccinated adults recently diagnosed with COVID-19, 600 mg twice a day for 4 weeks lowered soluble P-selectin, IL-1 beta and IL-2 against placebo. IL-6 and CRP were primary outcomes and are not among the markers reported as reduced [9].
Population
70 adults with type 1 or type 2 diabetes [5], and 60 adults aged 19 to 53 with COVID-19 not in hospital [9].
Sources
[5], [9]
Caveats
Two small trials. A systematic review of 47 RCTs lists 10 that measured blood or tissue markers but gives no pooled estimate [10].
ReishiLimited

Immune function

One small trial found changes in lab immune markers in older women, without measuring infections or illness.

The numbers and sources
Effect
In 39 older women who completed 8 weeks of 2,000 mg a day of dry extract, stimulated lymphocyte proliferation rose, Th17 cells fell and Th2 cells rose, with changes in several immune gene expressions against placebo [5]. In cancer trials, T-cell percentages rose modestly [1].
Population
60 older women enrolled, 39 completed [5]. Cancer patients [1].
Sources
[5], [1]
Caveats
Lab markers, not infection or illness rates. High dropout in [5].
SaunaLimitedStudied in men

Inflammation markers

Mixed results: frequent sauna users had lower C-reactive protein in a study of 2,084 Finnish men, while a 4-week trial in 15 men found no change.

The numbers and sources
Effect
Mean C-reactive protein was 2.41, 2.00 and 1.65 in men using a sauna once, 2 to 3 times and 4 to 7 times a week, an inverse association that held after adjustment [23]. In 15 untrained young men randomized to cycling 3 times a week with or without a 30-minute sauna afterward, high-sensitivity C-reactive protein did not differ between groups after 4 weeks. Gut bacteria and intestinal barrier markers did not differ either [24].
Population
2,084 Finnish men aged 42 to 60 without acute or chronic inflammation, measured once [23]. 15 men aged 22 on average [24].
Sources
[23], [24]
Caveats
The larger study is cross-sectional, a single measurement that cannot show sauna use came first. Its abstract gives the unit as mmol/L, and C-reactive protein is usually reported in mg/L, so the unit may be a reporting error [23]. The trial was a pilot [24].
ZincLimited

Infections other than colds

Limited evidence: one trial found fewer infections in children with sickle cell anemia, while another found no change in deaths in infants with sepsis.

The numbers and sources
Effect
In children with sickle cell anemia, 20 mg zinc sulfate a day for 6 months led to 80 infections against 124 on placebo (305.7 against 480.7 per 100 person-years, IRR 0.62, 95% CI 0.45 to 0.86) [11]. In infants hospitalized with suspected sepsis, 10 mg a day for 14 days with antibiotics did not change death in hospital (4.1% against 4.9%, RR 0.83, 95% CI 0.60 to 1.15) or by 12 weeks (9.0% against 8.6%, RR 1.05, 95% CI 0.84 to 1.32) [12]. In adults hospitalized with COVID-19, ICU admission or death occurred in 5.7% on 90 mg zinc acetate a day and 23.5% on standard care alone (OR 0.21, 95% CI 0.03 to 0.96) [13].
Population
100 children aged 1 to 4 in Uganda, all on hydroxyurea [11]; 3,153 infants aged 3 to 59 days in India and Nepal [12]; 71 adults in one Spanish hospital [13].
Sources
[11], [12], [13]
Caveats
Each result comes from one trial in a different population. The sepsis trial was underpowered because events were fewer than expected, and vomiting was slightly more common on zinc [12]. The COVID-19 trial was small and open-label [13].

Not shown to help28

EchinaceaStrong

How long a cold lasts

Not shown to shorten colds in people with new colds. Only 1 of 7 treatment trials found echinacea better than placebo.

The numbers and sources
Effect
Of 7 treatment trials reporting cold duration, only 1 found echinacea better than placebo [1]. A 2019 meta-analysis found a mean difference of -0.45 days, with a confidence interval crossing zero [2]. In 719 patients aged 12 to 80, blinded echinacea shortened colds by a nonsignificant 0.53 days (95% CI -1.25 to 0.19) and lowered global severity by a nonsignificant 28 points [4].
Population
People with new-onset colds [1][2][4].
Sources
[1], [2], [4]
Caveats
The 719-person trial had more variability than expected, which limited power to detect small effects. It was funded by the NIH's National Center for Complementary and Alternative Medicine [4].
FerritinStrong

Effect of inflammation on the result

Shown to read about 30% higher when inflammation markers are raised, which hides iron deficiency in a share of people.

The numbers and sources
Compared with
Ferritin in people in the same studies without raised inflammation markers
Effect
In a meta-analysis of 32 studies of apparently healthy people, ferritin was 49.6% higher with C-reactive protein above 5 mg/L and 38.2% higher with raised alpha-1-acid glycoprotein. Across stages it was 30% higher in the incubation stage, 90% higher in early convalescence and 36% higher in late convalescence. Overall, inflammation raised ferritin by about 30% and led to a 14% underestimate of iron deficiency (95% CI 7 to 21) [14]. In pooled survey data, the share of people with depleted iron stores was 4% in the highest tenth of C-reactive protein and 30% in the lowest among preschool children, and 6% against 29% among women of reproductive age. Adjusting for inflammation raised the estimated prevalence by 7 to 25 percentage points in children and 2 to 8 points in women [15]. WHO sets higher deficiency cut-offs for people with infection or inflammation for this reason [2].
Population
8,796 infants, children, men and women in 32 studies [14]. 27,865 preschool children in 15 surveys and 24,844 women in 8 surveys [15].
Sources
[14], [15], [2]
Caveats
The surveys' authors state that more work is needed to validate the adjustment methods, and WHO notes that different adjustment approaches give widely varying prevalence estimates [15][2].
NutsStrong

Inflammation and blood vessel function

Not shown to lower C-reactive protein in two meta-analyses of nut trials. Blood vessel widening, a measure of vessel function, improved in one.

Compared with control conditions the abstract does not describe [13]; control arms without nuts [4]

The numbers and sources
Effect
Across 25 trials, nuts made no significant difference to C-reactive protein (WMD -0.01 mg/L, 95% CI -0.06 to 0.03) or other inflammation markers. Flow-mediated dilation rose 0.79% (95% CI 0.35 to 1.23) across 8 trials [13]. Tree nuts had no significant effect on C-reactive protein in 61 trials [4].
Population
32 randomized trials in adults, 652 participants for vessel function [13]. 2,582 adults in 61 trials [4].
Sources
[13], [4]
Caveats
The authors note the C-reactive protein result may have been influenced by two studies [13].
Antibiotics (common outpatient)Moderate

Common cold and green nasal discharge

Not shown to help the common cold or persisting colored nasal discharge against placebo, while adults given antibiotics for a cold had more side effects.

Compared with placebo [6]

The numbers and sources
Effect
Lack of cure or persistent cold symptoms did not differ (RR 0.83, 95% CI 0.60 to 1.14). Persisting purulent rhinitis did not differ (RR 0.73, 95% CI 0.47 to 1.13). Adverse effects were more frequent on antibiotics for colds (RR 1.8), in adults (RR 2.62) but not children (RR 0.91), and for purulent rhinitis (RR 1.46) [6].
Population
Up to 1,147 people in 6 common cold trials and 723 in 4 purulent rhinitis trials [6].
Sources
[6]
Caveats
Trials used various antibiotics and outcome measures, and some participants may have had chest or sinus infections [6].
Cold water immersionModerate

Inflammation

Not shown to lower inflammation. Inflammatory markers rose in the hour after immersion, and post-exercise reviews found no change in C-reactive protein or interleukin-6.

The numbers and sources
Effect
In healthy adults, inflammation markers were higher immediately after immersion (SMD 1.03, 95% CI 0.37 to 1.68) and 1 hour after (SMD 1.26, 95% CI 0.59 to 1.94) [14]. Across meta-analyses of post-exercise immersion, there were no significant effects on C-reactive protein or interleukin-6 up to 72 hours [6]. Five weeks of immersion after interval training did not change inflammatory markers in muscle [12].
Population
Healthy adults in 11 randomized trials [14]. Healthy people in 15 meta-analyses of post-exercise immersion [6]. 16 men [12].
Sources
[14], [6], [12]
Caveats
The acute rise was measured within an hour and the reviewers call it an acute inflammatory response. Long-term markers in everyday users were not pooled [14].
GlutamineModerate

Immune function in athletes

Not shown to change lymphocyte or leukocyte counts in athletes.

The numbers and sources
Effect
Glutamine did not change lymphocyte or leukocyte counts. Neutrophils fell at doses above 200 mg per kg (WMD -605.77), and glutamine dipeptide raised post-exercise blood glucose by 0.51 mmol/L [1].
Population
Athletes [1].
Sources
[1]
Caveats
Cell counts are not infection rates. The neutrophil confidence interval as printed (-1200.0 to 52.1) crosses zero, which conflicts with its reported p-value of 0.03 [1].
HomeopathyModerateStudied in children

Colds and respiratory infections in children

Not shown to prevent or cure respiratory infections in children. Every trial at low risk of bias found no benefit over placebo.

Compared with identical placebo or self-selected conventional treatment [11]

The numbers and sources
Effect
Non-individualized products had little preventive effect (OR 1.14, 95% CI 0.83 to 1.57). Individualized prevention had little effect on antibiotic use (OR 0.79, 95% CI 0.35 to 1.76). Short-term cure (OR 1.31 favoring placebo, 95% CI 0.09 to 19.54) and long-term cure (OR 0.99) were too uncertain to judge. All studies at low risk of bias showed no benefit, while trials at uncertain or high risk of bias reported benefits [11].
Population
8 trials with 1,562 children aged 0 to 16 with upper respiratory tract infections [11].
Sources
[11]
Caveats
Evidence was low or very low in quality, with high attrition and selective reporting. Three trials were funded by homeopathy manufacturers [11].
Nordic dietModerate

Inflammation markers

Not shown to change CRP, IL-6 or TNF-alpha in adults in Nordic countries.

Compared with control groups the source does not describe [8], [9]

The numbers and sources
Effect
In 6 trials (613 adults), CRP did not change (-0.17 mg/L, 95% CI -0.69 to 0.35), and TNF-alpha and IL-6 did not change in 3 trials [8]. A 2025 meta-analysis also found no effect on CRP [9].
Population
Adults in Nordic countries [8]; adults in 15 RCTs [9].
Sources
[8], [9]
Caveats
Short trials. In SYSDIET, IL-1 receptor antagonist rose on the control diet and not on the Nordic diet [4], a finding not tested in the pooled analyses.
Vitamin A (retinol) levelModerate

Infection and inflammation

Shown to read 11% to 24% lower in people with signs of infection, which can count people as deficient whose stores are adequate.

The numbers and sources
Compared with
Plasma retinol in people in the same studies with no raised acute phase protein
Effect
In a meta-analysis of 15 studies of apparently healthy people, plasma retinol was 13% lower in the incubation phase of infection, 24% lower in early convalescence and 11% lower in late convalescence than in people with no raised acute phase protein. Retinol was 25% higher in people with a normal C-reactive protein than in people with a raised one [9]. In blood samples from a hospital laboratory, median plasma vitamin A fell by more than 40% in the most inflamed group, and the authors conclude that the result can be interpreted reliably only when C-reactive protein is below 10 mg/L [11]. WHO states that clinical and subclinical infection can lower serum retinol by as much as 25% [2]. A US report explains the fall by transient decreases in retinol-binding protein and transthyretin, which carry retinol in blood, even when liver retinol is adequate [5].
Population
15 studies of apparently healthy people [9]. 2,217 samples from 1,303 patients in one UK laboratory [11].
Sources
[9], [11], [2], [5]
Caveats
The meta-analysis authors ask that surveys measure C-reactive protein and alpha-1-acid glycoprotein alongside retinol so that results can be corrected [9]. A European guideline states that there is no universal adjustment for inflammation [12].
Vitamin CModerate

COVID-19

Not shown to lower death, ventilation or hospital stay in patients with COVID-19 in randomized trials. A review adding cohort studies found lower mortality.

The numbers and sources
Effect
Across 9 RCTs, vitamin C did not change death from any cause (RR 0.92, 95% CI 0.83 to 1.02), need for ventilation (RR 0.98, 95% CI 0.87 to 1.11) or length of hospital stay (MD -0.63 days, 95% CI -3.04 to 1.78) [3]. Across 13 studies, short-term mortality (RR 0.92, 95% CI 0.72 to 1.17) and in-hospital mortality (RR 1.05, 95% CI 0.95 to 1.16) did not change [4]. A review that also pooled cohort and retrospective studies found lower mortality (OR 0.64, 95% CI 0.51 to 0.80) [5].
Population
Patients with COVID-19, 9 RCTs [3]; 13 studies with 12,062 participants [4]; 22 studies with 3,429 hospitalized patients [5].
Sources
[3], [4], [5]
Caveats
Sources disagree. The review that found lower mortality pooled cohort and retrospective studies [5]. Certainty ranged from very low to moderate [4].
Vitamin DModerate

Respiratory infections

Not shown to clearly lower respiratory infections in people. An earlier review found one point fewer, which lost significance in a 2025 update.

The numbers and sources
Effect
Across 37 trials with primary outcome data (48,488 participants aged 0 to 95), 61.3% on vitamin D and 62.3% on placebo had one or more acute respiratory infections (OR 0.92, 95% CI 0.86 to 0.99, I² 35.6%) [12]. A 2025 update by the same group added 3 trials with 16,085 participants. Across 40 studies (61,589 participants), the effect was no longer statistically significant (OR 0.94, 95% CI 0.88 to 1.00, P 0.057, I² 26.4%), and age, baseline vitamin D status, dosing frequency and dose size did not modify it [13]. In 17 pediatric RCTs (18,372 children), infections did not fall significantly (RR 0.82, 95% CI 0.64 to 1.06), and daily doses of 1,000 IU or less had RR 0.42 (95% CI 0.27 to 0.66) [14]. A Cochrane review of children up to 5 found RR 0.95 (95% CI 0.91 to 1.00) for the share making infection-related healthcare visits, and no change in visits per child [15]. Across 6 RCTs (15,677 participants), tuberculosis infection (OR 0.95, 95% CI 0.79 to 1.14) and active tuberculosis (OR 0.77, 95% CI 0.56 to 1.05) did not differ from placebo [16]. In 1,747 adults newly diagnosed with COVID-19, a 4-week course did not change healthcare visits or death (28% against 29%, OR 0.97, 95% CI 0.75 to 1.24) [17].
Population
46 double-blind RCTs, 75,541 participants [12]; 40 placebo-controlled studies, 61,589 participants [13]; children in 17 RCTs [14] and 107 RCTs from pregnancy to age 5 (31,521 participants) [15]; people without active tuberculosis [16]; adults in the US and Mongolia [17].
Sources
[12], [13], [14], [15], [16], [17]
Caveats
The 2021 and 2025 analyses come from the same group and share most trials. The 2021 absolute difference was one percentage point [12]. The 2025 funnel plot was asymmetric (Egger P 0.002) [13]. The Cochrane result rests on low-certainty evidence [15].
ZincModerate

Catching a cold, zinc taken daily

Not shown to prevent colds in children and adults taking zinc daily.

The numbers and sources
Effect
Zinc made little or no difference to the risk of developing a cold (RR 0.93, 95% CI 0.85 to 1.01, 9 studies, 1,449 participants, low certainty). When colds occurred, their length did not differ (MD -0.63 days, 95% CI -1.29 to 0.04, moderate certainty) [1].
Population
15 prevention trials in children and adults [1].
Sources
[1]
Caveats
Trials mixed common cold and upper respiratory infection definitions.
16 more
Absolute neutrophil count (ANC)Limited

Infections in people with the Duffy-null genotype

Limited evidence: one study of two population samples found the Duffy-null genotype lowered neutrophil counts and was not linked to more infections.

The numbers and sources
Effect
In UK Biobank participants of Black African or Caribbean ethnicity, the mean count was 2,820 cells/µL with the Duffy-null genotype and 4,430 without it. About 21% of people with the genotype had a count below 2,000, against 1.1% of the others. The genotype was not associated with a higher risk of viral or bacterial infection in UK Biobank or in a Danish sample. Among people with the genotype, 17.4% of those with a count of 1,500 to 2,000 had an infection against 18.8% of those with a count of 2,000 to 7,500 (RR 0.94, 95% CI 0.81 to 1.10) [15].
Population
UK Biobank, including 7,450 people with the Duffy-null genotype, and the Danish iPSYCH case-cohort sample [15].
Sources
[15]
Caveats
Observational. Only 323 people with the genotype had a count of 1,000 to 1,500, and the authors state that the small numbers below 1,500 preclude firm conclusions about infection risk at those counts [15].
CBDLimited

Inflammation markers

Limited evidence: CBD and THC trials found no significant change in IL-6, IL-8, IL-10 or TNF-alpha.

The numbers and sources
Effect
In a meta-analysis of randomized trials of CBD and THC, pooled changes in IL-6 (SMD -0.17, 95% CI -0.56 to 0.23), IL-8 (-0.30, 95% CI -0.62 to 0.01), IL-10 (-0.10) and TNF-α (-0.09, 95% CI -0.45 to 0.27) did not differ significantly from control [8].
Population
13 trials, with 2 to 4 trials and about 78 to 129 people per arm in each pooled marker [8].
Sources
[8]
Caveats
The review pooled CBD and THC interventions. GRADE certainty ranged from very low (IL-6) to moderate (IL-8, TNF-α) [8]. Blood markers, not symptoms.
Copper levelLimited

Inflammation and infection

Limited evidence: serum copper rises with inflammation because ceruloplasmin is an acute phase protein, so a result inside the range can hide low body copper.

The numbers and sources
Compared with
Copper status in the same person without inflammation
Effect
A US report states that ceruloplasmin is an acute phase protein that increases markedly with liver disease, malignancy, inflammatory diseases, myocardial infarction and a variety of infectious diseases, and that with any of these conditions copper deficiency might not be diagnosed from serum copper or ceruloplasmin [3]. A European guideline states that, unlike most micronutrients, plasma copper increases with an inflammatory response, and sets a higher cut point for likely deficiency when C-reactive protein is above 20 mg/L [4]. In 2,217 blood samples from a hospital laboratory, plasma copper and vitamin E were the only measured micronutrients that did not fall as C-reactive protein rose [12].
Population
2,217 samples from 1,303 patients screened for micronutrients in one UK laboratory between 2001 and 2011 [12].
Sources
[3], [4], [12]
Caveats
The laboratory study's abstract does not report by how much copper rose. The two official sources state the direction of the effect and do not give a correction factor [3][4].
CuppingLimited

Inflammation, wet cupping

One trial of 126 people with metabolic syndrome found wet cupping did not change high-sensitivity C-reactive protein over 12 weeks.

Compared with dietary advice alone [13]

The numbers and sources
Effect
Serum hs-CRP did not change between groups (p > 0.05) or over time when wet cupping was added to dietary advice, measured at 6 and 12 weeks [13].
Population
126 patients aged 18 to 65 with metabolic syndrome [13].
Sources
[13]
Caveats
One trial without a sham.
Deer antler velvetLimited

Arthritis

One trial found elk velvet antler did not differ from placebo over 6 months in 168 adults with rheumatoid arthritis.

The numbers and sources
Effect
In 168 people with rheumatoid arthritis on standard treatment, 6 months of elk velvet antler did not differ from placebo on joint pain and swelling, disease activity, function, quality of life or CRP [4]. A systematic review identified 2 rheumatoid arthritis trials and 1 osteoarthritis trial, and its authors wrote that the osteoarthritis findings may have some promise [5].
Population
Adults with stage 2 to 3 rheumatoid arthritis [4]; RCTs [5].
Sources
[4], [5]
Caveats
One osteoarthritis trial.
ElderberryLimited

Preventing colds

Limited evidence: cold episodes did not differ significantly from placebo in air travelers, and a review found elderberry may not lower cold risk.

The numbers and sources
Effect
In 312 air travelers, cold episodes were 12 on elderberry and 17 on placebo, a difference that was not significant (p 0.4) [3]. The systematic review found elderberry may not reduce the risk of developing a cold [2].
Population
Air travelers [3], 5 RCTs [2].
Sources
[3], [2]
Caveats
Few prevention trials.
Green-lipped musselLimited

Inflammation markers (CRP)

Small trials found no change in CRP in knee osteoarthritis, and lower airway inflammation markers in asthma.

The numbers and sources
Effect
In 100 adults with knee osteoarthritis, CRP and ESR did not differ between powder and placebo after 12 weeks [3]. In the asthma crossover trial, airway markers including cysteinyl leukotrienes and 8-isoprostane in exhaled breath fell on PCSO-524 against placebo [6].
Population
100 adults with knee osteoarthritis [3] and 20 adults with asthma [6].
Sources
[3], [6]
Caveats
Two small trials measuring different markers. No meta-analysis of inflammation markers was located.
High-sensitivity C-reactive protein (hs-CRP)LimitedStudied in women

Detecting endometriosis

Limited evidence: in a Cochrane review, three studies of hs-CRP gave sensitivities of 41% to 83% and specificities of 47% to 87% against surgery.

The numbers and sources
Compared with
Endometriosis seen at surgery (laparoscopy)
Effect
Three studies with six data sets tested hs-CRP against surgical findings at cut-offs that varied, with sensitivities of 0.41 to 0.83 and specificities of 0.47 to 0.87. The highest figures, 0.83 and 0.87, came from one study of 95 women sampled in the follicular phase. Five of the data sets drew on overlapping groups of women, so the results were not pooled. The review set a sensitivity of 0.94 and a specificity of 0.79 as the accuracy a blood test would need to replace diagnostic surgery. Across 122 blood markers in 141 studies, none consistently met its criteria for replacing surgery or for triage, and the reviewers concluded that none was accurate enough for clinical use outside research [23].
Population
506 participants in three studies of hs-CRP, within 141 studies of 15,141 women [23].
Sources
[23]
Caveats
The review rated every included study as of poor methodological quality, and the women were already having surgery, with endometriosis in 10% to 85% of them, so the figures may not hold for women tested for symptoms alone. Its searches ran to 2015 [23].
High-sensitivity C-reactive protein (hs-CRP)LimitedStudied in women

Uterine fibroids

One US survey found no clear link between uterine fibroids and CRP above 3 mg/L.

The numbers and sources
Effect
In the same national survey that linked endometriosis to higher CRP, the authors report that no substantive patterns emerged between uterine fibroids and CRP above 3.0 mg/L or leukocyte telomere length [24].
Population
2,342 women in the National Health and Nutrition Examination Survey, 1999 to 2002 [24].
Sources
[24]
Caveats
One cross-sectional analysis with self-reported diagnoses, and the abstract gives no odds ratio for fibroids [24]. No meta-analysis of inflammation markers in women with fibroids was located.
Lymphatic drainage massageLimited

Lymphedema after breast cancer, compared with placebo drainage

One trial of 194 people found adding manual lymphatic drainage to decongestive therapy reduced arm swelling no more than placebo drainage.

Compared with placebo manual lymphatic drainage, all groups also receiving compression, skin care, exercise and education [2]

The numbers and sources
Effect
After 3 weeks of intensive treatment, excess arm volume fell by 5.3 percentage points with fluoroscopy-guided drainage, 5.2 with traditional drainage and 5.4 with placebo drainage. Fluoroscopy-guided drainage differed from placebo by -0.2 percentage points (95% CI -2.1 to 1.8). Fluid built up at the shoulder and trunk in all groups, and secondary outcomes showed no clinically important differences [2].
Population
194 people with chronic lymphedema on one side after breast cancer, at 5 hospitals in Belgium, followed through 6 months of maintenance treatment and 6 months of follow-up [2].
Sources
[2]
Caveats
One trial in chronic lymphedema. Participants and assessors were blinded [2].
MultivitaminLimited

Immunity and COVID-19

One trial found no difference in COVID-19 cases or symptomatic illness against placebo in older US adults, in a secondary analysis.

The numbers and sources
Effect
In the final COSMOS year (2020), COVID-19 incidence did not differ between multivitamin and placebo (HR 0.93, 95% CI 0.81 to 1.07). Among 338 cases with symptom data, symptomatic illness did not differ significantly (OR 0.70, 95% CI 0.44 to 1.11). In the per-protocol analysis, the odds ratio was 0.60 (95% CI 0.37 to 0.99) [11].
Population
18,205 adherent COSMOS participants as of January 2020 [11].
Sources
[11]
Caveats
A secondary analysis with self-reported infections. The per-protocol result is secondary.
Why limited
This rests on one trial, of 21,442 people. Strength counts independent studies that agree, so one trial is limited at any size.
Ozone therapyLimited

COVID-19

Limited evidence: pooled trials found ozone did not shorten hospital stays or reduce intensive care admissions or deaths from COVID-19, though fewer PCR tests stayed positive.

Compared with placebo, sham, saline or standard care [12]

The numbers and sources
Effect
In meta-analyses of randomized trials, ozone therapy lowered PCR positivity at follow-up (RR 0.07, 95% CI 0.01 to 0.34), which the reviewers judged a clinically unimportant surrogate. Hospital stay, intensive care admission and mortality did not differ [12].
Population
Meta-analyses of trials in patients with COVID-19, within an umbrella review of 7 meta-analyses [12].
Sources
[12]
Caveats
Certainty was low or very low for every outcome in the umbrella review, and safety outcomes were reported inconsistently [12].
ReishiLimited

Alongside cancer treatment

Limited evidence: tumor response with reishi alongside chemotherapy or radiotherapy was not clearly better, and no trial recorded long-term survival.

The numbers and sources
Effect
In 5 RCTs, patients given reishi alongside chemotherapy or radiotherapy were more likely to respond than those on chemotherapy or radiotherapy alone (RR 1.50, 95% CI 0.90 to 2.51). Reishi alone did not show the same response. CD3, CD4 and CD8 percentages rose by 3.91%, 3.05% and 2.02%, and four studies reported better quality of life. No trial recorded long-term survival [1].
Population
Patients with cancer [1].
Sources
[1]
Caveats
Trial quality was generally unsatisfying and reporting inadequate. The reported interval for tumor response crosses 1 [1].
Selenium levelLimited

Inflammation and infection

Limited evidence: plasma selenium falls as inflammation rises, by about half when C-reactive protein is above 80 mg/L, without a matching loss of body selenium.

The numbers and sources
Compared with
Plasma selenium in patients of the same laboratory with the lowest C-reactive protein
Effect
In blood samples from a hospital laboratory, plasma selenium fell with increasing severity of the acute inflammatory response, beginning at C-reactive protein of only 5 to 10 mg/L, and the median fell by more than 40% in the most inflamed group. The authors conclude that a plasma selenium can be interpreted reliably only when C-reactive protein is below 10 mg/L [13]. A European guideline gives expected falls of 15% to 25% at C-reactive protein of 10 to 40 mg/L, about 35% at 41 to 80 mg/L and about 50% above 80 mg/L. It attributes the fall to selenium moving out of the circulation, because plasma selenium returns to normal in many cases without supplementation [2].
Population
2,217 samples from 1,303 patients screened for micronutrients in one UK laboratory between 2001 and 2011 [13].
Sources
[13], [2]
Caveats
The change varied markedly from patient to patient, so no single correction fits an individual result [13].
Thyroid peroxidase antibodies (TPO antibodies)Limited

Differences between assays

One study found upper reference limits for TPO antibodies from 6.4 to 28.7 IU/mL across five automated assays run on the same men.

The numbers and sources
Compared with
Four other automated immunoassays run on the same blood samples
Effect
Using the reference group that the National Academy of Clinical Biochemistry guideline specifies, the upper reference limits in men and women were 28.7 and 29.0 IU/mL on Immulite 2000 XPi, 24.6 and 25.4 on Maglumi 2000 Plus, 6.4 and 6.9 on Kryptor Compact Plus, 8.3 and 10.0 on Phadia 250 and 14.2 and 17.9 on Liaison XL. These were lower than the limits stated by the manufacturers, except for Liaison XL in women, by 11.3% to 66.8%. The five methods did not compare well with each other, although all were standardized against the same international reference preparation [2].
Population
120 men and 120 women under 30 with TSH of 0.5 to 2.0 mIU/L, a normal thyroid ultrasound and no thyroid or other autoimmune disease [2].
Sources
[2]
Caveats
One Italian study of 240 people. The authors attribute the differences to the TPO antigen preparations the assays use [2]. A result is read against the cutoff of the assay that produced it.
Thyroid peroxidase antibodies (TPO antibodies)Limited

Thyroid disease when only thyroglobulin antibodies are positive

One large US survey found thyroglobulin antibodies without TPO antibodies were not significantly associated with thyroid disease, while TPO antibodies were.

The numbers and sources
Effect
In the NHANES III survey, thyroglobulin antibodies were positive in 10.4% and TPO antibodies in 11.3% of people who reported no thyroid disease. TPO antibodies were significantly associated with hypothyroidism or hyperthyroidism and thyroglobulin antibodies were not. Thyroglobulin antibodies alone, in the absence of TPO antibodies, were not significantly associated with thyroid disease [16].
Population
17,353 people aged 12 and older representing the US population from 1988 to 1994, of whom 16,533 reported no thyroid disease [16].
Sources
[16]
Caveats
One cross-sectional survey, using the assays of its time. Positive antibodies were more common in women, rose with age and were less common in Black participants (4.5%) than in white participants (12.3%) [16].

Checked, no increase found2

ARBs (angiotensin receptor blockers)Moderate

Angioedema

Not shown to raise angioedema above placebo in trials, 0.11% against 0.07%. Cases have been reported, some in people who had it on an ACE inhibitor.

The numbers and sources
How common
Uncommon: between 1 in 1,000 and 1 in 100 people.
Effect
Across 19 trials, the weighted incidence of angioedema was 0.11% on ARBs (95% CI 0.09 to 0.13) and 0.07% on placebo, a difference that was not significant, against 0.30% on ACE inhibitors [25]. The losartan label lists angioedema, including swelling of the larynx and glottis with airway obstruction, among reactions reported after marketing, and notes that some of these people had angioedema before on other drugs including ACE inhibitors [10].
Population
35,479 people on ARBs in 19 trials with 122,293 person-years of follow-up [25].
Sources
[25], [10]
Caveats
Post-marketing reports give no rate [10].
ElderberryLimited

Immune overstimulation

Limited evidence: a review found no studies linking elderberry to clinical inflammatory outcomes, and rated the evidence on harms uncertain.

The numbers and sources
How common
How often is not known: the sources give no rate.
Effect
A systematic review prompted by COVID-19 concern about cytokine storm found no studies linking elderberry to clinical inflammatory outcomes. Three studies measured cytokines ex vivo after elderberry, and the effect on markers may decline with ongoing use. Against oseltamivir, an elderberry-containing product may be associated with fewer flu complications and adverse events [2].
Population
5 RCTs and 3 ex vivo cytokine studies [2].
Sources
[2]
Caveats
The review rated the evidence on harms uncertain [2].

Not studied in people8

AntihistaminesUntested

Anaphylaxis

Not studied in people. A Cochrane review found no randomized trial of H1 antihistamines for anaphylaxis, a use that rests on how they block histamine.

No study in people. This rests on how it acts in the body.

The numbers and sources
Effect
A Cochrane review searched for randomized and quasi-randomized trials of H1 antihistamines for anaphylaxis with or without shock and found none that met its criteria. The authors stated they could not draw conclusions for practice [11].
Population
No trials located [11].
Sources
[11]
Caveats
Antihistamines block histamine, one of several mediators of anaphylaxis. Without trials, any effect on outcomes such as survival is unknown [11].
BPC-157Grey marketUntested

Inflammation

Not studied in people. Animal and lab studies report lower inflammatory cytokines.

No study in people. This rests on animal studies, lab studies.

The numbers and sources
Effect
A 2025 systematic review of mostly preclinical studies found that BPC-157 reduced inflammatory cytokines and acted on pathways involved in cell growth and new blood vessel formation [16]. A 2026 review reports reduced inflammatory cytokine activity and improved small blood vessel integrity across preclinical models [24].
Population
Animal models and cell studies [16][24].
Sources
[16], [24]
Caveats
The review abstracts give no effect sizes for inflammation. The human research located is limited to small pilot studies of musculoskeletal pain, interstitial cystitis and intravenous use [24].
Chiropractic and spinal manipulationUntested

Preventing infections or boosting immunity

Not studied in people. A systematic review found no clinical study of spinal manipulation preventing infectious disease or improving its outcomes.

No study in people. This rests on how it acts in the body.

The numbers and sources
Effect
Reports posted by a chiropractic trade association in March 2020 claimed adjustments affect the immune system. More than 150 researchers from 11 countries reviewed the reports and found no valid clinical evidence for the claim [14]. A systematic review searching to April 2020 found no clinical studies of manipulation for preventing infectious disease or improving outcomes in people with one [13].
Population
No study population. Two association reports reviewed [14], and 13 studies screened [13].
Sources
[14], [13]
Caveats
The claim rests on a proposed link between the spine, the nervous system and immunity, not on any trial in people [14].
GroundingUntested

Antioxidant effect of the Earth's electrons

Not studied in people. Grounding researchers propose that electrons from the ground act as antioxidants around injured tissue, a mechanism no human study has measured.

No study in people. This rests on how it acts in the body.

The numbers and sources
Effect
The main hypothesis of one review is that free electrons from the Earth's surface spread into the body and have antioxidant effects, creating an antioxidant microenvironment around injury that limits damage from reactive oxygen species [6]. An earlier review proposes that losing contact with the Earth's electrons contributes to physiological dysfunction [10].
Population
No people. Hypotheses in reviews [6][10].
Sources
[6], [10]
Caveats
The authors of both reviews have ties to the company sponsoring grounding research [6].
IV vitamin therapyUntested

Immunity in healthy people

Not studied in people as a drip. In experimental sepsis, high-dose IV vitamin C acted on immune signaling, which no trial in healthy people has tested.

No study in people. This rests on how it acts in the body.

The numbers and sources
Effect
A systematic review describes proposed antioxidant and immune-modulating actions of high-dose IV vitamin C, including down-regulation of inflammatory STAT1 and PD-L1 pathways in experimental sepsis. Its clinical evidence came from sepsis and cancer care [9]. No trial located tested a drip for preventing infections in healthy people.
Population
No study population. A review of mechanistic and clinical studies [9].
Sources
[9]
Caveats
Mechanisms seen in experimental sepsis do not show what a drip does for a healthy immune system.
Mouth tapingUntested

Oral health, energy and immunity

Not studied in people. Social media videos credit mouth taping with better sleep and oral health, and with more energy and immunity.

No study in people. This rests on what users report.

The numbers and sources
Effect
A review of TikTok videos about mouth taping found the most common proposed advantages were improved sleep and oral health. The same review reports claims of increased energy, immunity and dental health, and found few of them had been evaluated [5].
Population
The first 50 TikTok videos meeting the review's criteria [5].
Sources
[5]
Caveats
Claims in videos are not measurements. No study located tested mouth taping for oral health, dental health, energy or immunity.
Oregano oilUntested

Infections and candida treated by mouth

Not studied in people. Mice with a body-wide candida infection survived more often when fed oregano oil, and in the lab it stopped candida growth.

No study in people. This rests on animal studies, lab studies.

The numbers and sources
Effect
In culture, oregano oil at 0.25 mg/mL completely inhibited Candida albicans growth. In mice with systemic candidiasis, oregano oil given by mouth daily for 30 days was followed by 80% survival with no candida in the kidneys, while mice fed olive oil alone died within 10 days [7]. A 2025 systematic review of lab studies of four oregano species found antifungal activity against Candida, Aspergillus and Penicillium species, linked to carvacrol, thymol and related compounds [8].
Population
Mice, 6 per group, and Candida albicans cultures [7]. Lab studies of fungal strains [8].
Sources
[7], [8]
Caveats
One small mouse study and lab work. Results in culture do not show what happens in the body. No trial of oral oregano oil for candida or other infections in people was located.
TB-500Grey marketUntested

Inflammation

Not studied in people. In lab brain immune cells and in a mouse model of Alzheimer's disease, TB-500 lowered signs of inflammation.

No study in people. This rests on lab studies, animal studies.

The numbers and sources
Effect
In a microglia cell line, the brain's immune cells, TB-500 suppressed the nitric oxide production and pro-inflammatory cytokine expression set off by bacterial lipopolysaccharide. In 5xFAD mice, a model of Alzheimer's disease, it reduced glial activation in the brain [18].
Population
Cells in the lab, and 5xFAD mice [18].
Sources
[18]
Caveats
The study measured brain inflammation in a model of Alzheimer's disease, not the injury or joint inflammation TB-500 is taken for. It tested another thymosin beta-4 fragment, Ac-SDKP, alongside TB-500 [18].