Whole body

Shown to raise fatigue and constipation against placebo in chemotherapy trials, with fatigue in 13% against 2% and constipation in 9% against under 1%.

Shown to raise side effects during short courses, with at least one adverse event in 49.2% on 15 days of prednisone against 23.9% on placebo.

Shown to raise nausea, dry mouth, sweating, constipation and dizziness against placebo, with nausea in 23% to 30% against 8% to 12% on placebo.

Shown to raise nausea, diarrhea, tremor and sweating against placebo, with nausea in 26% on sertraline against 12% on placebo.

Shown to raise sleepiness and fatigue, about twice as often as placebo in add-on epilepsy trials, with fatigue in 15% against 11% in migraine trials.

Shown to raise headache, sinus symptoms, back pain and muscle pain slightly over placebo. Headache affected 50% on the drug and 47% on placebo.

Shown to raise dry mouth, blurred vision, swelling and fatigue against placebo, with dry mouth in 34% of outpatients against 20% on placebo.

After the emergency dose, 18% reported headache and 12% nausea, as often as on levonorgestrel. In fibroid trials, headache and breast tenderness did not differ from placebo.

Small trials found fishy body odor in 36% of children on choline against 8% on placebo, in children with fetal alcohol spectrum disorder.

One trial found more bleeding after frail older adults were switched from a vitamin K antagonist to a DOAC, 101 of 662 against 62 of 661.

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

One trial, stopped early, found more deaths in extremely preterm infants given myo-inositol, 18% against 11% on placebo.

The clomiphene label lists ovarian enlargement in 13.6% and hot flushes in 10.4% of women in its trials. Against clomiphene, letrozole had fewer hot flushes but more fatigue and dizziness.

Linked to weight gain in 70% of long-term users in one survey, with cataracts reported by 15% and fractures by 12%.

One trial found 80% of people who quit with e-cigarettes were still vaping a year later, against 9% still using nicotine replacement.

Shown to raise major bleeding by about 43% against no aspirin, to 4.1% against 3.2% over 7.4 years in one trial and 3.8% against 2.8% in another.

Shown to cause major bleeding in about 5% of people with atrial fibrillation over the trials, a rate similar to or below warfarin's 5.9%.

FDA states that warfarin can cause major or fatal bleeding. Major bleeding occurred in about 6% of people with atrial fibrillation over the pivotal trials.

Shown in about 9% of patients over a course of acupuncture, most often bleeding, bruising or pain at the needle site.

Shown to raise major bleeding when clopidogrel was added to aspirin, 3.7% against 2.7% in acute coronary syndromes, and about 60% more with courses beyond a year.

Shown to raise major bleeding when aspirin was combined with low-dose rivaroxaban, 3.1% against 1.9% on aspirin alone. Labels warn of the same with warfarin.

FDA states older people with dementia-related psychosis on antipsychotics had about 1.6 to 1.7 times the death rate of placebo, 4.5% against 2.6% over 10 weeks.

Linked to about 2.4 times the major bleeding when an NSAID was taken alongside an anticoagulant, and to 3.1% against 1.9% when rivaroxaban was added to aspirin.

Shown to slightly raise colds, diarrhea, joint pain and tiredness in label trials, with upper respiratory infection in 4.3% against 2.5% on placebo.

Linked to adrenal insufficiency in about 6.8% of people with asthma using only inhaled steroids, more at high doses and after more than a year.

Shown to raise fatigue, 6 to 9% against 4% on placebo.

One trial found more deaths on aspirin than placebo in healthy adults aged 70 or over, driven by cancer deaths in 3.1% against 2.3%.

Limited evidence: in two Norwegian trials, folic acid with B12 raised cancer from 8.4% to 10.0% and deaths from 13.8% to 16.1% in people with heart disease.

One trial found 7 deaths on mirtazapine against 1 on placebo by week 16 in people with dementia, a difference of marginal statistical significance.

Shown in about 1 per 10,000 patients. Two collapsed lungs were recorded among 229,230 patients.

Prospective surveys recorded major adverse events after about 1 in 1,000 treatments or fewer. Case reports describe collapsed lungs after needling over the chest wall.

Linked to higher risk of cancers of the mouth, throat, esophagus, bowel, larynx, liver and breast, rising with dose. Breast cancer rose 5% at up to one drink daily.

Shown to raise swelling, joint pain and carpal tunnel symptoms, with swelling about 6 times as likely in adults with deficiency.

Linked to about 15% to 28% more deaths in nappers in four meta-analyses of cohorts, mainly with naps of an hour or more, rising in a straight line.

Shown to raise side effects such as headache and flushing, about 1.7 times placebo for sildenafil, with tadalafil 10 mg lowest among starting doses.

Linked to 23% higher death at the highest processed meat intake than the lowest, in pooled cohorts. A later review rated the link very small and low certainty.

Linked to about two to three times the death rate of never smokers in large cohorts, and to more than a decade of lost life.

Shown to slow growth in children, with about 2 cm less height and 2.7 kg less weight over 3 years of daily methylphenidate.

FDA states pure or highly concentrated caffeine products have been linked to at least two US deaths, and seizures can occur with rapid intake of about 1,200 mg.

Linked to 20% higher cancer death at the highest egg intake than the lowest, in pooled cohorts rated low to moderate in certainty.

Shown to about double in-hospital deaths at high doses in two intensive care trials. A later review of 23 trials found no mortality change.

Shown to slow weight and head growth slightly in healthy breastfed infants, while lowering anemia at 6 months.

Linked to 20% to 31% higher all-cause mortality in cohort studies, higher when carbohydrate was replaced with animal foods and lower when replaced with plant foods.

Linked to a 530% rise in reported child ingestions in the US from 2012 to 2021, mostly in children 5 and under, with two deaths.

Estrogen plus progestin added 12 major events per 10,000 women a year at 50 to 59 and 38 at 70 to 79. Estrogen alone had 19 fewer and 51 more.

Shown to suppress the adrenal glands, with adrenal insufficiency in 1.4% of asthma patients after under 28 days of use and 27.4% after more than a year.

Shown to raise death about 43% in kidney transplant recipients against other regimens, at transplant doses with other immunosuppressants.

Shown to raise death from any cause 8% to 17% with beta-carotene in smokers and asbestos-exposed workers. Deaths did not differ in mostly nonsmoking physicians.

Shown to raise death or organ dysfunction with intravenous vitamin C in sepsis, and deaths in severe burns. This concerns critically ill patients, not oral supplements.

Shown in a review of trials, surveillance and reports to cause mostly mild, reversible side effects: nausea, headache, stomach upset, menstrual changes, acne, itching and rash.

Linked to about twice the odds of a bleeding admission within 15 days of an antibiotic in older warfarin users, and over four times with azole antifungals.

Linked to about twice the odds of stomach and gut bleeding when an NSAID was taken with warfarin, compared with warfarin alone.

Linked to about 35% more major bleeding when an SSRI was taken alongside an oral anticoagulant, in a meta-analysis of eight observational studies.

The label warns that serotonin syndrome may occur when amphetamines are taken with other drugs acting on serotonin. It gives no rate.

Small weight gains were seen in label trials, about 1 to 2 pounds on terazosin and 0.6 kg on doxazosin against small losses on placebo.

The label warns of serotonin syndrome with other serotonergic drugs and lists MAOIs as a contraindication, after hyperpyretic crises, convulsions and deaths with the combination.

The label reports nausea, headache and malaise after abrupt stopping, and irritability, restlessness and sleep disturbance during gradual reduction. It gives no rate.

The label states that SSRI and SNRI antidepressants affect platelet activation and may increase the risk of bleeding when taken with clopidogrel. It gives no rate.

The label warns of neuroleptic malignant syndrome, a potentially fatal reaction reported with antipsychotic drugs including quetiapine. It gives no rate.

The label reports serotonin syndrome with buspirone, mostly when taken with other serotonergic drugs such as SSRIs, SNRIs or triptans. It gives no rate.

Linked in two Swedish cohorts to 15% higher death per daily glass of milk in women and 3% in men, and to more hip fractures in women.

The labels list SSRI and SNRI antidepressants among drugs that increase the risk of bleeding when taken with apixaban or rivaroxaban. They give no rate.

Linked in a systematic review to smoking, alcohol and other substance use, short sleep, low academic performance, ADHD symptoms, suicide risk and dental erosion in people under 21.

US emergency department visits involving energy drinks doubled from 10,068 in 2007 to 20,783 in 2011, a national surveillance report found.

Linked to more injury than alcohol alone in a systematic review, and to less sedation and more stimulation in pooled studies.

The label reports cyclosporine raises ezetimibe and cyclosporine levels, cholestyramine lowers ezetimibe exposure, and combining it with fibrates other than fenofibrate is not advised.

Case reports: a US public health investigation of two severe illnesses, one fatal, in people drinking kombucha daily, and a report of severe liver injury.

Case reports: 7 clinic patients had nausea, vomiting and chills minutes after compounded IV glutathione. FDA linked this to possible high endotoxin levels.

The label lists headache, irritability, depression, fatigue, swelling and injection site reactions, breast growth in men and ovarian hyperstimulation in women. It gives no rates.

FDA states it found 99 adverse event cases consistent with belladonna toxicity, including infant deaths and seizures, possibly related to homeopathic teething products.

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

The label states use with strong CYP3A4 inhibitors such as ritonavir, ketoconazole or clarithromycin can raise steroid levels and heart side effects. It gives no rate.

Limited evidence: after the 1997 FDA unit-dose packaging rule, iron poisoning deaths in children under 6 fell from 29 in 10 years to 1 in 5 years.

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

Linked to fiber intake below guideline goals in US adults on low-carbohydrate diets, a shortfall the general population also showed.

The label warns that mirtazapine can precipitate serotonin syndrome, with higher risk alongside other serotonergic drugs. It gives no rate.

The label lists MAOIs as a contraindication because the combination increases the risk of serotonin syndrome. It gives no rate.

The label reports dizziness, abnormal dreams, electric shock sensations, agitation, anxiety, nausea and sweating after stopping mirtazapine, particularly abruptly. It gives no rate.

One trial found more appetite, mild leg swelling and muscle pain, with weight up 2.7 kg against 0.8 kg on placebo, in older adults.

The label reports serotonin syndrome with ondansetron, mostly with other serotonergic drugs such as SSRIs and SNRIs, some cases fatal. It gives no rate.

Linked to more deaths within a year than naproxen in adults with osteoarthritis starting tramadol, 23.5 against 13.8 per 1,000 person-years.

One trial found one case of serotonin syndrome-like symptoms with SAMe added to an antidepressant. Other side effects were no higher than on placebo.

One large cohort linked being a definite evening type to 10% higher death from any cause than being a definite morning type.

The label warns that SNRIs can cause serotonin syndrome, more often with other serotonergic drugs. It gives no rate.

The label lists MAOIs as a contraindication, within 14 days before or 5 days after duloxetine, because of an increased risk of serotonin syndrome.

The label warns that SSRIs can cause serotonin syndrome, alone or more often with other serotonergic drugs. It gives no rate.

The label lists MAOIs as a contraindication, within 14 days in either direction, because the combination raises the risk of serotonin syndrome.

The label warns of an increased risk of neoplasms and lists active cancer as a contraindication. IGF-1 rose 81% in one trial.

Animal and lab studies only: EFSA concluded the extract can induce DNA damage, and that its safety was not established under any condition of use.

The label reports decreased sweating and raised body temperature, mostly in children and some after hot weather, occasionally needing hospital care. It gives no rate.

The label warns that trazodone can cause serotonin syndrome, more often with other serotonergic drugs. It gives no rate.

The label lists MAOIs as a contraindication because the combination increases the risk of serotonin syndrome.

Case reports: three people developed withdrawal symptoms despite gradual tapering. Pooled reviews of antidepressants give no rate for trazodone alone.

One trial found more deaths on tretinoin 0.1% than vehicle in older veterans and stopped early. The authors judged a causal link unlikely.

Case reports: US poison centers recorded 8,269 liquid nicotine exposures in children under 6 from 2012 to 2017, mostly by swallowing.

Limited evidence: one review found low selenium intake in European vegans.

Linked to a third of emergency hospital admissions for adverse drug events among US adults aged 65 or over, in national surveillance data from 2007 to 2009.

The label lists aspirin, clopidogrel and other antiplatelet and anticoagulant drugs as raising bleeding risk with warfarin. Studies of anticoagulants with aspirin found about 1.7 times the bleeding.

The label warns that garlic and ginkgo may add to bleeding, and that coenzyme Q10, St. John's wort and ginseng may weaken warfarin. Few controlled studies exist.

FDA states long-term inorganic arsenic exposure is linked to skin, bladder and lung cancers, names children as most vulnerable, and sets an action level for infant rice cereal.

FDA states hyperbaric devices are not proven to cure cancer, Lyme disease, autism or Alzheimer's disease, and that explosions and fires have occurred in unreviewed chambers at unaccredited facilities.

Shown to cause adverse effects in 30% of treated participants against 10% of controls across 24 randomized trials, more often above 2.0 ATA or beyond 10 sessions.

Shown to injure the middle ear often: 43.2% of 236 hospital patients had barotrauma, most of it minor, and children in two trials had more ear problems than controls.

FDA states it has not cleared or approved any whole-body cryotherapy device for medical treatment, and names asphyxiation, frostbite, burns and eye injury as hazards.

Small studies found core temperature kept falling after leaving cold water in 10 of 11 swimmers, and only 5 of 10 swimmers lasted 90 minutes at 10°C.

Case series: 77 people died in 35 hyperbaric chamber fires worldwide from 1923 to 1996, every fatal fire in an oxygen-enriched atmosphere.

One small study found 45 minutes in a 65°C far-infrared sauna raised core temperature 1.4°C, doubled heart rate and shrank plasma volume 11.6% in healthy adults.

Case series: sauna deaths in Finland ran below 2 per 100,000 people a year, and half or more of those who died had been drinking alcohol.

In a survey of 482 regular sauna users, reported adverse reactions were mostly minor, chiefly dizziness, dehydration and headache. Two reports involved hospital care.

A scoping review found 16 documented adverse events after whole-body cryotherapy in 5 case reports and 2 trials, 4 of them graded serious.