Primary Care Corner with Geoffrey Modest MD: Changing diet changes lifespan

by Geoffrey Modest MD

A recent review of the data from the Nurses’ Health Study and the Health Professionals Follow-up Study from 1998 to 2010 found that mortality tracks with changes in diet (see DOI: 10.1056/NEJMoa1613502).


— 47,994 women in the Nurses’ Health Study and 25,745 men in the Health Professionals Follow-up Study were assessed, correlating changes in diet quality from over 8- to 16-year time periods and hazard ratios for total and cause-specific mortality

— the diet metrics used were the Alternate Healthy Eating Index–2010 score (Including 11 food components, each scored 0-10), the Alternate Mediterranean Diet score (including 9 components, each scored 0- 1), and the Dietary Approaches to Stop Hypertension (DASH) diet score (including 8 components, each scored 1- 5)

–Dietary changes were assessed at baseline and every 4 years thereafter

–Covariates: information on the participants lifestyle and risk factors for cardiovascular disease were assessed and updated every other year, including age, weight/BMI, smoking status, use of aspirin, multivitamins, postmenopausal hormone replacement therapy/oral contraceptives, menopausal status, physical activity, hypertension/ hypercholesterolemia/diabetes recently diagnosed by a physician, alcohol use, smoking, quintiles of total energy intake and changes in these levels

— They assess both shorter-term clinical changes (baseline to 8  year follow-up, 1986-1994) as well as longer-term changes (baseline to 12 and 16 year follow-ups, 1986- 2002)

— Those with greater improvement in diet quality were younger, had a lower baseline diet score (ie less good diet), engaged in more physical activity, had lower alcohol intake, and in general reported increased consumption of whole grains, vegetables, and n-3 fatty acids, and less sodium intake.

— Those with consistently high quality diet at baseline and 12 years later were older, had lower BMI, were less likely to be current smokers, and were more physically active than those with consistently poor diet quality.


— In the Nurses’ Health Study: 5967 deaths, including 1115 deaths from cardiovascular disease and 2089 deaths from cancer, over 544,973 person-years of follow-up

— In the Health Professionals Follow-up Study​, 3979 deaths, including 1226 deaths from cardiovascular disease and 1192 deaths from cancer, over 286,402 person-years of follow-up

— All-cause mortality:

​– those with the greatest diet quality improvement (13-33%), relative to those with a stable diet, had an 9% decreased mortality by the Alternate Healthy Eating Index–2010 score, HR 0.91 (0.85-0.97); 16% by the Alternate Mediterranean Diet score, HR 0.84 (0.78-0.91); and 11% by the DASH diet score, HR 0.89 (0.84- 0.95​)

​– those with decreasing diet quality (vs stable change) had a 12% increased mortality by the Alternate Healthy Eating Index–2010 score, HR 1.12 (1.05-1.19); 6% by the Alternate Mediterranean Diet score, HR 1.06 (0.99-1.13, nonsignificant); and 6% by the DASH diet score, HR 1.06 (1.00-1.12​)​

​–those with the most impressive decrease in death from any cause were those who had the largest improvement from poor quality diet vs those who remained on a poor quality diet 12 years later: 15% decreased mortality by the Alternate Healthy Eating Index–2010 score; 23% by the Alternate Mediterranean Diet score; and 28% by the DASH diet score

–those with a consistently high quality diet had a 14% decreased mortality by the Alternate Healthy Eating Index–2010 score; 11% by the Alternate Mediterranean Diet score; and 9% by the DASH diet score

— a 20 percentile improvement in diet scores was significantly associated with a 8- 17% decrease in total mortality with the use of the 3 diet indexes

— cardiovascular mortality:

— a 20 percentile improvement in diet scores was associated with a 7-15% decrease with the use of the Alternate Healthy Eating Index–2010  and the Alternate Mediterranean Diet scores

— In the fully-adjusted model, a 20 percentile improvement during the 12 year period of 1986-1998 was associated with:

— by Alternate Healthy Eating Index score —  death from any cause: 17%, death from cardiovascular cause: 15%, death from cancer: 6% (cancer nonsignificant)

— by Alternate Mediterranean Diet score — death from any cause 8%, death from cardiovascular cause 7%, death from cancer 2% (cancer nonsignificant)

— by DASH score — death from any cause 10%, death from cardiovascular disease 4% (nonsignificant), death from cancer 9% (significant, and largely due to deaths from lung cancer)

— there was a graded improvement in death from any cause in all of the indexes​ when assessing the 20 percentile diet improvement from 8 years to 12 years ​to 16 years. Also true for cardiovascular deaths but not cancer


— prior studies have shown that the above dietary indexes correlate with reductions in death from any cause, death from cardiovascular disease, and deaths from cancer, leading to the scientific report of the 2015 Dietary Guidelines Advisory Committee recommending these diet indexes(see​ ). These prior studies looked at baseline diet and subsequent events. This study, by contrast, looked at changes in diet and clinical outcomes.

— overall, this study found that a dietary change over 8 years, their earliest evaluation, did find a decreased all-cause mortality by each of the diet measurements. This improvement was more impressive after the 12 year evaluation and more so after 16 years. We do not have data on shorter term improvements: would patients see benefit after only 4 years?  2 years?  There have been studies on short-term changes in fat consumption and changes in the lipid composition of atheromatous coronary artery plaques (eg: within 60 days of changing diet to increased fish oils, there was a shift to a thicker fibrous cap with less inflammation, presumably making the atheromas more stable and less likely to rupture and cause an acute coronary syndrome). This might provide a mechanism for clinical benefit at a much earlier time period (though the intervention would, no doubt, need to be maintained).

— of course, the results of the long-term nutrition studies, including this one, are based on observational data and, as such​, there may have uncontrolled potential biases. So, one cannot attribute causation​: those people who changed their diets to healthier ones may well have changed other aspects of their life (exercise, or other unmeasured healthy lifestyle changes, either known or unknown), and perhaps these other changes were really the ones leading to increased longevity, not the dietary change.

so, I think this study really does add a lot to buttress our efforts in recommending that patients improve their diet: there were very impressive gains in total and cardiovascular mortality by significantly improving diet quality as measured by these dietary indexes, and after only 8 years (the shortest interval measured here).  In fact the largest improvements were in those going from a poor quality diet to a good quality one.  so though the longer one eats a healthy diet, the better, it’s never to late to change….  and perhaps this is a significant motivator for some patients.