Revisiting infradian rhythms in depressive symptoms: wearable signals identify individualised mood cycles in healthy adults

Infradian rhythms of human heart rate reveal distinct multiday cycle periods in healthy adults

Summary Background Chronobiology research has historically focused on circadian rhythms; however, longer infradian rhythms are prevalent in human physiology and may have important implications for health and wellbeing. Previous studies have identified widespread infradian rhythms across human physiology, often in the context of hormonal regulation and disease. Despite growing evidence of their ubiquity, the mechanisms, significance, and clinical relevance of these rhythms remain poorly understood, largely reflecting a lack of longitudinal datasets and robust detection methods. The emergence of new wearable technologies enables rich, continuous data capture within individuals, allowing physiological rhythms to be studied at scale. Methods This retrospective, observational cohort study analysed young adults (n = 623), with up to four years of wearable and questionnaire data collected through the University of Notre Dame’s (USA) NetHealth project. Participants who recorded at least three months of continuous (>80% data completeness) heart rate data were included and significant infradian rhythms were identified using wavelet analysis. Unsupervised non-negative matrix factorisation was performed to cluster similar wavelet power spectrum distributions. Individuals’ heart rate rhythms were compared to known environmental cycles (day-of-week, lunar, seasonal) and considering demographics and social networks. A second, smaller cohort (n = 70) with heart rate and menstrual timing was included to analyse the interplay of hormonal regulation on monthly cycles. Multinomial logistic regression and the Kruskal–Wallis test were applied to quantify the effects of environmental, behavioural and demographic factors on heart rate rhythms. Findings Significant infradian rhythms of heart rate were detected in 70% (369/525) of the cohort and 36% (187/525) had two or more rhythms. Annual, biannual and 10-week rhythms were the most common. Within the 4–45-day band, individuals clustered into three multiday chronotypes based on dominant periodicities in their wavelet power spectra: weekly (∼7 days), monthly (∼25–35 days), and multi-month (>35 days). Heart rate rhythms were influenced by environmental cycles (day-of-week and seasonality) and were synchronised to the menstrual cycle in most menstruating females, although monthly rhythms were also observed in males and menopausal women. Interpretation The prevalence of infradian, or multiday heart rate rhythms in healthy young people motivates further scientific investigation to understand the mechanisms of these rhythms and their potential association with autonomic function, and risk of disease or disease-specific symptoms. Funding This study was funded by an 10.13039/501100000923Australian Research Council Grant (DP240102899). This study also received funding from the Australian National Health and Medical Research Council Medical Research Future Fund (MRFF 2039966).

Multiday rhythms shape mood dynamics in depression

Circadian rhythm disruption and multiple sclerosis: The role of affective symptoms and temperamental traits.

Systematic assessment of mood and circadian rhythms may offer a more comprehensive framework for identifying clinically relevant chronobiological disruption in patients with multiple sclerosis.

Parsing inter-individual variability in the digital phenotype across the menstrual cycle

Characterizing Nonlinear Heterogeneous Sleep Trajectories across the Menstrual Cycle in a Transdiagnostic Psychiatric Population.

Autonomic and limbic system functioning, early life stress, and depressive symptoms in adolescence

Background Adolescence is a critical window for autonomic and limbic development and a period of elevated risk for depression. Although autonomic and limbic system function have, independently, been linked to depression, they are coupled and both influenced by early life stress (ELS). We do not yet know, however, whether the association of autonomic activity and limbic connectivity with depressive symptoms differs across dimensions of ELS. Methods Participants were 93 youth (Mage = 12.3) who completed the Trier Social Stress Test (TSST) with continuous electrocardiogram (ECG) recording. ECG-derived high-frequency heart rate variability (RespHRV) was assessed across the baseline, reactivity, and recovery phases of the TSST. Hippocampal-amygdala connectivity was estimated from resting-state data acquired just before the TSST using the Tian atlas. We tested whether RespHRV recovery and connectivity were jointly associated with depressive symptoms and whether ELS dimensions of threat, unpredictability, and deprivation moderated this association. Results Greater RespHRV during recovery was associated with more negative right hippocampal tail-left lateral amygdala connectivity (β = −0.42, q = 0.002). In a joint model, threat (β = −0.34, p = .007) and deprivation (β = −0.36, p = .009 independently moderated this association. Specifically, more negative connectivity predicted fewer symptoms when paired with weaker RespHRV recovery in high-ELS youth, but with stronger RespHRV recovery in low-ELS youth. Conclusions Whether autonomic-limbic coupling buffers or confers risk for depression depends on the early environment in which it developed.