(PS10-C68) Circadian Rhythm Preferences Similarly Predict Sleep for Individuals with BPD, GAD, and Healthy Controls

Research suggests that individuals who do not live in line with their bodies’ natural tendency to wake and sleep are at risk for a number of health and mental health difficulties (Phillips, 2009). Individuals with psychopathology experience circadian rhythm disturbances (Eidelman, et al., 2012), but it has been proposed that individuals with borderline personality disorder (BPD) have circadian rhythm disturbances above and beyond those with other mental health difficulties (McGowan & Saunders, 2021). The current sample allows direct comparisons to be made between individuals with BPD, individuals with another mental health disorder (generalized anxiety disorder [GAD]), and healthy controls. The goal of this secondary analysis was to examine whether scores on the morningness-eveningness questionnaire (MEQ; Horne & Ostberg, 1977) correlated with bed times and rise times of individuals over a 14-day span (as would be expected; Horne & Ostberg, 1977), and whether those correlations differed significantly by participants’ diagnoses (e.g. BPD, GAD, or healthy control). Based on the theory that individuals with BPD have more circadian rhythm difficulties, we hypothesized that MEQ would be less predictive of rise times and bedtimes for individuals with BPD as compared to healthy controls. There was no hypothesis about how MEQ would interact differently for BPD versus GAD.

The sample used for this analysis has been described in full elsewhere (Fitzpatrick, Varma, & Kuo, 2020). Individuals with diagnoses of BPD or GAD, and healthy controls (n = 40 in each group) were given daily sleep diaries to complete for 14 consecutive days. On day 7 of their participation, they completed questionnaires including the MEQ. MEQ significantly predicted rise times and bed times (both p < 0.001) when it was the sole predictor considered. We used generalized estimating equations to measure the effect of MEQ, diagnosis, and their interaction on rise times and bed times (both time-varying across 14 days). For both models, we found no significant interaction between MEQ and diagnosis. For the model predicting rise times, there was a significant effect of MEQ (p = .024). For the model predicting bedtimes, the effect of MEQ was not significant, but in the direction we would expect (higher scores, which correspond to more morningness, predict earlier bedtimes; p = .06).

 
These results suggest that individuals with BPD, GAD, and healthy controls do not significantly differ in the degree to which their morningness-eveningness preferences predict their rise times and bedtimes over a two-week period. The predictive power of the MEQ was significant in the entire sample (p < 0.001), suggesting that participants across groups are heeding their individual body clocks. In addition, trend-level results suggest the opposite direction of our hypothesis: MEQ was more predictive of rise times for individuals with BPD (but notably this finding was not significant). This data from a direct comparison between the groups contradicts theories that individuals with BPD specifically experience circadian disturbance.