Cohort studycited by the Tayebat archivefull text checked★ key study A-SOD-05 in-depth review
Sugary and diet drinks and death in 118,000 US adults
Long-Term Consumption of Sugar-Sweetened and Artificially Sweetened Beverages and Risk of Mortality in US Adults.
Overview
Malik VS et al.. Circulation 2019 · PMID 30882235 · DOI 10.1161/CIRCULATIONAHA.118.037401
Two or more sugary drinks a day: 21% higher death risk; diet drinks linked only at the highest intake and only in women.
Does the study support the specific claim? Partly · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Are SSB and ASB associated with total and cause-specific mortality?
- How was the study done?
Two prospective cohorts (HPFS 1986-2014; NHS 1980-2014); repeated FFQs; Cox models.
- How many people, animals or samples were studied?
118,363 (37,716 men; 80,647 women); 36,436 deaths; Health professionals and nurses free of chronic disease at baseline
- What was actually found?
SSB >=2/d vs <1/mo: total mortality HR 1.21 (1.13-1.28); CVD mortality 1.31 (1.15-1.50); cancer mortality 1.16 (1.04-1.29). ASB: only highest category, total 1.04 (1.02-1.12), CVD 1.13 (1.02-1.25); association in NHS but not HPFS (P interaction 0.01); not with cancer mortality.
- How large was the observed effect?
HR 1.21 total, 1.31 CVD (SSB)
- What do the results mean in everyday life?
Regular sugary drinks are linked with earlier death.
- What are the study's limitations?
Health professionals (limited generalisability); Observational design; residual confounding possible; Self-reported dietary intake
- What does it say about Elawady's specific claim?
- Do the results apply to Germany and Europe?
High.
- What can explicitly not be concluded from this study?
That diet drinks are as harmful as sugary ones; causality
Scientific in-depth review
1. Full reference
- Authors
- Malik VS, Li Y, Pan A, De Koning L, Schernhammer E, Willett WC, Hu FB
- Journal
- Circulation
- Year
- 2019
- Identifier
- PMID 30882235 DOI 10.1161/CIRCULATIONAHA.118.037401 · PMC6488380
- Verification
- verified – PubMed E-utilities efetch, 2026-10-25
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
NIH grants (UM1 CA186107, UM1 CA176726, UM1 CA167552, P01 CA87969, P01 CA055075, R01 HL034594, HL088521, HL35464, HL60712). WCW and VSM are on a pro bono retainer for the Center for Science in the Public Interest in litigation related to SSBs and have consulted for the City of San Francisco on soda warning labels.
2. Research question
Are SSB and ASB associated with total and cause-specific mortality?
3. Study design and population
- Study design
- Cohort study – Two prospective cohorts (HPFS 1986-2014; NHS 1980-2014); repeated FFQs; Cox models.
- Population
US female nurses (NHS, aged 30-55 at entry) and male health professionals (HPFS, aged 40-75), predominantly non-Hispanic white, free of diabetes, CVD and cancer at baseline
- Country
USA
- Period
NHS 1980 - June 30, 2014 (34 years); HPFS 1986 - January 31, 2014 (28 years)
- Sample size
118,363 (80,647 women; 37,716 men); 36,436 deaths (7,896 CVD, 12,380 cancer) over 3,415,564 person-years
- Inclusion and exclusion criteria
Excluded baseline diabetes, CVD or cancer, >70 FFQ items blank, missing SSB data, implausible energy intake (<500/>3500 kcal women; <800/>4200 kcal men)
4. Methods and endpoints
- Exposure / intervention
SSB (colas, other carbonated and non-carbonated sugar-sweetened drinks; fruit juice excluded) and ASB intake, updated every 4 years; categories <1/mo, 1-4/mo, 2-6/wk, 1-<2/d, >=2/d (ASB also 2-<4/d and >=4/d)
- Comparator
<1/month
- Endpoints
Total, CVD and cancer mortality
- Statistical methods
Age-stratified Cox models per cohort, pooled; time-varying covariates; mutual adjustment SSB/ASB; BMI in models; stopping diet updates after major diagnoses; sensitivity analyses (cumulative average, baseline only, 8-year lag, AHEI, adjustment for intermediate conditions); substitution analysis
5. Actual results
- Results
Pooled SSB vs <1/mo: 1-4/mo 1.01 (0.98-1.04), 2-6/wk 1.06 (1.03-1.09), 1-<2/d 1.14 (1.09-1.19), >=2/d 1.21 (1.13-1.28); p-trend <0.0001; per serving/day HR 1.07 (1.05-1.09); age-adjusted >=2/d 1.52 (1.43-1.61) attenuated by adjustment. Women >=2/d 1.25 (1.16-1.34), men 1.12 (1.00-1.26) (p-interaction 0.02). CVD mortality >=2/d 1.31 (1.15-1.50), per serving 1.10 (1.06-1.14); cancer mortality 1.16 (1.04-1.29), breast cancer mortality 1.34 (1.00-1.80). After adjusting for incident intermediate conditions, SSB per serving 1.05 (1.03-1.07). ASB pooled: 0.96 (0.93-0.99), 0.97 (0.95-1.00), 0.98 (0.94-1.03), >=2/d 1.04 (1.02-1.12) for total; CVD >=2/d 1.13 (1.02-1.25). NHS: >=2/d 1.10 (1.04-1.16), >=4/d 1.30 (1.15-1.46), CVD >=4/d 1.43 (1.10-1.87); HPFS: no association (0.99, 0.91-1.07). NHS ASB association disappeared after adjusting for intermediate conditions (1.00, 0.94-1.06) and in most sensitivity analyses. Replacing 1 serving/day SSB with ASB: total mortality HR 0.96 (0.94-0.98), CVD 0.95 (0.90-0.99), cancer 0.96 (0.93-1.00).
- Absolute risks
not reported as risk differences; deaths and person-years by category given (e.g. NHS SSB >=2/d: 861 deaths in 84,884 person-years vs <1/mo: 11,149 in 1,127,585), crude rates not age-standardised.
- Authors' conclusion
SSB consumption was positively associated with mortality, mainly CVD, with a graded dose association; the association of high ASB intake with total and CVD mortality among women requires confirmation; replacing SSB with ASB was associated with moderately lower mortality; ASBs could replace SSBs among habitual consumers, but high ASB consumption should be discouraged and water emphasised.
6. Strengths
- Very large cohorts with 28-34 years of follow-up and >90% follow-up per cycle
- Repeated validated FFQs; time-varying exposure and covariates
- Graded dose-response, consistency across cohorts for SSB
- Extensive sensitivity analyses and approaches to limit reverse causation
7. Weaknesses
- Observational; residual confounding (SSB drinkers smoke more, poorer diet; AHEI 42.7 vs 50.6)
- Reverse causation likely for ASB (higher BMI and hypertension among ASB drinkers) (authors)
- Predominantly white health professionals; limited generalisability
- Self-reported intake; FFQ validity modest for some beverages (r 0.36 non-cola carbonated)
- Many outcomes and subgroups
- Two authors have SSB-litigation-related pro bono/consulting roles (disclosed)
8. Risk of bias
- Orienting tool
- ROBINS-E
- 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
Structured single-reviewer judgement from full text; a formal assessment would require two independent reviewers and a protocol. Risk of bias is higher for the ASB findings than for SSB.
| Domain | Judgement | Rationale |
|---|---|---|
| Confounding | some concerns | Extensive time-varying adjustment for lifestyle and diet; large attenuation from age-adjusted (1.52) to full model (1.21) indicates confounding, residual confounding possible. |
| Exposure measurement | some concerns | Validated repeated FFQ; moderate validity for some items; measurement error likely non-differential. |
| Selection of participants | low | Healthy baseline cohorts; exclusions prespecified. |
| Post-exposure interventions | some concerns | Diet changes after diagnosis handled by stopping updates; switching SSB to ASB because of illness possible (reverse causation for ASB). |
| Missing data | low | >90% follow-up per cycle; >97% deaths ascertained; carry-forward of missing FFQ values. |
| Outcome measurement | low | Deaths from NDI/vital records with physician review of cause. |
| Selection of reported result | some concerns | Many categories, outcomes and subgroups; ASB signal emphasised only at highest categories in one cohort. |
Full-text review (25 October 2026): Added person-years, per-serving HRs (SSB 1.07 total, 1.10 CVD), sex-specific estimates, attenuation after intermediate conditions, ASB >=4/day NHS findings and their disappearance in sensitivity analyses, SSB-to-ASB substitution HR 0.96, author litigation-related COI; bias domains added, overall unchanged (some_concerns).
9. Assessment of the statistical interpretation
SSB: modest, graded relative association (7% per daily serving; 21% at >=2/day), consistent across cohorts and robust to sensitivity analyses. ASB: weak, non-linear signal confined to very high intake in women, which disappeared after adjustment for intermediate conditions and in most sensitivity analyses; substitution of SSB by ASB was associated with lower mortality.
10. Relevance to the specific claim
Supports C-SOD-03; mixed for C-SOD-10.
11. Independent contrary findings
HPFS null for ASB.
12. Transferability to Germany and Europe
High.
13. Overall methodological judgement
Strong cohort evidence for an association of SSB with mortality; weak and probably confounded evidence for ASB, and the substitution analysis contradicts the idea that diet drinks are as harmful as sugary ones.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | high |
| Directness | direct |
| Reproducibility | replicated |
| Consistency with the overall evidence | consistent |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | essentially correct but incomplete |
14. Open questions
Causality; effects of very high ASB intake; generalisability beyond white US health professionals.
15. Linked fact checks
- Sugar-sweetened soft drinks raise the risk of coronary heart disease, cardiovascular death and early death (e.g., two or more daily servings: +35% coronary heart disease in women; +31% cardiovascular and +21% total mortality). – Studies supporting the claim
- Diet soda is not a safe alternative to sugary soda: artificially sweetened drinks are linked to higher mortality, especially circulatory deaths. – Studies contradicting or limiting the claim
Use by the Tayebat archive
Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.
- What the archive says about the study
34-year follow-up of 118,000+ US adults: two or more daily sugary drinks raised cardiovascular mortality by 31% and total mortality by 21%; ASB at highest intake linked to a 13% rise in CVD mortality.
- A. What was actually studied?
SSB/ASB and mortality.
- B. What did the authors find?
As reported in results.
- C. What conclusions do the authors themselves draw?
SSB associated with mortality; ASB association in women needs confirmation.
- D. What does the archive derive from it?
Uses the ASB result to argue diet soda is not a safe alternative.
- E. Is this derivation scientifically justified?
SSB numbers correct. For ASB the archive omits that the association was only in the highest category (and >=4/day), only in women, disappeared after adjustment for intermediate conditions and in most sensitivity analyses, and that replacing one SSB per day with an ASB was associated with 4% lower total and 5% lower CVD mortality; the authors explicitly suggest ASB as a replacement for habitual SSB consumers.
- Adequacy of the interpretation (by the archive)
- essentially correct but incomplete