Association between rest-activity rhythm and diabetic retinopathy among US middle-age and older diabetic adults

Diabetic retinopathy (DR), is the leading cause of vision impairment and blindness among work-age adults, characterized by the progressive retina vascular damage as a severe complication of diabetes mellitus. Paralleling the rise of diabetes, the prevalence of DR had been increasing globally. According to a recent meta-analysis, the global prevalence of DR among the diabetic population is 22.27%, and the number of DR is projected to reach 160.50 million by 2045 (1). DR manifests through series of stages, from microaneurysms abnormalities to proliferative DR marked by neovascularization and retinal detachment (2). Understanding the risk factors and pathophysiological mechanisms of DR is crucial for developing effective prevention and treatment strategies.

Circadian rhythm is an intrinsic 24-hour cycle regulating various biological processes in organisms. One critical output of circadian rhythm is the rest-activity rhythm (RAR), which encompasses the patterns of physical activity and rest throughout the day and night (2). Disturbance of RAR, often resulting from irregular sleep patterns, shift work, or exposure to artificial light (3–5), has been implicated in various metabolic diseases and cardiovascular disorders, such as hypertension, diabetes, and so on (2, 6, 7).

Emerging evidence indicates that the disruption of circadian rhythm may also influence the diverse facets of DR, from alterations in clock genes to changes in rhythmic clinical manifestations (8, 9). Notably, a remarkable decrease of a diverse array of clock genes were observed in retina of diabetes rodent model (10, 11). Our recent study revealed that diabetes reshapes the rhythmic profile of retinal transcriptome (12). Aberrant expression or loss of core clock genes could accelerate retinal endothelial dysfunction and increase retinal permeability (13). Furthermore, disturbed rhythms in retinal structures and physiological functions, such as retinal thickness and melatonin rhythm, have been observed in DR (14). A recent cross-sectional study explored alteration of circadian rhythm in DR by examining 24-hour melatonin secretion, intrinsically photosensitive retinal ganglion cell (ipRGC) and RAR (15). The results indicated that DR patients exhibited lower melatonin output and reduced ipRGC function. However, due to the small sample size (n=25), no significant change in RAR was observed. In the present study, we aimed to explore the association between RAR metrics and DR based on a large, population-based cohort, hypothesizing that irregular RAR may be a potential risk factor for the development and progression of DR. This will be investigated through a cross-sectional study among middle-age or older diabetic participants (age≥40) within the U.S. National Health and Nutrition Examination Survey (NHANES).

The present study uncovered the association between DR and RAR parameters in a representative U.S. working-age or older diabetes population. Notably, our findings demonstrated that the risk of DR increases with the higher IV value and lower M10 value. This indicates that fragmented rhythms and lower peak activity are more common among DR participants. These findings highlight the potential advantages of maintaining a regular sleep-activity rhythm and having high activity levels during the daytime in terms of delaying the onset of DR.

It is well documented that circadian rhythm plays a crucial role in regulating glucose homeostasis (20). Previous studies have revealed that the disruptions of circadian rhythm, induced by lifestyle factors such as shift work (21) and social jetlag (22), are associated with type 2 diabetes. Glucose homeostasis is essential for maintaining the normal physiological function of retinal cells. Hyperglycemia triggers a cascade of complex pathophysiological changes in retinal cells (23). For example, chronic hyperglycemia can cause endothelial dysfunction, pericyte loss, and the perturbations of endothelial glycocalyx in the retina (24, 25). In addition, hyperglycemia activates retinal microglia and Müller cell, inducing the release of various inflammatory cytokines and vascular growth factors (26, 27). The production of reactive oxygen species and the formation of advanced glycation end products also increase under poorly controlled blood glucose. These processes exacerbate retinal vascular damage, promote neovascularization, and ultimately lead to the progression of DR (28).

Additionally, clinical findings and laboratory studies have indicated that circadian rhythm disturbances are directly associated with DR (29). Melatonin, a hormone acting as a dark signal for the retina, was found to be decreased in DR patients (15, 30). Furthermore, DR patients suffer from ipRGC dysfunction (15), which is a crucial component of the light entrainment in synchronization of the central clock to environment cues (31). Our previous study found that circadian oscillations of the transcriptomic profile in the retina were reshaped in streptozocin-induced diabetic mouse model (12). The expression of clock genes was altered in retina of diabetic mice (32), and the loss of core clock gene was related to the phenotype of DR. For instance, Per2 mutant mice showed DR-like retinal vascular, including increased retinal vascular permeability and avascular capillaries (13). Interestingly, deletion of another core clock gene Bmal1, reduced neovascularization and vascular leakage, indicating Bmal1 could drive neovascularization (33). In a summary, circadian rhythm disruption and altered expression of clock genes play crucial roles in the development of DR. The development of wearable device Actigraphy enables individuals to conduct 24-hour sleep and activity monitoring. The introduction of the RAR, generated from Actigraphy data, provides a robust quantification of daily rhythm. Recently, researchers have uncovered the association between RAR and various diseases, such as Parkinson’s disease, non-alcoholic fatty liver disease (34, 35). Additionally, it was found that blunted RAR is strongly correlated with biological aging and may negatively affect well-being and longevity (36). However, there are limited studies on relationship between RAR and DR.

Recently, Mayuko et al. examined RAR in relation to common diabetic complications, including DR (37). The findings of our study share similarities with those of Mayuko et al., but also exhibit differences. They stated that the presence of DR was associated with higher IV and lower daytime activity, consistent with our findings. However, Mayuko et al. found that RA and L5 were also significantly associated with DR, which was not observed in our study. The differences could be explained by a few reasons. Firstly, considering the complex survey design, sample weights for each individual was calculated when conducting analysis in our study, whereas Mayuko et al. had a limited sample size. Secondly, Mayuko et al. further graded DR into simple diabetic retinopathy and proliferative diabetic retinopathy, whereas we only considered the presence of DR without considering its severity.

M10, representing the mean activity level during the most active 10-hour period within 24-hour cycle, which is used to analyze activity patterns. In this study, M10 was found to be negatively associated with the risk of DR. It is well known that environmental and lifestyle modifications can alter circadian rhythm (38). Research in animals has shown that db/db mice, a well-acknowledged diabetic mouse model, showed sporadic wheel-running activity with no consistent pattern (39). There is substantial evidence supporting the insight that increasing physical activity alleviates the progression of DR (40). Several prospective cohort studies have reported that higher physical activity is independently associated with a lower incidence of DR (41, 42). The mechanism is speculated to be that physical activity might improve glycemic control and 25-hydroxyvitamin D level (43, 44). It was reported that 25-hydroxyvitamin D could inhibit retinal inflammation and retinal neovascularization in vivo and in vitro (45–47). Thus, we speculate that subjects with high activity pattern (high M10 value) may have better glycemic control and higher level of 25-hydroxyvitamin D, thereby ameliorating DR-related pathogenesis.

Intradaily variability (IV), refers to fluctuation or change in an individual’s activity levels and rest periods within a single day. High IV indicates a less stable or more fragmented rest-activity pattern, with frequent transitions between activity and rest states. This instability generally results from irregular sleep patterns, inconsistent wake times or frequent napping. A case-control study compared sleep quality among diabetic patients with and without DR, and found that poor sleep, sleep latency and daytime dysfunction were more common in DR patients (48). A recent meta-analysis also reported poor sleep satisfaction was associated with a 2-fold risk of having DR (48). Sirimon et al. found DR patients were more likely to have insomnia and greater sleep variability (49). Melatonin, the primary hormone involved in regulating the sleep-wake cycles, is known that melatonin deficiency is responsible for sleep disorders (50). Melatonin exhibits multiple protective effects that could benefit DR patients, including anti-oxidative, anti-inflammatory and anti-apoptotic properties (51). Accordingly, we speculate that some participants, due to a lack of melatonin, may lose its protective effect on the retina and become more prone to developing DR. Additionally, melatonin-deficient participants are more likely to experience sleep disorders or fragmented rest-activity patterns, as evidenced by elevated IV value in RAR.

The main strength of this study is its nationally representative and diverse population-based cohort, which enhances the generalizability of the findings. However, there are several limitations to consider when interpreting the results. Firstly, due of the cross-sectional study design, causality between RAR metrics and DR cannot be established. Given that advanced stages of DR can cause severe vision loss and dysfunction of ipRGCs, leading to misalignment of circadian photoentrainment, it is possible that blunted RAR is a result of DR. Secondly, the diagnosis of DR was obtained from the self-reports in the medical conditions questionnaire, which could lead to diagnostic omissions. Additionally, in this study, only the presence of DR was available for analysis without considering the severity of DR. Future research should incorporate fundus photo and optical coherence tomography images for more accurate DR diagnosis and grading.

We demonstrated that altered RAR parameters are associated with the onset of DR. Notably, the significant associations were observed for IV, reflecting the fragmentation of rhythm, and for M10, reflecting the daytime activity level. These findings indicate that the fragmented rhythm and lower peak activity may increase the risk of DR. Further interventional research aimed at improving circadian rhythm through lifestyle, behavioral and environmental modification should be conducted to elucidate the casual links between circadian disruption and DR.