The study objectives were to measure dietary diversity (DD) of an urban US population and to determine if associations of 10 year atherosclerotic cardiovascular (ASCVD) risk with DD were independent of dietary quality. Participants were drawn from the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study, wave 4 (n = 2066, 1259 African Americans (AA), 807 Whites (W). Three DD measures were derived from 2 days of 24 h recall data collected with the USDA automated-multiple-pass-method. Count was based on consumption of at least half an equivalent of food from 21 subgroups. Evenness was calculated using Berry Index (BI) and BI-adjusted by food health values. Dissimilarity was calculated by Mahalanobis Distance. Diet quality was assessed by Mean Nutrient Adequacy (MAR) and DASH scores. Associations of DD and quality with ASCVD risk, calculated using 2013 American College of Cardiology and American Heart Association guidelines, were assessed with multivariable regression. Covariates included income, education, food security, and energy/kg weight. Count and MAR were positively associated whereas dissimilarity was negatively associated with ASCVD risk. There was no evidence that evenness contributed to cardiovascular health. The findings suggest more diversity in food attributes and diets rich in micronutrients rather than increased count support cardiovascular health.

Little is known about the effects of diet quality through adulthood and its association with malnutrition later in life. The first research objective was to evaluate diet quality assessed by Mean Adequacy Ratio (MAR) of United States African American and White adults (n = 2066), examined at baseline and two follow-up waves in the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study. The sample was split into cohorts by age at study baseline: Younger, <50, and older, ≥50 years. The second objective was to assess the association of MAR and risk for malnutrition in adults who were ≥60 years at wave 4 (n = 746). The Mini Nutritional Assessment was used to determine risk for malnutrition. At each of the three study waves, 17 micronutrients from two 24 h dietary recalls were used to calculate MAR. Over 13 years MAR changed minimally in the younger cohort as they aged from early to middle adulthood. In contrast, a statistically significant decline in MAR was observed for the older cohort between baseline (2004–2009) and wave 4 (2013–2017), with a greater degree of worsening at low energy levels. The risk for malnutrition was significantly associated with consuming a diet low in energy, lower protein as a percent of energy at baseline, as well as being food insecure, a current smoker, and having income <125% poverty. The risk for malnutrition was not associated with a change in protein intake in years prior to age 60, change in MAR scores across waves, MAR at wave 4, age, sex, race, or having hypertension or diabetes. These longitudinal study findings revealed that diet quality was not predictive of risk for malnutrition.

Over time, adherence to healthy behaviors may improve physical and mental strength which is essential for successful aging. A plausible mechanism is the reduction of inflammation. Research on the association of risky health behaviors on change in strength with age is limited. This study examined changes in the inflammatory potential of the diet, smoking, illicit drug use with changes in strength in a racially and socioeconomically diverse adult sample from the Healthy Aging in Neighborhoods of Diversity Across the Life Span study. The dietary inflammatory index (DII) was calculated from 35 food components derived from multiple 24-h dietary recalls. Strength was evaluated by handgrip strength (HGS), SF-12 PCS and SF-12 MCS (physical and mental component scores). Repeated measures analyses were used to examine associations. At baseline, mean age was 48.4 ± 0.25 years, 56% of the sample were women, and 58% African American. Significant 4-way interactions were found between age, race, socioeconomic status, and DII for women, on change in HGS (p < 0.05) and in SF-12 PCS (p < 0.05) and for men, in change in SF-12 PCS (p < 0.05). Improvements in SF-12 MCS were associated with all three health behaviors as main effects. This study provided evidence that changes towards improving healthy behaviors, diet with anti-inflammatory potential, not smoking cigarettes and not using illicit drugs, were associated with improved strength. Health professionals, especially registered dietitians and health coaches, should create lifestyle interventions to reduce inflammation targeting change in more than one risky health behavior.

Knowledge of various aspects of dietary diversity (DD)—an essential healthful dietary component—across adulthood is limited. This study examined three DD aspects over time in racially diverse adults. Participants were from the National Institute on Aging, Healthy Aging in Neighborhoods of Diversity across the Life Span study. DD measures were calculated at baseline (N = 2177), and first and second examination follow-ups (N = 2140 and N = 2066, respectively) using two 24-h recalls. The count was based on the consumption of ≥50% of an equivalent from 21 food groups. Evenness was derived using the Berry-Index adjusted by the food’s health value; dissimilarity, by Mahalanobis Distance. Mixed-effects linear regression models were conducted to test changes in DD across adulthood, adjusting for sex, race, poverty status and education as fixed effects, and adjusting for smoking, age and energy as time-dependent variables. Only dissimilarity showed significant interactions of time × race (p = 0.0005), and time × poverty status (p = 0.0325), indicating a slower rate of increase over time in dissimilarity scores among Whites compared with African–Americans and those with income >125% poverty versus <125% poverty. A significant interaction between time×energy (p < 0.0001) was noted for both evenness and dissimilarity scores. To our knowledge, this is the first study to document the differential change in dissimilarity scores by race and income over time.

The authors wish to make a correction to Table 2 in the published version of their paper