Diet soda, stroke and dementia in Framingham

Overview

Pase MP et al.. Stroke 2017 · PMID 28428346 · DOI 10.1161/STROKEAHA.116.016027

Daily diet-soda drinkers had about three times more ischaemic strokes and Alzheimer's disease, but with very few cases and wide uncertainty.

Does the study support the specific claim? Partly · Methodological quality: moderate

Original publication

Plain-language summary

Scientific in-depth review

1. Full reference

Authors

Pase MP, Himali JJ, Beiser AS, Aparicio HJ, Satizabal CL, Vasan RS, Seshadri S, Jacques PF

Journal

Stroke

Year

2017

Identifier

PMID 28428346 DOI 10.1161/STROKEAHA.116.016027 · PMC5405737

Verification

verified – PubMed E-utilities efetch, 2026-10-25

Retracted

no

Corrections

none

Extent checked

full text

Conflicts of interest and funding

NHMRC (Pase), NHLBI Framingham contracts, NIA and NINDS grants, USDA ARS agreement, Evans Scholar award; all other authors report no disclosures.

2. Research question

Are sugar- or artificially sweetened beverages associated with incident stroke or dementia?

3. Study design and population

Study design

Cohort study – Prospective cohort (Framingham Offspring); FFQ at exams 5-7; 10-year surveillance; Cox models.

Population

Framingham Heart Study Offspring participants (predominantly of European descent): stroke sample aged >45 (mean 62, 45% men); dementia sample aged >60 (mean 69, 46% men)

Country

USA

Period

FFQ at exams 5 (1991-1995), 6 (1995-1998), 7 (1998-2001); 10-year surveillance from exam 7

Sample size

2888 (stroke; 97 strokes, 82 ischaemic) and 1484 (dementia; 81 cases, 63 AD); analysed models smaller (e.g. cumulative ischaemic stroke Model 2: 58 events/2137; AD Model 3: 57/1285)

Inclusion and exclusion criteria

Excluded prevalent stroke or significant neurological disease and age <=45 (stroke analysis); prevalent dementia, MCI or neurological disease and age <=60 (dementia analysis); cumulative intake required exam 7 plus >=1 earlier FFQ

4. Methods and endpoints

Exposure / intervention

Recent (exam 7) and cumulative (average exams 5-7) intake of total sugary beverages, sugar-sweetened soft drinks and artificially sweetened soft drinks (ASB categories 0/week, >0-6/week, >=1/day)

Comparator

0/week (abstract states '<1 per week') or <1/day for total sugary beverages

Endpoints

Incident stroke, ischaemic stroke, all-cause dementia, Alzheimer's disease

Statistical methods

Cox regression; Model 1 (age, sex, education, calories), Model 2 (+diet quality DGAI, physical activity, smoking), Model 3 (Model 1 + SBP, hypertension treatment, CVD, AF, LVH, cholesterol, HDL, diabetes, waist-hip ratio, APOE e4); mediation analyses; no correction for multiple comparisons

5. Actual results

Results

ASB daily cumulative vs none (Model 2): ischaemic stroke HR 2.96 (95% CI 1.26-6.97); AD HR 2.89 (1.18-7.07). Recent ASB and all stroke: >0-6/week 1.78 (0.98-3.23), >=1/day 1.87 (0.90-3.90) (Model 2); 1.59 (0.92-2.75) and 1.79 (0.91-3.52) (Model 3). Dementia associations with cumulative ASB were no longer significant in Model 3: all-cause dementia >=1/day HR 1.70 (0.80-3.61), AD 2.03 (0.83-4.97). Excluding prevalent diabetes (Model 1): dementia HR 2.45 (1.07-5.59), AD 3.23 (1.22-8.52); diabetes a partial mediator. Excluding prevalent hypertension, all-stroke association attenuated (>=1/day HR 1.43, 0.40-5.11). Sugar-sweetened soft drinks and total sugary beverages not associated with stroke or dementia (e.g. SSB >3/week all stroke HR 0.69, 0.29-1.62).

Absolute risks

not reported in full text (event counts only: 97 strokes, 81 dementia cases over 10 years)

Authors' conclusion

Artificially sweetened soft drink consumption was associated with a higher risk of stroke and dementia.

6. Strengths

• Prospective, community-based cohort with repeated validated FFQs
• Rigorous stroke and dementia adjudication (DSM-IV, NINCDS-ADRDA)
• Sequential adjustment models and mediation analyses

7. Weaknesses

• Few events; very wide CIs
• Reverse causation/confounding by indication: daily ASB drinkers had more diabetes (22% vs 7%), higher BMI and hypertension treatment
• Dementia associations not robust to cardiometabolic adjustment (Model 3)
• Many exposure-outcome-model combinations without multiple-comparison correction (authors)
• Self-reported FFQ; single, predominantly white cohort (authors)
• Reference category inconsistently described (0/week vs <1/week)

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

high

Rationale

Structured single-reviewer judgement from full text; a formal assessment would require two independent reviewers and a protocol.

Full-text review (25 October 2026): Added model structure, Model 3 results (dementia HR 1.70, 0.80-3.61; AD 2.03, 0.83-4.97 - not significant), mediation analyses, baseline diabetes imbalance (22% vs 7%), no multiplicity correction, funding; bias domains added, overall unchanged (high).

9. Assessment of the statistical interpretation

Imprecise estimates from few events, selected from many comparisons; the dementia signal disappears after adjustment for cardiometabolic factors (HR 1.70, 0.80-3.61), consistent with reverse causation (people with diabetes choosing diet drinks). Sugar-sweetened drinks showed no association, which the authors call 'intriguing'.

10. Relevance to the specific claim

Only direct evidence for C-SOD-06.

11. Independent contrary findings

WHI (I-SOD-06) found a smaller association (ischaemic stroke HR 1.31 at >=2/day).

12. Transferability to Germany and Europe

Moderate.

13. Overall methodological judgement

Hypothesis-generating; magnitude unreliable.

14. Open questions

Replication with adequate case numbers; mechanism.

15. Linked fact checks

• One diet soda per day nearly triples the risk of ischaemic stroke and of Alzheimer's dementia. – Studies with mixed or unclear results

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

One diet soda per day was associated with nearly triple the risk of ischemic stroke and Alzheimer's dementia (Framingham follow-up of 4,000+ adults).

A. What was actually studied?

Associations of SSB and ASB intake with stroke and dementia.

B. What did the authors find?

HR 2.96 and 2.89 for daily cumulative ASB intake; SSB not associated.

C. What conclusions do the authors themselves draw?

ASB consumption was associated with higher risk.

D. What does the archive derive from it?

Presents the HRs as a tripling of risk and attributes them partly to microbiome effects (Nature 2014).

E. Is this derivation scientifically justified?

Numbers correct for Model 2, but 'more than 4,000' is inaccurate (2,888 and 1,484 in overlapping samples); the archive omits that the dementia association was no longer significant after cardiometabolic adjustment, the strong diabetes imbalance, the very wide CIs and the absence of multiple-comparison correction; the microbiome link is speculative.

Adequacy of the interpretation (by the archive)

partly correct but over-interpreted