Heart rate and its variability as an indicator of mental health in male prisoners

Heart rate (HR) and its variability (HRV) has been proposed as a marker for depressive symptoms and other aspects of mental health. However, the real correlation between them is presently uncertain, as previous studies have generally been conducted on the basis of small samples. In a sample of 113 adult male prisoners, we analyzed correlations between five measures of HR/HRV and five psychological measures of mental health aspects (depression, state and trait anxiety, and social relationships). We used Nadaraya-Watson non-parametric regression in both directions and age-stratified Spearman correlation to detect possible relations. Despite strong correlations among HR/HRV measures and among psychological measures, correlations between HR/HRV and psychological measures were low and non-significant for the overall sample. However, we found an age dependency, suggesting some correlations in younger people (HR with STAI-State, r = 0.39; with HADS-Anxiety, r = 0.52; both p < .005). Overall, the general utility of HR/HRV as a marker for mental health across populations remains unclear. Future research should address age and other potential confounders more consistently.

💡 Research Summary

This paper investigates whether resting heart rate (HR) and heart‑rate variability (HRV) can serve as physiological markers of mental health in a relatively large sample of male prisoners (N = 113). The authors measured five HR/HRV parameters (mean HR, RMSSD, SDNN, LF, HF, and the LF/HF ratio) from a 5‑minute electrocardiogram recorded under quiet, seated conditions. Psychological status was assessed with five validated scales covering depressive symptoms (HADS‑Depression), anxiety (HADS‑Anxiety, STAI‑State, STAI‑Trait), and social relationships (WHO‑QOL‑BREF social domain).

Statistical analysis proceeded in two stages. First, the authors applied Nadaraya‑Watson kernel regression in both directions (HR/HRV → psychology and psychology → HR/HRV) to explore potential non‑linear associations without imposing a parametric model. Second, they stratified the sample by age (≤30 years vs. >30 years) and computed Spearman rank correlations within each stratum, thereby testing whether age moderated any relationships.

Across the entire cohort, correlations between any HR/HRV metric and any psychological measure were weak (|r| < 0.20) and non‑significant (p > 0.05). This finding contradicts earlier, smaller studies that reported moderate to strong links between autonomic indices and depressive or anxious symptomatology. However, when the sample was divided by age, a different picture emerged. In the younger subgroup (approximately 38 participants, age ≤30), mean HR showed a moderate positive correlation with state anxiety (STAI‑State, r = 0.39, p < 0.005) and with HADS‑Anxiety (r = 0.52, p < 0.005). HRV indices, by contrast, remained largely unrelated to the psychological scales in both age groups, and the LF/HF ratio did not display the expected pattern of sympathetic‑parasympathetic balance.

The authors interpret these results as evidence that HR may reflect acute stress or arousal in younger inmates, whereas HRV—particularly when measured over a brief five‑minute window—fails to capture the more chronic autonomic dysregulation associated with mood and anxiety disorders in this population. They acknowledge several methodological constraints: (1) the cross‑sectional design precludes causal inference; (2) potential confounders such as smoking, alcohol use, medication, sleep quality, and underlying cardiovascular disease were not fully controlled; (3) the short recording duration limits the reliability of low‑frequency components and may underestimate true variability; and (4) multiple testing corrections were not explicitly reported, raising the possibility of type‑I error.

In the discussion, the paper emphasizes that the prison environment introduces unique stressors (e.g., limited physical activity, altered circadian rhythms, heightened vigilance) that may interact with age‑related physiological resilience. Consequently, the generalizability of HR/HRV as universal mental‑health biomarkers is called into question. The authors propose that future investigations should adopt longitudinal designs with repeated HR/HRV assessments, incorporate comprehensive lifestyle and health covariates, and compare incarcerated cohorts with community samples to disentangle environment‑specific effects.

In conclusion, the study finds no robust evidence that resting HR or HRV can serve as blanket indicators of depression, anxiety, or social functioning among adult male prisoners. Nonetheless, the age‑specific association between HR and state anxiety suggests that autonomic markers may retain some predictive value in younger, perhaps more physiologically responsive, subpopulations. To realize the clinical potential of HR/HRV in correctional settings, researchers must systematically address age, comorbid health conditions, and methodological rigor in future work.