(P26-31-26) Beef Consumption Patterns Differ Between Mainland and Island-Dwelling Puerto Rican Adults

Objectives: Compare beef consumption patterns between Puerto Rican adults living in Boston, MA and in Puerto Rico.

Methods: Diet from FFQ administered in the prospective cohorts, the Boston Puerto Rican Health Study (BPRHS) at baseline (n=1491), ~2.2 y follow-up (FU, n=1203), ~6.2 y FU (n=811), and ~12.7 y FU (n=550), and the Puerto Rico Observational Study of Psychosocial, Environmental, and Chronic disease Trends (PROSPECT) (baseline, n=1210) was used to assess intake of total beef, lean vs non-lean beef as energy-adjusted servings/day (s/d), and minimally (MP) vs ultra processed (UP) beef as % kcal. ANCOVA tested differences in beef intake between BPRHS (mean of waves) and PROSPECT. Foods were ranked within beef subgroups. K-means clustering identified subgroups of dietary patterns (% total protein). Co-consumption network analysis (CCN) mapped structural relations between foods consumed with beef (% kcal).

Results: In BPRHS and PROSPECT, mean ages were 60 ± 8 and 54 ± 12y; 71% and 77% female. Adjusted mean (95CI) intake of total beef did not differ between cohorts [0.34 (0.32, 0.36) vs. 0.35 (0.33, 0.37), FDR p = 0.36] while> lean beef [0.08 (0.07, 0.08) vs. 0.05 (0.04, 0.05) s/d] and MP beef [81 (80, 82) vs. 73 (72, 75) % kcal] was higher in BPRHS vs. PROSPECT; non-lean [0.23 (0.23, 0.26) vs. 0.28 (0.26, 0.29) s/d] and UP [18 (17, 19) vs. 24 (23, 25) % kcal] were lower (FDR p < 0.001 all). The top 5 beef foods in BPRHS were homemade, while in PROSPECT they were from fast-food. In both cohorts, MP beef included ground beef, flank steak, and stew meat; UP was mainly fast-food burgers. Protein patterns separated into 1) beef and white grains (mostly from non-lean and MP), 2) low-fat milk, and 3) chicken and white grains in BPRHS; and 1) beef and white grains, 2) fish, and 3) eggs in PROSPECT. CCN showed beef consumers co-consume pork and lamb (partial rho = 0.34 BPRHS, 0.35 PROSPECT), the strongest correlations in both networks. Beef was also co-consumed with rice and pasta, eggs, and non-beef processed meat (partial rho = 0.18, 0.15, 0.14, respectively) in BPRHS.

Conclusions: Total beef intake was similar; however, those on mainland consume leaner, MP beef (homemade), while those in Puerto Rico consume more non-lean, UP beef (fast-food burgers). Beef is an important source of protein in both cohorts; leaner and less processed beef should replace current choices in Puerto Rico.

Funding Sources: The Beef Checkoff