Dietary recommendations for disease prevention emphasize making multiple dietary changes, including increasing intake of fruits, vegetables and grains and decreasing fat intake by consuming lean meats and poultry, and low-fat dairy products. The authors of this prospective study examine the all-cause mortality and cause-specific mortality of a large cohort of women, using an overall measure of diet quality derived from the current food-based dietary guidelines.

Design and Participants

The analytic cohort for this study consisted of 42,254 women who returned food frequency questionnaires between 1987 and 1989 (phase 2) as part of the National Cancer Institute's Breast Cancer Detection Demonstration Project (BCDDP). The questionnaire, which consisted of 62-items, was validated for use with older women. Questions about frequency and portion size of foods consumed in the previous year were included in the instrument.

A second questionnaire was mailed during phase 3 of the BCDDP (1993-95). Information about exogenous hormone use, medical history, end-points other than breast cancer, alcohol use, vitamin use, physical activity and updated family and reproductive history was also collected during both phases

Women who did not respond during phases 2 and 3 were interviewed, if possible, by phone. Extensive efforts were made to locate women in phase 3, including a search of the National Center for Health Statistic's National Death Index through December 1993. Phase 2 was considered the baseline for analysis.

Diet Quality Measurement
Diet quality was assessed using a Recommended Foods Score (RFS). The sum of 23 questionnaire items corresponding to foods recommended by the dietary guidelines - fruits, vegetables, whole grains, lean meats or meat alternates, and low-fat dairy products - contributed to the score. The maximum score possible was 23. The other 39 items on the questionnaire were excluded. Portion size information was also excluded.

• Based on the distribution, RFS scores in the cohort were divided into quartiles: quartile 1, mean score 7 (0-8; n=8890); quartile 2 mean score 10.0 (9-11; n=12,070); quartile 3 mean score 12.0 (12-13; n=9088); quartile 4 mean score 15 (14-23; n=12,206). The risk for mortality for the upper three RFS quartiles was compared with the risk for the lowest RFS quartile.
• The median follow-up time was 5.6 years and the mean age at baseline was 61.1 years (40-93 years). The mean (SE) RFS score of the analytic cohort was 11.04(0.02)

Results

Generally, subjects with higher RFS were slightly older, more educated, physically active, likely to drink alcohol, use supplements regularly, and less likely to smoke. More than 87% or the cohort was white and had 12 or more years of education. Risk for Mortality
There were 2062 deaths due to all causes. The RFS was inversely associated with all-cause mortality.

• The relative risk for all-cause mortality for the upper three RFS quartiles compared to the lowest quartile was 0.82 (0.73, Confidence Interval [CI] 95%) for quartile 2; 0.71 (0.62-0.81, CI 95%) for quartile 3; and 0.69 (0.61-0.78, 95% CI) for quartile 4. Relative risk ratios were adjusted for education, ethnicity, age, body mass index, smoking status, alcohol use, level of physical activity, menopausal hormone use, and history of disease. (Cox regression model for trend; P<.001). These results were unaffected when 223 deaths not confirmed by death certificates were excluded.
• Respondents in the highest quartile of RFS also had at least 30% lower risk than those in the bottom quartile for all-sites cancer, stroke, and heart disease.
• The RFS-mortality association remained significant when data were analyzed excluding subjects who reported a history of cancer, diabetes or heart disease. It also remained significant when the RFS-mortality was reexamined to exclude the possibility of those reporting a poor diet at baseline due to pre-clinical disease (P<.001).
• For all-sites cancer (P<.001), coronary heart disease (P<.001), and stroke (P=.001), respondents in the highest quartile of RFS had at least a 30% lower risk than those in the lowest quartile.

Daily Mean Intake of Energy and Selected Nutrients

• RFS was positively associated with energy and protein intake, percentage of energy from carbohydrate and micronutrient intake. Percentage of calories from fat was inversely associated (mean 39%, 36%, 34%,32% for quartiles 1 - 4 respectively; P<.05).
• The results also suggest a qualitative difference in food selection associated with higher RFS scores. Mean energy intake of quartile 4 was 131% of quartile 1, however mean levels of dietary fiber, vitamin C, folate, and pro-vitamin A carotenoids in quartile 4 were 200%, 230%, 181% and 253% respectively of the mean levels in quartile 1.

Comment

The results of this study support the food-based guidelines, and suggest that complying with current dietary recommendations is associated with improved health outcomes. "Dietary patterns characterized by consumption of fruits, vegetables, whole grains, low-fat diary products, lean meats and poultry is associated with a lower risk of mortality." Women in the highest RS quartile had a 30% lower risk of multivaritate-adjusted, all-cause mortality compared to those with the lowest RFS scores. The authors suggest that "increasing the intake of recommended foods without undue emphasis on learning about hidden, fat, total amount and type of fiber or individual vitamins and minerals, may represent a practical recommendation for improving health."