This study aimed to examine the associations between cognition, skin tone, and stress in the form of allostatic load, perceived stress, adverse life events, and perceived discrimination in a heterogeneous sample of Puerto Rican adults.

Participants included 1,502 Puerto Ricans (71% female; M age = 57.06 ± 7.60, range 45-75; M years of education=9.32 ± 2.78) from the first wave of the Boston Puerto Rican Health Study (BPRHS). Participants completed a battery of measures to assess cognition, socioeconomic status, acculturation, and psychosocial status. Skin tone (i.e., dark, medium, light, and white) was ascribed by trained interviewers. Structural equation modeling was used to analyze the associations between skin tone groups, stressful life events (i.e., perceived discrimination and number of adverse life events), perceived stress, allostatic load, and cognition.

The model fit examining the relationship between stress and cognition demonstrated adequate fit even after adjusting for age, sex, acculturation, education, and socioeconomic status (χ2 = 1079.98 (p<0.01), CFI = 0.92, TLI = 0.89, RMSEA = 0.06, SRMR= 0.05). When considering skin tone, model fit was also acceptable for all skin tone groups. Greater allostatic load predicted lower levels of cognition in the overall sample (β=-0.07, p<.01, CI: -0.12: -0.02) and in individuals with darker skin tones (β=-0.24, p=<.05, CI: -0.37: -0.10). Higher levels of perceived stress predicted lower levels of cognition in the overall sample (β=-0.16, p<.01, CI: -0.22: -0.09) and in all skin tone groups except in individuals with white skin tones. Stressful events significantly predicted cognition in the overall sample (β=0.12, p<.01, CI: 0.03: 0.21).

Results provide evidence of an interaction in which the relationships between stress and cognition differs across skin tone groups, even within an ethnic group. This suggests a modifying role of intersecting identities (e.g., ethnicity by skin tone) in understanding health outcomes. Additional research with a robust, diverse sample of H/Ls is needed to continue examining the heterogeneity of heritage groups across various health outcomes.