Mediterranean Diet and Changes in Weight among Older Puerto Rican Adults

Abstract

ObjectivesTo examine the association between a Mediterranean diet score (MDS) and its components with 2‐yr weight change among older Puerto Ricans.MethodsCohort (Boston Puerto Rican Health Study) of 1,127 participants with mean age 57 y (SD: 7), 73% female, and 58% with a BMI ≥ 30 kg/m2. Diet over the previous year was ascertained at baseline using an FFQ modified and validated for use in this population. Adherence to a Mediterranean diet was assessed using a 0‐ to 9‐point scale. Individuals received 1 point for reporting intakes higher than the cohort energy‐adjusted and sex‐specific median serving for either whole grains (0.65 and 0.67 s/d, men and women, respectively), fruits (1.08 s/d for both sexes), vegetables (1.60 and 1.48 s/d), legumes (5.44 and 4.59 s/wk), fish (0.96 and 0.81 s/d), or the monounsaturated to saturated fatty acid ratio (1.16 and 1.18 units/d). Conversely, 1 point was given for estimated consumption lower than the median serving for meat (4.82 and 4.02 s/d, men and women, respectively) or dairy products (1.74 and 1.57 s/d); 1 point was awarded for moderate consumption of alcohol (2 s/d for men and 1 s/d for women). Weight at baseline and at 2‐yr follow‐up (median, inter‐quartile range: 2.01 y, 1.96 to 2.16 y) was measured in duplicate and the average at each visit used in all analyses.ResultsThe average MDS was 4.3 (SD: 1.7). In multivariate models, lower adherence to a Mediterranean diet was associated with greater weight gain during follow‐up [mean (95% CI); 1.51% (0.33, 2.70) for MDS 0–3 and −0.38% (−1.33, 0.57) MDS 6–9; P‐trend=0.001.; Table]. In similar models, vegetable consumption was significantly associated with changes in weight [0.86% (0.09, 1.63) for low and −0.39% (−1.15, 0.38) for high; P=0.004]. Other components of the MDS were not significantly associated with 2‐yr weight change. Our results remained similar after accounting for baseline total energy intake.ConclusionsA dietary pattern that is less consistent with a Mediterranean diet was associated with weight gain over a 2‐yr follow‐up period. Our results suggest that a Mediterranean diet may be beneficial for weight maintenance among older Puerto Ricans.Support or Funding InformationThis study was supported by NIH grants P01‐AG023394, P50‐HL105185, and R01‐AG027087. Relation between baseline Mediterranean Diet Score and 2‐yr changes in weight (n=1127) Model Covariates Mediterranean Diet Score P‐trend 0 to 2 3 4 5 6 to 9 n=164 n=180 n=248 n=239 n=296 1 age, sex, height, time between visits 1.23% (0.10, 2.36) 0.77% (−0.32, 0.63) −0.32% (−1.26, 0.63) −0.71% (−0.32, 1.86) −0.60% (−1.47, 0.27) 0.002 2 + high school education,currently smoking, physical activity score, diabetes, heart disease 1.51% (0.33, 2.70) 1.18% (−0.01, 2.36) 0.02% (−1.00, 1.03) −0.37% (−1.40, 0.66) −0.38% (−1.33, 0.57) 0.001 3 + total energy 1.55% (0.36, 2.74) 1.24% (−0.06, 2.43) 0.04% (−0.98, 1.06) −0.39% (−1.41, 0.64) −0.37% (−1.32, 0.58) 0.001