A prospective study of carbohydrate intake and risk of all-cause and specific-cause mortality

Abstract

To investigate the associations between carbohydrate intake and the risk of overall and specific-cause mortality in a prospective cohort study. Diet was measured using 24h dietary recalls. Underlying cause of death was identified through linkage to the National Death Index. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using Cox proportional hazards regression. During a median follow-up of 7.1years among 35,692 participants who aged 20-85years, a total of 3854 deaths [783 cardiovascular disease (CVD)-specific and 884 cancer-specific death] were identified. Carbohydrate intake was not associated with risk of overall mortality (multivariable-adjusted HR comparing extreme quartiles 1.03, 95% CI 0.94, 1.13, ptrend = 0.799), while higher fiber intake was associated with lower mortality risk (HR 0.86, 95% CI 0.77, 0.95, ptrend = 0.004). Replacing 5% of energy from carbohydrate with both plant fat and plant protein was associated with 13% (95% CI 8%, 17%) and 13% (95% CI 3%, 22%) lower risk of total and CVD mortality, respectively. Whereas a positive or null association was found when replacing carbohydrate withboth animal fat and animal protein. Higher carbohydrate-to-fiber ratio was associated with increased risk of overall (HR 1.20, 95% CI 1.09, 1.33, ptrend < 0.001) and cancer-specific (HR 1.17, 95% CI 0.95, 1.44, ptrend = 0.031) mortality. Our findings suggested that high fiber diet or diet with low carbohydrate-to-fiber ratio was associated with lower long-term death risk, and provided evidence for the health benefit from dietary substitution of both plant fat and plant protein for carbohydrate.