Ahmad · JAMA network open 2024 · prospective cohort study · n=25315

Mediterranean Diet Adherence and Risk of All-Cause Mortality in Women.

Cited 41 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective non-randomized cohort study with long-term follow-up

PubMed 38819819 · doi:10.1001/jamanetworkopen.2024.14322 · record verified 2026-08-27

What was done

This prospective cohort study examined 25,315 initially healthy female participants from the Women's Health Study (enrolled 1993–1996) followed for a mean of 24.7 years. Mediterranean diet adherence was computed on a 0 to 9 scale from baseline food-frequency questionnaires. Thirty-three blood biomarkers spanning lipids, lipoproteins, inflammation, insulin resistance, and metabolites were measured at baseline. Cox proportional hazards regression assessed all-cause mortality risk, and mediation analyses evaluated the relative contributions of these cardiometabolic biomarkers to the association.

What was found

Over follow-up, 3,879 deaths occurred among the 25,315 participants. Compared with low adherence (score 0–3), multivariable-adjusted hazard ratios for mortality were 0.84 (95% CI, 0.78–0.90) for middle adherence (score 4–5) and 0.77 (95% CI, 0.70–0.84) for upper adherence (score 6–9; P for trend < .001). After further adjustment for lifestyle factors, hazard ratios were 0.92 (95% CI, 0.85–0.99) and 0.89 (95% CI, 0.82–0.98), respectively (P for trend = .001). Biomarker mediation analysis showed small molecule metabolites explained 14.8% of the association, inflammatory biomarkers explained 13.0%, triglyceride-rich lipoproteins explained 10.2%, body mass index explained 10.2%, and insulin resistance explained 7.4%; branched-chain amino acids, HDL, LDL, glycemic measures, and hypertension each contributed <3%.

Why it matters

This study provides evidence on the specific molecular pathways linking Mediterranean diet adherence to longevity, identifying small-molecule metabolism and inflammation as primary contributors over traditional lipid or blood pressure pathways.

Limits

The study is observational and prone to residual confounding. Diet and biomarkers were assessed only at baseline and could change over 25 years. The sample was predominantly White (94.9%) female health professionals, which limits generalizability to men and racially diverse populations. Diet data relied on self-reported food-frequency questionnaires.

Cited by