Circadian rhythms, ∼24 h oscillations driven by the body's internal clock, regulate a range of physiological processes (e.g. vascular function, heart rate) and health outcomes (Chan et al. 2026). Chronotype reflects the timing of an individual's sleep–wake cycle and activity patterns and is shaped by genetic, environmental and developmental factors. While chronotype is typically measured based on self-reported questionnaires, determining one's circadian timing poses greater challenges. For example, although dim light melatonin onset (DLMO) is considered a gold standard method for assessing circadian phase, its reliance on controlled laboratory conditions, serial biological sampling and strict behavioural restrictions makes it costly, low-throughput and impractical for large-scale or real-world applications. Other methods include genotyping of clock-gene variants to predict genetic chronotype predisposition, but there is still much work to be done in this area regarding accessibility and accuracy (Malone et al. 2025). Understanding one's chronotype is important because circadian disruptors, such as misaligned light–dark cycles and caloric intake, known zeitgebers (i.e. synchronizers) of the primary circadian clock, are associated with increased chronic disease risk (Chan et al. 2026). Thus, determining a clinically relevant, accessible and consumer-friendly measure of circadian timing is warranted. A recent study in The Journal of Physiology by Chan et al. sought to (1) develop a circadian timing metric called heart rate phase (HRP) derived from heart rate measured by wearable technology, (2) determine associations between HRP and chronotype-associated genetic variants, and (3) determine associations between HRP and chronotype-associated diseases (Chan et al. 2026). Chan et al. used heart rate data collected via Fitbit from the ‘All of Us’ Research Program, an NIH-funded project that includes health data from over 600,000 participants across the USA. To determine HRP, mean participant heart rate was aggregated over 5 min periods for each weekday, resulting in a series of 2016 time points, which were fit to a sine curve with a 24 h period. The HRP can be interpreted as the time at which the sine curve crosses its midpoint on the upswing, between the overnight trough and the daytime peak. Exclusion criteria included a sine curve goodness of fit (R2) less than 0.5, fewer than 30 days of heart rate data and an HRP greater than 3 SD from the mean. Notably, this reduced the cohort from 53,686 participants with adequate Fitbit data to 31,422 participants. To validate HRP as a proxy measure of chronotype, the authors examined whether 110 single-nucleotide variants known to be associated with chronotype predicted HRP using linear regression, adjusting for confounding variables including age, sex and genomic ancestry. Additionally, they assessed correlations among HRP, physical activity and age, variables demonstrating known associations with chronotype. Having validated HRP, they completed a phenome-wide association study using logistic regression to test associations between HRP and 3272 health conditions, collected from electronic health records, which were adjusted for weekly step count, age, sex and ancestry. Finally, to determine whether chronotype acts causally through HRP to influence disease risk, Mendelian randomization was conducted using the morningness-associated genetic variant, rs1144566, as the instrumental variable. The final analytical sample mean age was 53 ± 16 years, and the mean HRP was 9.48 ± 1.57. A subset of 22,653 participants had whole genome sequencing data available, and of those, 15,960 had linked electronic health records. Using linear regression of HRP on circadian-associated genetic variants, the authors reported the most significant associations between HRP and four variants clustered near the RGS16 and RNASEL genes, a locus consistently linked to morningness chronotype in prior genome-wide association studies and meta-analyses, supporting HRP as a metric grounded in circadian biology. Additionally, the authors reported correlations among HRP, physical activity and age. Specifically, later HRP was correlated with decreased weekly step count (r = −0.31) and, separately, earlier HRP was correlated with increased age (r = −0.24). Together, these findings support HRP as both a genetically and behaviourally validated proxy of chronotype. Among the 3272 disease conditions tested, there were associations between HRP and 22 different chronotype-associated conditions. These disease conditions were separated into four different clusters: addiction and mood disorders, sleep disorders, metabolic disorders and pregnancy-related conditions (see figure 3D in Chan et al. for a visual depiction). Broadly, later HRP was associated with addiction and mood, sleep, and metabolic disorders, while earlier HRP was associated with pregnancy-related conditions. For example, later HRP was associated with increased type 2 diabetes (T2DM) risk (odds ratio, OR = 1.09 1.06, 1.13) and elevated HbA1c (OR = 1.07 1.03, 1.11). Given the associations between the morningness genetic variant, rs1144566, earlier HRP and decreased T2DM risk, the authors used Mendelian randomization to test whether the variant may causally protect against T2DM through its effect on HRP. The analysis suggested a causal relationship in models of both T2DM (P = 0.00027) and elevated HbA1c (P = 0.021). The authors provide strong proof of concept data for HRP to serve as a new measure of chronotype and circadian timing that is associated with behavioural health and several measures of chronic disease risk. The Mendelian randomization analysis suggests a causal link between rs1144566 through HRP to T2DM risk, which meaningfully strengthens the clinical implications. While direct comparisons with existing literature are not yet possible given HRP's novelty, prior investigations of chronotype may help contextualize these findings. The authors’ finding of morningness being protective against T2DM is supported by several large epidemiological studies that report eveningness chronotypes are associated with increased T2DM risk (Heikkinen et al. 2025). Though findings are mixed, eveningness chronotype associations with increased risk of T2DM among males have been observed, suggesting possible sex differences in these chronotype–metabolic disease relationships (Heikkinen et al. 2025). Given that the present study used a sample consisting of two-thirds female participants and the known sex differences in circadian rhythms, including that of heart rate, whether HRP is sex-dependent should be investigated. The present study found associations between morningness and pregnancy-related conditions broadly, including pregnancy itself, second-trimester pregnancy, and conditions of labour and delivery. These findings may simply reflect an association between morningness and fertility rather than specific adverse or beneficial pregnancy outcomes. Studies of pregnancy and chronotype are generally mixed; some find that evening chronotype is associated with adverse pregnancy outcomes and chronotype shifts during the perinatal period (Verma et al. 2024). This suggests a further need to study chronotype, HRP and pregnancy-related changes throughout the circadian cycle. There are several interesting future directions to continue building upon these findings. First, the authors note HRP was calculated over 1 week from Sunday to Sunday. While the median participant had 296 days of data, HRP, as it is calculated in this paper, may not reflect the variation in one's schedule throughout the week, such as between work days and free days, as well as variation throughout the year. Sleep timing regularity, such as the onset or midpoint of sleep and its standard deviation, may be a stronger predictor of cardiometabolic health than sleep duration (Windred et al. 2024). Thus, implementing an algorithm that would allow for the dynamic calculation of HRP and HRP-related variables may help to examine differential associations between workday and free day HRP, in addition to HRP regularity (Bowman et al. 2021). Determining whether ‘catch-up’ sleep (i.e. sleeping in) after night work changes how HRP is applied and interpreted may be of interest. Chan and colleagues adjusted for sleep duration but did not exclude sleep disorders. Excluding diagnosed sleep disorders (and ideally untreated obstructive proxies such as snoring history, if available in health records) would clarify whether HRP–disease associations reflect circadian misalignment per se, or are confounded by sleep pathology that independently drives cardiometabolic risk. Prospective analyses linking baseline HRP to incident disease, which may be feasible within ongoing All of Us follow-up or via accelerometer-derived rhythm analogues in the UK Biobank, would establish temporal precedence and reduce concerns about reverse causation. Lastly, and most relevant to physiology laboratories such as ours, deep physiological phenotyping of individuals may identify subclinical mechanisms linking HRP to disease risk before overt pathology develops. For example, applying HRP to emerging adults may be of interest, as this population demonstrates dysregulated sleep patterns and high sleep timing irregularity relative to other age groups and may represent an ideal time for behavioural intervention before the onset of disease. Such studies should incorporate gold-standard measures of endothelial function (e.g. flow-mediated dilatation), insulin sensitivity (glucose tolerance-derived indices) and meal-timing patterns to inform the design of circadian-targeted interventions (e.g. light and meal timing, exercise) for primordial prevention. In summary, the authors present a novel, continuous measure of chronotype, which was validated against known chronotype-associated genetic variants. Earlier HRP was associated with several morningness chronotype genetic variants, and delayed HRP was associated with T2DM, which is supported by previous findings relating chronotype and chronic disease. The authors offer a strong proof of concept for HRP to serve as a new measure of chronotype and circadian timing. Given that many wearable technology devices offer sleep monitoring metrics, HRP may be included as an accessible measure for consumers to understand their chronotype and the impact of daily activities on their overall health. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared. Conception or design of work: all authors. Drafting the work or revising it critically for important intellectual content: all authors. All authors have approved the final version to be published and agree to be accountable for all aspects of the work. This work was supported in part by NIH LRP L70HL181838 (S.O.S.). We would like to thank Austin T. Robinson and Nathaniel D. M. Jenkins for their feedback on the manuscript.
Muma et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: