Under pressure: Salivary cortisol, psychosocial stress, and type 2 diabetes in five American Indian communities

Document Type



Background: Many American Indian (AI) communities experience significant health inequities, including type 2 diabetes (T2D). Disease onset, progression, and complications of T2D for AIs are worsened by disproportionate exposure to psychosocial stressors. Although much research has examined the effects of stress on T2D complications, there is very little known about how stress affects these processes in AI populations. The current study describes feasibility and compliance to a community-based salivary cortisol collection protocol in 5 tribal communities. We also explore potential confounding factors and correlates of cortisol patterns, the latter of which include psychosocial stressors identified by AI community members. Methods: The Maawaji’ idi-oog Mino-ayaawin (Gathering of Health) study is a community-based participatory research (CBPR) collaboration between the University of Minnesota (UMN) and five Anishiinabe (Ojibwe) communities in the upper Midwestern US. Clinical staff at each participating medical facility generated lists of possible participants using random sampling techniques. Inclusion criteria were a diagnosis of diabetes within 5 years of the sampling data, 18 years of age or older, and self-identified as AI. Data were collected from three sources: 1) Computer-Assisted Personal Interview (CAPI) surveys; 2) medical chart reviews, including lists of relevant prescribed medications, collected by clinical staff; and 3) salivary cortisol levels, self-collected from participants at home. Participants collected salivary samples four times throughout the day to capture their diurnal pattern. At the time of each sample collection, participants also completed the Subjective State Scale (SSS) to assess their environments, behavior, and subjective feelings of stress. This procedure occurred in four waves at 6-month increments. Results: A total of 194 AI participants (109 females) with T2D ranging in age from 18-77 years (M = 46.32) were involved in baseline data collection. Participants adhered closely to salivary cortisol protocol collection requirements. As expected, individual cortisol indices were quite variable between participants, but most demonstrated a healthy diurnal pattern (a rise after waking and gradual decrease throughout the day). Consistent with other research, waking value was the only index that significantly varied by gender, with men (M = 9.42 nmol/L) showing higher values than women (M = 7.26 nmol/L). Waking cortisol values were significantly positively associated with AUCg values (r = .42, p < .001), suggesting that these values are a useful indicator of total daily cortisol output. Smoking was the only self-reported temporal behavior that seemed to influence sample values: smoking was positively associated with waking cortisol values (r = .23, p < .01). Very few chronic stressors were associated with baseline cortisol indices. We expect that chronic exposure to stress will be linked with intra-individual variation over time, rather than inter-individual differences at baseline; preliminary findings for longitudinal data will be presented. The influence of medications on cortisol indices will also be presented. Conclusions/Implications: Findings demonstrate the feasibility of using at-home collection procedures to measure cortisol as a stress biomarker. The findings will also allow us to determine how chronic illness and stressors may influence cortisol activity over time, and examine both intra- and inter-individual differences in responses to stress.

Publication Date


This document is currently not available here.