Body weight and death risk worldwide

Overview

Global BMI Mortality Collaboration. Lancet 2016 · PMID 27423262 · DOI 10.1016/S0140-6736(16)30175-1

Severe obesity (BMI 40+) was linked to nearly 3 times the death risk of normal weight.

Does the study support the specific claim? No · Methodological quality: high

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Global BMI Mortality Collaboration

Journal

Lancet

Year

2016

Identifier

PMID 27423262 DOI 10.1016/S0140-6736(16)30175-1

Verification

verified – PubMed efetch 2026-10-25

Retracted

no

Corrections

none

Extent checked

full text

Conflicts of interest and funding

UK MRC, British Heart Foundation, NIHR Cambridge BRC, US NIH; funders had no role. Several writing-committee members report industry funding or fees (e.g. one full-time Roche employee with shares; fees from Novartis, Amgen, AstraZeneca, Boehringer Ingelheim, Janssen; research funding from Nestle, GSK, Merck, Pfizer).

2. Research question

BMI and all-cause mortality

3. Study design and population

Study design

Meta-analysis – Individual-participant-data meta-analysis (Global BMI Mortality Collaboration) of 239 prospective cohort studies with a prespecified analysis plan (2013); PRISMA-IPD reporting

Population

Adults aged 20-90 years, BMI 15-<60, in 32 countries (Asia, Australia/New Zealand, Europe, North America); primary analysis restricted to never-smokers without known chronic disease at baseline who survived 5 years

Country

Asia, Australia/NZ, Europe, North America

Period

Studies published January 1970 - January 2015; median follow-up 13.7 years (IQR 11.4-14.7)

Sample size

10,625,411 participants in 239 studies; primary analysis 3,951,455 participants in 189 studies with 385,879 deaths

Inclusion and exclusion criteria

Prospective cohorts or consortia with >=100,000 participants, data on weight, height, age, sex; not selected on chronic disease; deaths recorded; median follow-up >=5 years. Participants with BMI <15 or >=60, age <20 or >90 excluded; primary analysis excluded ever-smokers, pre-existing chronic disease and first 5 years of follow-up

4. Methods and endpoints

Exposure / intervention

Baseline BMI (nine categories)

Comparator

BMI 22.5-<25

Endpoints

All-cause mortality; cause-specific mortality (CHD, stroke, other CVD, cancer, respiratory)

Statistical methods

Study-specific Cox models stratified by age and sex; inverse-variance random-effects meta-analysis; floating variance CIs; fractional polynomial nadir; HR per 5 kg/m2; population-attributable fractions; I2; multiple sensitivity analyses

5. Actual results

Results

Primary analysis (reference BMI 22.5-<25.0): BMI 15-<18.5 HR 1.51 (1.43-1.59); 18.5-<20 1.13 (1.09-1.17); 20-<22.5 1.00 (0.98-1.02); 25-<27.5 1.07 (1.07-1.08); 27.5-<30 1.20 (1.18-1.22); obesity grade 1 1.45 (1.41-1.48); grade 2 1.94 (1.87-2.01); grade 3 (40-<60) 2.76 (2.60-2.92). Europe grade 3 HR 3.04 (2.84-3.27). HR per 5 kg/m2 above 25: 1.31 (1.29-1.33) overall; 1.52 (1.47-1.56) at age 35-49 vs 1.21 (1.17-1.25) at 70-89; men 1.51 (1.46-1.56) vs women 1.30 (1.26-1.33). Population-attributable fraction of all-cause mortality due to overweight/obesity: 19% North America, 16% Australia/NZ, 14% Europe, 5% east Asia. Crude analyses without exclusions gave inverted HRs for overweight and less than half the excess risk for grade 1 obesity.

Absolute risks

Absolute death rates in appendix only (not in main text file); population-attributable fractions 14% (Europe) to 19% (North America); absolute excess death rate for grade 1 obesity about three times greater in men than in women (appendix).

Authors' conclusion

Associations of overweight and obesity with higher all-cause mortality are broadly consistent across four continents, supporting strategies to combat the entire spectrum of excess adiposity.

6. Strengths

• Individual-participant data from about 97% of eligible participants; prespecified analysis plan
• Strict handling of reverse causation and smoking (never-smokers, no chronic disease, 5-year exclusion)
• Very large number of deaths; precise estimates
• Robust to many sensitivity analyses

7. Weaknesses

• Observational; BMI is the only adiposity measure (authors)
• Study results generally not adjusted for ethnicity or socioeconomic status (authors)
• South Asia, Africa, Latin America poorly represented (authors)
• No formal risk-of-bias assessment of contributing cohorts
• Single baseline BMI; restriction to never-smokers may limit generalisability
• Does not address effects of intentional weight loss or surgery

8. Risk of bias

Orienting tool

AMSTAR 2 (adapted to IPD meta-analysis of cohorts)

Type of assessment

structured assessment based on the full text by a single reviewer (no formal double review)

Full formal assessment

no

Overall risk of bias

some concerns

Rationale

Changed from low: methodologically strong IPD meta-analysis, but underlying data are observational with age/sex adjustment only and no formal per-study risk-of-bias assessment; causal interpretation relies on restriction strategies and concordant Mendelian randomisation evidence. Structured single-reviewer judgement from full text; a formal assessment would require two independent reviewers and the protocol.

Full-text review (25 October 2026): Added full HR table (nine BMI groups), regional (Europe grade 3 HR 3.04), age and sex modification, population-attributable fractions, inclusion/exclusion and restriction strategy, funding/COI; AMSTAR 2 items added; bias changed low -> some_concerns (observational, age/sex-only adjustment, no per-study RoB).

9. Assessment of the statistical interpretation

Very precise relative estimates; grade 3 obesity (BMI 40-<60) is associated with about 2.8 times the all-cause mortality of BMI 22.5-<25 (about 3.0 in Europe). These are associations in never-smokers without disease; they provide context for the excess mortality of people eligible for bariatric surgery but do not show the effect of weight loss or surgery.

10. Relevance to the specific claim

Context: severe obesity itself roughly triples mortality, explaining why surgical patients remain below general-population life expectancy (C-BAR-10).

11. Independent contrary findings

12. Transferability to Germany and Europe

Includes Europe.

13. Overall methodological judgement

Strong context evidence.

14. Open questions

Effect of intentional weight loss on mortality; body composition; under-represented regions (including the Middle East/Egypt).

15. Linked fact checks

• Sleeve gastrectomy (and bariatric surgery in general) is a lethal operation that kills patients / shortens life. – Studies contradicting or limiting the claim
• Even the largest long-term study (SOS) showed that surgery patients still die 5.5 years earlier than the general population. – Studies contradicting or limiting the claim