What are the odds of being harmed by taking over-the-counter painkillers regularly?
Evidence quality 4.63/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 5/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 4/5
- D4 Uncertainty
- 4/5
- D5 Scope
- 5/5
- D6 Prose
- 5/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 5/5
Lifetime probability · lifetime, subgroup
1 in 10
10% lifetime chance
Most people underestimate this.
range 1 in 20 to 1 in 5.0
● your factors — click this risk ▾ to reveal
≈ As likely as
Perceived
OTC painkillers sit in an unusual perceptual slot: readily available on grocery shelves, universally familiar, and mentally filed alongside vitamins and antacids rather than alongside drugs. The usual public intuition is that the ceiling of harm from ibuprofen, naproxen, aspirin, or acetaminophen taken "a few times a week" is a mild stomach ache or a headache that doesn't go away. That intuition undercounts a very real chronic-use hazard: serious GI bleeding, acute kidney injury, and acetaminophen-driven acute liver failure each have measurable annual event rates in habitual users. No rigorous nationally-representative survey measures the perceived-harm distribution for chronic OTC analgesic use, so this is flagged as intuition, not a polled number.
Rough estimate: Most chronic users guess the lifetime serious-adverse-event risk well below 1 in 100
Source: editorial intuition, not polled
Actual
~1 in 100 per year serious GI event (chronic NSAID user)
chronic NSAID user (≥3×/week, >1 year), US adult, mid-life onward
Show derivation
Reference subgroup: a US adult who uses an OTC NSAID (ibuprofen, naproxen, or low-dose aspirin taken for pain rather than cardioprotection) or acetaminophen three or more times per week across roughly 20 years of adult life, without gastroprotection from a proton-pump inhibitor. The ~10% headline is a rounded mid-point for the cumulative probability of a clinically significant adverse event — upper GI bleed or ulcer requiring medical attention, acute kidney injury, an NSAID-attributable cardiovascular event, or acetaminophen-associated hepatotoxicity — across that chronic-use window. Bracketed 5-20% to reflect (a) the Lanas 2005 nationwide cohort incidence of ~122 major GI events per 100,000 persons per year averaged across the full population, which rises to roughly 1 per 100-1000 per year among habitual NSAID users and to 1 per ~110 per year in chronic users over 75, and (b) the Marcum/Hanlon 2010 figure of ~41,000 hospitalizations and ~3,300 deaths per year among older adult NSAID users in the US. Compounding an annual event hazard of ~0.5% per year (middle-aged chronic user weighted-average across GI, renal, CV, and hepatic endpoints) over 20 years: 1 - (1 - 0.005)^20 ≈ 0.095, rounded to ~10%. The hazard rises steeply with age: the same compounding over 20 years of chronic use starting at age 65 (annual hazard closer to 1%) produces ~18%. The hazard falls steeply for users under 50 without risk factors. Scope is subgroup_lifetime because this is a per-chronic-user cumulative probability, not a general-population lifetime risk, and not directly comparable to the population-scope figures on other Likelier pages.
Caveats: This entry is specifically the cumulative serious-adverse-event probability for …
This entry is specifically the cumulative serious-adverse-event probability for a chronic OTC analgesic user — operationally a US adult taking an NSAID or acetaminophen three or more times per week across roughly 20 years of adult life without gastroprotection. It is not the risk from short-course or occasional use, where the per-event hazard is low enough that the compounded probability rounds to near-baseline. Acute use for a headache, a hangover, a sprained ankle, a menstrual cramp, or a post-operative week is not what this number is measuring. The headline also mixes four distinct endpoints — upper GI bleed or ulcer, acute kidney injury, an NSAID-attributable cardiovascular event, and acetaminophen-associated hepatotoxicity — into a single composite "clinically significant harm" outcome; readers interested in a specific endpoint should read the regional_breakdown and personal_factor_multipliers rows. The per-agent risk profile differs: naproxen has the lowest cardiovascular signal in the Trelle meta-analysis; ibuprofen and diclofenac sit higher; ibuprofen and naproxen share a broadly similar GI profile at OTC doses; low-dose aspirin carries its own substantial contribution to the combined GI mortality total (~1/3 per Lanas 2005). Acetaminophen at or below the 4 g/day label ceiling is largely safe for the liver in the absence of alcohol, chronic liver disease, or concurrent use of a second acetaminophen-containing product — the unintentional overdose pattern in Lee's registry is dominated by precisely those three modifiers. Individual outcomes depend on age, sex, dose, frequency, duration, specific agent, coadministration (PPIs, anticoagulants, antiplatelets, SSRIs), pre-existing GI/renal/hepatic/cardiovascular disease, H. pylori status, alcohol intake, and a long list of pharmacogenomic modifiers. The 5-20% uncertainty band is wide on purpose: composite adverse-event probabilities for chronic OTC use are less tightly constrained than the aggregate annual mortality/hospitalization totals.
Regional breakdown
The headline figure averages across very different populations. Here’s how the probability varies by geography or context:
| Region / context | Lifetime probability | Notes |
|---|---|---|
| Occasional OTC user (<1×/week, any age, no risk factors) | 1 in 500 |
Near-baseline. Acute AEs at label dose and frequency are rare; most epidemiological signal comes from chronic users. |
| Chronic NSAID user (3+×/week, middle-aged, no PPI) | 1 in 10 |
Headline subgroup. ~0.5% annual serious-AE hazard (GI + renal + CV composite) compounded over ~20 years of chronic use. |
| Chronic NSAID user, age 65+ (no PPI) | 1 in 5.6 |
Peptic ulcer hospitalization rate 2-6 per 1,000/year per Marcum/Hanlon 2010; compounded annual hazard closer to 1%. |
| Chronic acetaminophen at ≤4 g/day label dose, no alcohol | 1 in 100 |
Hepatotoxicity at or below the label ceiling is rare in the absence of alcohol or pre-existing liver disease; ~30,000 acetaminophen-tox hospitalizations/year spread across ~50M adult users is ~0.06%/year. |
| Chronic acetaminophen user + ≥2 drinks/day | 1 in 20 |
Alcohol-induced CYP2E1 upregulation sharply raises NAPQI production at any given acetaminophen dose; unintentional hepatotoxicity concentrated in this group per Lee 2008. |
| Chronic NSAID + daily low-dose aspirin (combined GI risk) | 1 in 6.7 |
Upper-GI-bleed hazards multiply rather than add; up to one-third of NSAID/aspirin GI deaths attributable to low-dose aspirin per Lanas 2005. |
Risks at similar odds
Other risks with roughly the same likelihood — useful for calibration.
Inheriting AUD risk
If a parent had alcohol use disorder, what are the odds you'll develop alcohol use disorder yourself?
Compulsive sexual behavior
What are the odds of developing compulsive sexual behavior disorder?
Smokeless tobacco
What are the odds of dying from a smokeless-tobacco-related disease as a regular user?
Untreated back pain disability
What are the odds of chronic disability from not treating back pain?
Recently viewed on this device
Stored locally — clear anytime.
Pick challenger
The chronic-use regime is where the numbers get interesting. For a US adult who takes an OTC NSAID three or more times a week across twenty years of adult life without a proton-pump inhibitor, the cumulative probability of a clinically significant adverse event (an upper GI bleed, an acute kidney injury, an NSAID-attributable cardiovascular event, or acetaminophen-linked hepatotoxicity) lands somewhere around 1 in 10, bracketed between 1 in 20 and 1 in 5 depending on age and comorbidity. Lanas and colleagues’ nationwide cohort put the crude major-GI-event rate at roughly 122 per 100,000 person-years averaged across the full population, with 15.3 deaths per 100,000 NSAID users per year once you concentrate the hazard on the subset that actually takes the drugs. Marcum and Hanlon’s US figures are in the same neighbourhood: about 41,000 hospitalizations and 3,300 deaths per year from NSAID adverse effects among older adults alone. Acetaminophen runs on a different track but ends up in a similar order of magnitude, accounting for roughly half of acute liver failure cases in the US per Lee’s Acute Liver Failure Study Group registry, with a 30% mortality rate once ALF develops.
The interesting feature of this entry is the shape of the perceived/actual gap. OTC status itself is a powerful perceptual signal: drugs you can put in a grocery basket next to the toothpaste are drugs that, most adults reason, cannot plausibly cause a hospital admission. That reasoning was built for short-course use, which is overwhelmingly what the OTC label envisions, and it breaks under chronic use. The annual hazard in habitual users is small in absolute terms (on the order of half a percent per year for a middle-aged user, closer to a full percent for someone over 65), but it compounds. Twenty years of a habit that looks like “a couple of ibuprofen most mornings for the knee” produces a cumulative event probability in the neighbourhood of lifetime car-crash mortality, and the public intuition has not caught up to that arithmetic. This is one of the entries where actual risk exceeds perceived risk by roughly an order of magnitude, which is why it is tagged underrated rather than calibrated.
Where the headline doesn’t apply: everywhere outside the tightly-defined chronic-use subgroup. Occasional users, younger adults without GI, renal, or cardiovascular risk factors, and people taking a short course for a specific acute indication sit at or near baseline — the per-event hazard at label dose and label frequency is genuinely small, and the compounded figure on this page is about exposure duration, not exposure itself. Specific-agent choice matters: naproxen has the lowest cardiovascular signal in the Trelle network meta-analysis; ibuprofen and diclofenac sit higher; low-dose aspirin carries its own non-trivial share of NSAID- attributable GI mortality in Lanas’s data. Acetaminophen is largely benign for the liver when kept below 4 g/day in the absence of alcohol, chronic liver disease, or a second acetaminophen-containing product in the same medicine cabinet. The exception, and the reason acetaminophen leads the US acute-liver-failure table, is the interaction with heavy alcohol use and with acetaminophen-containing prescription opioids: roughly half of the US acetaminophen ALF cases in Lee’s registry are unintentional overdoses driven by exactly those two modifiers, and the combined hazard is far outside anything the OTC label implies.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] American Journal of Gastroenterology (Lanas et al.) — A nationwide study of mortality associated with hospital admission due to severe gastrointestinal events and those associated with nonsteroidal antiinflammatory drug use
A nationwide study of mortality associated with hospital admission due to severe gastrointestinal events and those associated with nonsteroidal antiinflammatory drug use- Statistic
121.9 major GI events per 100,000 persons per year; 21.0-24.8 NSAID/aspirin-attributable deaths per million population per year; 15.3 deaths per 100,000 NSAID/aspirin users per year; up to one-third of those deaths attributable to low-dose aspirin- Excerpt
“"Mortality rates associated with either major upper or lower GI events are similar but upper GI events were more frequent. [...] NSAID/aspirin use was associated with 21.0-24.8 deaths per million people per year, of which up to one-third can be attributed to low-dose aspirin use." ”
- Source data from
- 2005-08-01
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Lanas 2005 is the canonical nationwide incidence anchor for NSAID-attributable serious GI events. 121.9 events per 100,000 person- years averaged across the full population corresponds to ~0.12% per year; concentrating that hazard on the subset of the population that actually uses NSAIDs chronically pushes the per-user rate to roughly 0.5-1% per year depending on age and comorbidity. 15.3 deaths per 100,000 users per year is the per-user case-fatality anchor. The one-third attributable to low-dose aspirin matters because the cardioprotective-aspirin population is partially but not entirely disjoint from the OTC-painkiller population in the public conversation. Used as the primary relative-risk and event-incidence basis for the ~0.5% annual per-chronic-user hazard used in the normalized compounding.
- Independence
- Lanas 2005 is a Spanish nationwide hospital-admissions cohort. Generalizing to US adults assumes comparable NSAID prescribing and GI epidemiology; the US per-capita figures (Marcum/Hanlon 2010 below) agree to within a factor of two. Treat as an independent European anchor cross-checked against the US estimate.
-
[2] Annals of Long-Term Care (Marcum & Hanlon) — Recognizing the Risks of Chronic Nonsteroidal Anti-Inflammatory Drug Use in Older Adults
Recognizing the Risks of Chronic Nonsteroidal Anti-Inflammatory Drug Use in Older Adults- Statistic
NSAID use causes an estimated 41,000 hospitalizations and 3,300 deaths each year among older adults in the US; peptic ulcer hospitalization rises from 1 per 1,000/year under age 50 to 2-6 per 1,000/year over 65; acute renal failure risk doubles within 30 days of initial NSAID use; ~40% of adults 65+ fill an NSAID prescription annually- Excerpt
“"NSAID use causes an estimated 41,000 hospitalizations and 3300 deaths each year among older adults. [...] the rate of hospitalizations for peptic ulcer disease (PUD) increases with age, from 1 per 1000 per year in those under 50 to 2-6 per 1000 per year in older adults (>65 yr). [...] the risk of ARF was increased nearly twofold for all NSAIDs (nonselective and COX-2 selective) within 30 days of initial use." ”
- Source data from
- 2010-09-01
- Accessed
- 2026-04-16 · archived copy
- Calculation
- The 41,000 hospitalizations / 3,300 deaths per year figure for older US adults on NSAIDs is the domestic anchor. Across a denominator of ~25 million older adult NSAID users (~40% of ~60 million US 65+ adults), that is roughly 164 hospitalizations per 100,000 users per year and 13 deaths per 100,000 users per year — within a factor of two of the Lanas Spanish figures, confirming the cross-population consistency. Peptic ulcer hospitalization rate of 2-6 per 1,000/year in older adults is the basis for the elderly row in the regional_breakdown and the ~1% annual GI-event hazard in the headline compounding for the over-65 subgroup.
- Independence
- Marcum/Hanlon synthesize Lanas and the ARAMIS database; partially dependent on Lanas 2005 above for the European incidence figures but draws on distinct US claims data for the US-specific hospitalization and mortality totals. Treat as a related US-focused line of evidence.
-
[3] Hepatology Research (Lee) — Acetaminophen-related acute liver failure in the United States
Acetaminophen-related acute liver failure in the United States- Statistic
Acetaminophen accounts for approximately 50% of acute liver failure cases in the US; 30% mortality once acute liver failure develops; nearly half of US acetaminophen ALF cases are unintentional overdoses of therapeutic products- Excerpt
“"Acetaminophen overdoses are the number one cause of acute liver failure (ALF) in the United States; they account for 50% of all cases of ALF and carry a 30% mortality. [...] nearly half of U.S. cases are unintentional, resulting from overuse of acetaminophen-containing pain relief products rather than deliberate self-harm." ”
- Source data from
- 2008-11-01
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Lee's US Acute Liver Failure Study Group registry is the definitive source that acetaminophen is the single leading cause of ALF in the US. The unintentional-overdose fraction (~50% of acetaminophen ALF cases) is load-bearing: these are chronic users who stayed at or slightly above the 4 g/day label limit, often while also drinking alcohol or combining an OTC acetaminophen product with an acetaminophen-containing prescription opioid. The ~500 deaths per year US figure commonly cited in the FDA literature is the product of ~1,600 acetaminophen-associated ALF cases per year and the ~30% fatality rate. Used as the anchor for the acetaminophen component of the compounded hazard and for the alcohol-acetaminophen interaction multiplier.
- Independence
- The US Acute Liver Failure Study Group registry is the primary data source; Blieden 2014 (below) and subsequent FDA regulatory documents draw directly on Lee's registry data. Treat Lee 2008, Blieden 2014, and the FDA figures as a single largely-dependent line of evidence.
-
[4] Expert Review of Clinical Pharmacology (Blieden et al.) — A perspective on the epidemiology of acetaminophen exposure and toxicity in the United States
A perspective on the epidemiology of acetaminophen exposure and toxicity in the United States- Statistic
~30,000 patients hospitalized yearly for acetaminophen toxicity in the US; ~6% of adults take doses exceeding 4 g/day at least occasionally; up to 50% of acetaminophen overdoses are unintentional; 17% of adults with unintentional overdose experience liver damage- Excerpt
“"Approximately 6% of adults are prescribed acetaminophen doses of more than 4 g/day [...] 30,000 patients are hospitalized yearly for acetaminophen toxicity [...] up to 50% of overdoses are unintentional in nature [...] 17% of adults with unintentional overdose experience liver damage." ”
- Source data from
- 2014-05-01
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Blieden 2014 is the behavioural-epidemiology source: ~6% of adult acetaminophen users exceed the 4 g/day label ceiling on at least some days, and ~30,000 are hospitalized per year with acetaminophen toxicity. 30,000 / ~50 million US adults using acetaminophen at least weekly ≈ 0.06% per year hospitalization rate — small at the population level, but concentrated in the chronic-user-plus-alcohol or chronic-user-plus-combination-product subgroups. This is the source for the regional_breakdown row on chronic acetaminophen with alcohol and the personal_factor_multiplier for heavy alcohol use.
- Independence
- Blieden draws on the same Acute Liver Failure Study Group registry as Lee 2008 above, cross-linked with Poison Control Center call data and the Kaufman Slone Survey of OTC use prevalence. Treat as partially dependent on Lee 2008; independent on the prevalence-of-use data.
-
[5] BMJ (Trelle et al.) — Cardiovascular safety of non-steroidal anti-inflammatory drugs: network meta-analysis
Cardiovascular safety of non-steroidal anti-inflammatory drugs: network meta-analysis- Statistic
Network meta-analysis of 31 RCTs and 116,429 patients: ibuprofen associated with a stroke RR of 3.36 (95% credibility interval 1.00-11.6); diclofenac with a cardiovascular-death RR of 3.98 (1.48-12.7); naproxen appeared least harmful- Excerpt
“"Although uncertainty remains, little evidence exists to suggest that any of the investigated drugs are safe in cardiovascular terms. Naproxen seemed least harmful. Cardiovascular risk needs to be taken into account when prescribing any non-steroidal anti-inflammatory drug." ”
- Source data from
- 2011-01-11
- Accessed
- 2026-04-16 · archived copy
- Calculation
- Trelle 2011 is the cardiovascular half of the chronic-NSAID hazard picture. The RCTs pooled here were mostly prescription-dose trials, so the absolute CV event rate attributable to OTC-dose ibuprofen is lower than the abstract's relative risks imply. For a middle-aged chronic user the CV component of the ~0.5% annual hazard is on the order of 0.1-0.2% per year, rising substantially in users with pre-existing cardiovascular disease. The differential hazard across specific NSAIDs (naproxen < ibuprofen < diclofenac) is the basis for the caveat distinguishing between naproxen and the other OTC NSAIDs.
- Independence
- Trelle 2011 is a pooled RCT meta-analysis, methodologically distinct from the Lanas and Marcum/Hanlon observational cohorts above. Independent line of evidence for the cardiovascular component.