Accelerometer-Measured Daily Behaviors That Mediate the Association Between Refractive Status and Depressive Disorders

Abstract

Purpose: To identify the accelerometer-measured daily behaviors that mediate the association of refractive status with depressive disorders and enhance the understanding of behavioral differences in depression.

Methods: Participants with baseline mean spherical equivalent (MSE) and 7-day accelerometer measurements from the UK Biobank were included in this cohort study. Refractive status was categorized as hyperopia and non-hyperopia. Four daily behaviors, including moderate to vigorous intensity physical activity (MVPA), light physical activity (LPA), sedentary, and sleep were recorded between 2013 and 2015. We also assessed 24-hour behavior patterns. Depression cases were defined through both questionnaires and hospital records over 10 years of follow-up.

Results: Among 20,607 individuals, every 0.5-diopter increase in MSE was associated with a 6% higher risk of depressive disorders, with hyperopia participants at a higher risk than non-hyperopia participants (odds ratio, 1.14; 95% confidence interval, 1.05–1.23; P = 0.001). MVPA and sleep time significantly correlated with depressive disorders, with odds ratios of 0.79 and 1.14 (P < 0.05). MSE showed significant correlations with all four behaviors. The effects of MVPA and sleep duration on MSE and depressive disorders varied throughout the day. Mediation analyses showed that MVPA and sleep partially mediated the relationship between MSE and depressive disorders, with 35.2% of the association between moderate to high hyperopia and depression mediated by MVPA.

Conclusions: Physical activity and sleep significantly mediate the relationship between MSE and depressive disorders.

Translational Relevance: The mediation effect of MVPA highlights its therapeutic potential in reducing the risk of depression among individuals with moderate to severe hyperopia. Interventions aimed at increasing daytime MVPA and decreasing daytime sleep could enhance mental health in this vulnerable group.

Introduction

Depression is a global health concern that affects more than 300 million individuals worldwide. It ranks among the leading causes of death and disability, and exerts a significant burden on individuals, caregivers, and health care systems.¹ Vision impairment is known to coincide with depression status,^2,3 and age-related refractive errors represent one of the most common eye conditions leading to visual impairments.^4,5 Our previous study indicated that hyperopia was significantly associated with an increased risk of incident clinically significant depression.⁶ However, community-based studies investigating the association between refractive errors and depression have yielded conflicting evidence,^7,8 leaving the relationship between depression and the full spectrum of mean spherical equivalent (MSE) unclear. 

Daily behaviors influence lifestyle and overall well-being, which play into the signs and symptoms associated with depression.^9,10 It is important to consider that individuals with uncorrected or improperly corrected refractive errors often experience visual discomfort. This discomfort can impede their ability to engage in normal daily activities. For example, although physical activity is known to improve mental health,^11,12 individuals with refractive errors may have their physical activity limited, which can negatively impact their overall physical and mental health. Additionally, there is evidence suggesting that visual disturbances can alter the natural perception of the light–dark cycle, leading to circadian disruptions.¹³ These disruptions are closely associated with sleep disorders and mood dysregulation, including depression. The existing epidemiological evidence regarding the association between sedentary behaviors, refractive status, and the risk of depression remains inconsistent.¹⁴ It should be noted that most studies on the associations between daily behaviors and depressive disorders are based on self-reported questionnaires, and there remains a need for more comprehensive research utilizing objective measures to assess the correlation between refractive status, behavioral patterns, and depression. Compared with traditional questionnaires, accelerometer tests use definite assessment criteria and provide objective measurements, which are less vulnerable to recall bias.¹⁵ Exploring accelerometer-measured daily behaviors, specifically in individuals with refractive errors, can offer new perspectives on the occurrence of depressive disorders and facilitate the development of targeted interventions for their rehabilitation. 

In addition, limited studies addressed whether the effect of refractive status on depressive disorders is mediated indirectly by daily behaviors. Knowledge of the pathway through which refractive status is related to depression may assist in developing more effective interventions to decrease the burden on mental health care for people with refractive errors. Therefore, in the present study, we used the UK Biobank cohort and sought to investigate the associations of refractive status and accelerometer-measured daily behaviors with the risk of depressive disorders and assess the mediation effect of daily behaviors for the association between refractive status and depressive disorders. 

Methods

Study Sample

The UK Biobank is a national, prospective, population-based study established to identify determinants of complex diseases and improve the health outcomes of British adults. Approximately 9.2 million participants aged 40 to 69 years in the UK's National Health Service residing near 1 of 22 assessment centers were invited. More than 500,000 (response rate of 5.5%) individuals volunteered to participate in the baseline assessment between 2006 and 2010. Data were collected at baseline through questionnaires, physical measures, and biological samples, and long-term longitudinal follow-up was conducted for a wide range of health-related outcomes.¹⁶ The UK Biobank was approved by the UK National Health Service National Research Ethics Service (Ref 11/NW/0382) and informed consent was obtained from all participants. Details of the methods, protocols, and definitions used in the study can be found on the UK Biobank website (https://www.ukbiobank.ac.uk/). 

All data analyzed herein were provided by UK Biobank under project reference and data transfer agreement #86091. The flowchart for selection of participants from the UK Biobank in the current analysis is shown in Supplemental Figure S1. 

Ascertainment of Refractive Status

In the UK biobank, refractive error was measured by an RC-5000 device (Tomey) from January 1, 2006, to October 31, 2010. The mean values of spherical power and cylindrical power were determined by averaging 10 repeated measurements for both the left and the right eyes. The MSE was calculated by adding one-half of the cylinder power to the sphere power, and was averaged across both eyes. Hyperopia was defined as an MSE of +1.00 diopters (D) or greater, and moderate to high hyperopia was defined as an MSE of +2.00 D or greater. Myopia (MSE of −1.00 D or less) and emmetropia groups (MSE ranging from −0.50 D to 0.50 D) comprised the non-hyperopia group and served as controls. To minimize measurement error, the minimum difference between refractive groups was set at 0.50 D. 

Individuals with a history of eye conditions that could influence refractive errors including cataracts, injury or trauma resulting in loss of vision, refractive laser eye surgery, and corneal graft surgery were excluded.¹⁷ To rigorously mitigate the potential impact of visual impairment on behavioral and depressive outcomes, individuals with visual impairment (defined as the presenting visual acuity worse than 0.3 logarithm of the minimum angle of resolution units [Snellen 20/40] in the better-seeing eye) were excluded. 

Accelerometer Test in the UK Biobank

A 7-day raw device-based physical activity data were collected using an accelerometer (Axivity AX3; designed by Open Lab, Newcastle University) from 2013 to 2015, with details on accelerometry data processing, analyses, and behavior classification available elsewhere.¹⁵ Information on accelerometer data collection and analysis in this study is summarized in the Supplemental Materials. A validated machine learning model using balanced random forests with Hidden Markov models was applied to identify different movement behaviors in 30-second time windows, including moderate to vigorous intensity physical activity (MVPA), light physical activity (LPA), sedentary behavior, and sleep. Participants with accelerometer data that could not be calibrated, with more than 1% of clips with values exceeding 8 g or falling below −8 g, or exhibited abnormal average acceleration exceeding 100 mg, were excluded from the analyses (Supplemental Fig. S1). 

Ascertainment of Depressive Disorders

Depression was identified through incident hospitalization records, which categorized the disorder based on the 10th edition of the International Classification of Diseases, including patients whose events occurred after the baseline and before the accelerometer test. Lifetime depression disorder, based on questions from the Composite International Diagnostic Interview Short Form questionnaire, or current depression symptoms over the past two weeks measured by the Patient Health Questionnaire-9, conducted from 2016 to 2017, were also used to define depression.¹⁸ We excluded participants with depression at baseline, which was determined using self-reported depression data and scores on the Patient Health Questionnaire-2. Summary information for each depression outcome is provided in Supplemental Table S1. Meeting any of these criteria resulted in an individual being classified as having depression in this study. 

Covariates

Potential confounding factors considered in the present analysis included age, sex, ethnicity, educational attainment, family history of severe depression, smoking status, physical activity level, body mass index, smoking status, drinking status, and comorbidities (hyperlipidemia, diabetes, and cancers). In the analysis related to daily behaviors we further adjusted for work shift history, healthy diet score and season of the year in which the accelerometer was worn.¹⁹ All variables adjusted for are detailed in Supplemental Table S1. 

Statistical Analysis

Baseline characteristics of participants were compared between those with depression and those without using analysis of variance tests for continuous variables and χ² tests for categorical variables. Subsequently, the following models were adjusted for covariates that were found to be associated with depression. 

For the present time follow-up population-based cohort study, logistic regression models evaluated the association between depression and refractive status or daily behaviors. MSE was considered a continuous and categorical variable (hyperopia and non-hyperopia) to investigate the impact of refractive status on depression. We then repeated the analysis without excluding participants who had depression at baseline, which increases the statistical power and robustness of the findings. Considering the interaction between gender and depressive disorders, we also explored the association between refractive status and depressive disorders within gender-stratified groups. 

Restricted cubic splines explored the relationship between continuous MSE and daily behaviors, with 4 knots at its quantiles. To assess whether temporal behavior patterns had effects on refractive status and depressive disorders, time spent in each pattern was subdivided into four categories (morning: 6:00–11:59; afternoon: 12:00–17:59; evening: 18:00–23:59; and late night: 0:00–5:59). 

Given that the refractive status assessment was conducted from 2006 to 2010, daily behaviors were measured from 2013 to 2015, and depressive disorders were evaluated from 2016 to 2017, there is a clear timeline among the three, fulfilling the requirements for a mediation analysis. Mediation analyses examined the involvement of daily behaviors in the pathways linking MSE to depressive disorders. A significant indirect role (mediation) was deemed present when the following conditions were met: (a) a significant relationship existed between the independent variable and the mediator, (b) a significant relationship existed between the independent variable and the dependent variable, (c) a significant relationship existed between the mediator and the dependent variable, and (d) the association between the independent and dependent variables, known as the direct role, was attenuated when the mediator was included in the regression model. Structural equation modeling, the Sobel test, and bootstrap estimation with 1000 bootstrap samples and 95% confidence intervals (CIs)²⁰ were used to test the indirect effect. Models were adjusted for age at baseline, sex, ethnicity, smoking status, education level, Townsend index, family history of severe depression, seasons of accelerometer wear, drinking status, body mass index, health diet score, history of work shift, physical activity level, and comorbidities (diabetes, hypertension, hyperlipidemia, and cancers). 

All statistical analyses were conducted using Stata, version 16.0 (StataCorp LLC, College Station, TX), and R, version 4.3.0 (The R Foundation for Statistical Computing, Vienna, Austria). A two-sided P value of less than 0.05 indicated statistical significance. 

Results

A total of 26,776 UK Biobank participants had available refractive data and accelerometer measurements. After excluding participants with no eligible refractive data (n = 4001), low-quality accelerometer data (n = 1640), and visual impairment (n = 528), a total of 20,607 participants were included in the final analysis. The study population consisted of 55.71% females, with a mean age of 56.52 years. The participants had a mean baseline MSE of −0.56 ± 2.78 D. Table 1 describes the baseline characteristics for included participants stratified by depression status. In general, participants with depression were more likely to have a higher MSE, were younger, female, of non-White ethnicity, less educated, with higher Townsend scores, current or former smokers, current or former drinkers, with a higher body mass index, with a family history of severe depression, and have a history of diabetes, hypertension, hyperlipidemia, and cancers compared with those without depression (all P < 0.05). 

Table 1.

 

Baseline Characteristics of the Study Participants Stratified by Depression Status

 

MSE and Daily Behaviors Associated With Depressive Disorders

After multivariable adjustment, every 0.5 D increase in MSE was associated with a 6% increase in the risk of depression (odds ratio [OR], 1.06; 95% CI, 1.02–1.11; P = 0.007). Participants with hyperopia had a higher risk of depressive disorders compared with non-hyperopic participants (OR, 1.14; 95% CI, 1.05–1.23; P = 0.001). For daily behaviors, MVPA was protective effect against depression risk (OR, 0.01; 95% CI, 0.002–0.06; P < 0.001), and longer sleep durations were associated with an increased risk of depressive disorders (OR, 6.21; 95% CI, 2.80–13.77; P < 0.001). There was no significant association observed between sedentary activity and LPA time with depression (Table 2). We observed similar results in the population that included individuals with baseline depression (Supplemental Table S2). In exploring the association between refractive status and depressive disorders stratified by gender, we did not observe any significant interaction effects, although the stratified analysis indicated a stronger correlation between women and depressive disorders (Supplemental Table S3). 

Table 2.

 

Association of Refractive Status and Daily Behaviors With Depressive Disorder

 

Associations of MSE and Daily Behaviors

In multivariate-adjusted models, the restricted cubic splines suggested significant nonlinear associations between MSE and sedentary time (P for nonlinear < 0.001), LPA time (P for nonlinear = 0.012). Overall associations were noted for MVPA time and MSE (P overall = 0.002), and sleep time and MSE (P overall < 0.001). Specifically, sedentary time exhibited a steep increase within the lower or higher MSE range (<−0.05 D or >−0.05 D). Similarly, LPA time showed a steep decrease within the lower or higher MSE range (<0.08 D or >0.08 D), and a similar pattern was observed for MSE and sleep time. In contrast, MVPA time showed a flattened trend within the lower MSE range (<−0.05 D) and decreased within the higher MSE range (>−0.05 D) (Fig. 1). 

Figure 1.

 

Restrict cubic splines for association between MSE and daily patterns. (A) Association between MSE and average time proportion of sedentary. (B) Association between MSE and average time proportion of LPA. (C) Association between MSE and average time proportion of MVPA. (D) Association between MSE and average time proportion of sleep. The 95% CIs of the adjusted ORs are represented by the shaded areas. Multivariable models were adjusted for age at baseline, sex, ethnicity, smoking status, education level, Townsend index, family history of severe depression, seasons of accelerometer wear, drinking status, body mass index, health diet score, history of work shift, physical activity level, and comorbidities (diabetes, hypertension, hyperlipidemia, and cancers).

Daily Behaviors Among Different Periods in 24 Hours Associated With MSE and Depressive Disorders

We further investigated whether focal periods in a 24-hour cycle for MVPA and sleep time changed associations. Regarding MSE, a significant negative correlation was observed as MVPA time increased in the afternoon, and evening. The strongest negative correlation was found for MVPA in the afternoon (β = −0.18; P = 0.001). However, we only observed a significant correlation between sleep time in the early evening and MSE (β = 0.16; P = 0.002) (Table 3). 

Table 3.

 

Association of Daily Behaviors With MSE and Depressive Disorders Among Different Time Periods in 24 Hours

 

Regarding depressive disorders, increased MVPA time in the morning, afternoon, and evening was protective effect against depressive disorders. Prolonged sleep in the afternoon, and early evening was associated with the development of depressive disorders, particularly in the afternoon (OR, 1.19; P < 0.001). In contrast, nighttime sleep was associated with a 12% decrease in the risk of depressive disorders (OR, 0.88; P = 0.001) (Table 3). 

Daily Behaviors Between MSE and Depressive Disorders

To further identify underlying behavioral pathways between refractive status and depressive disorders, MVPA and sleep time were entered as mediators. 

After multivariable adjustment, MSE was significantly associated with the development of depression (β = 0.005; P < 0.01) in the mediation model. When MVPA and sleep were included in the model these significantly mediated the effect size (β = −0.008 and β = 0.008, with P < 0.001 for both, respectively). All 95% CIs of the indirect effect in the bootstrap estimation were above 0, and the P value for the Sobel test was less than 0.05. Specifically, MVPA time accounted for 9.3% of the mediated association between MSE and depressive disorders, and sleep time accounted for 5.4% (Fig. 2 and Supplemental Table S4). 

Figure 2.

 

Mediation analysis between MSE and depressive disorders, with MVPA level and sleep level as mediators. Note: β represents the indirect effect through the mediator. β′ represents the total effect of MSE on depressive disorders. The MVPA time and sleep time partially mediated the relationship between MSE and depression, with mediation proportions of 9.3% and 5.4%, respectively. The MVPA time partially mediated the relationship between hyperopia and moderate to high hyperopia on depression, with mediation proportions of 5.5% and 35.2%, respectively (all P for indirect effect <0.05). However, the mediation analysis did not show a significant moderating effect of sleep time on the relationship between hyperopia and depression (P for indirect effect >0.05). Multivariable models were adjusted for age at baseline, sex, ethnicity, smoking status, education level, Townsend index, family history of severe depression, seasons of accelerometer wear, drinking status, body mass index, health diet score, history of work shift, physical activity level, and comorbidities (diabetes, hypertension, hyperlipidemia and cancers). ^aP or indirect effect < 0.05 level.

In addition, we explored whether MVPA and sleep time moderated the relationship between hyperopia and depression. Only MVPA significantly mediated the association between hyperopia and depression, aligning with the observed distribution in the histogram of refractive status, daily patterns, and depression disorder status. Participants with depression in the hyperopia group exhibit a more pronounced reduction in MVPA (Supplemental Fig. S2). Specifically, the unstandardized regression coefficient between hyperopia and depression decreased from a β of 0.040 to a β′ of 0.038 when MVPA was considered as a mediator, with a mediating ratio of 5.5% (the 95% CIs of indirect effect above 0 in the bootstrap estimation and the P value for the Sobel test was < 0.05). More important, when considering MVPA time as a mediator, the association between moderate to high hyperopia and depression demonstrated a complete mediating effect (P for total association > 0.05; P for indirect association < 0.05), with a mediating ratio of 35.2% (Fig. 2 and Supplemental Table S4). 

Discussion

This study is the first, to our knowledge, to examine the behavioral mechanisms underlying the association of MSE and depression, a topic of great interest, with a large sample (n = 20,607). We observed a positive association between higher MSE and an increased risk of depressive disorders. The results showed that both lower MVPA and longer sleep duration are correlated significantly and positively with depression. Moreover, all four daily behaviors showed a significant association with MSE. Furthermore, mediation analysis revealed that both MVPA and sleep duration played a significant role in mediating the association between MSE and depression. Particularly, MVPA showed a significant moderating effect on the relationship between moderate to high hyperopia and depression. These findings provide a behavioral basis for understanding how refractive status is associated with poor mental health, and this result in turn has implications for treatment because of the behavioral features identified. 

Although previous studies reported rates of depression are higher in older people with visual impairment,^21–24 the isolated effect of refractive errors on depression in middle-aged and older populations has not been fully explored until now. Our study demonstrated a dose–response relationship between continuous MSE and depressive disorders, providing evidence that the risk of depression escalates with higher levels of MSE. Considering rates of presbyopia in the aging population²⁵ and the limited efficacy of pharmaceutical treatments for depression, developing preventative lifestyle strategies to prevent the onset of depression in this population is of high priority. 

To date, there has been little work using accelerometer data to examine diurnal behavior patterns among older adults and their interaction between depression and refractive status. We observed that individuals with refractive errors tended to have longer sedentary times, shorter LPA times, and shorter sleep times, which conferred a greater depression risk. Regarding MVPA time, a more pronounced decrease was observed among individuals with hyperopia. Previous studies indicate similar patterns between refractive errors and activity in middle-aged and elderly populations, who typically exhibit lower levels of physical activity²⁶ and experience sleep disturbances.²⁷ The novelty of this study's methodology highlights the significance of considering daily behaviors as potential factors influencing refractive errors and manipulating these factors to negate the future risks of depression. For example, if people with uncorrected refractive errors were to receive single vision spectacles, it is possible that this strategy would allow for normal activity levels and sleep, thereby eliminating their risk of depression from vision impairment. 

Previous studies revealed that higher levels of physical activity were associated with a 17% to 21% lower risk of developing depression.^11,28 Our study indicates that MVPA, rather than LPA, is associated with a lower risk of depression. Several studies found that participants with impairments in visual acuity had slow gait speed, mobility limitations, and a greater risk for falls.^29,30 Improving refractive errors may enhance physical activity by providing clearer vision, thereby increasing confidence and safety during physical activities for individuals aged 55 years or older.³¹ This increase in physical activity can then contribute to health benefits, such as improved vascular function and oxygenation and regulation of the hypothalamic–pituitary–adrenal axis, as well as noradrenergic and serotonergic effects.^32–34 Additionally, increased physical activity can lead to the production of neurotrophic factors, which are known to alleviate negative emotions and mood symptoms in individuals experiencing emotional distress.^35,36 This sequence of improvements demonstrates how correcting refractive errors can be a key step in a chain of positive health outcomes, including the mitigation of depressive symptoms. Shorter sleep duration has been associated with an increased risk of depression.^37,38 This finding could be attributed to the fact that insufficient sleep may lead to increased daytime tiredness, including feelings of sleepiness and psychological fatigue,³⁹ which have been predictive of adverse outcomes in individuals with depression.⁴⁰ It is worth noting that previous research has shown that the relationships between sleep and variables such as physical activity, age, depression, and cancer can be more complex,⁴¹ and these complex interactions may not have been expressed fully in the current study. Future research should use more refined statistical methods and consider using more precise sleep measurement tools to help deepen our understanding of the causal relationships between sleep, refractive status, and mental health. There is also evidence that sedentary behavior independently increases the risk of depression and anxiety disorders in adults, but findings have been inconsistent.^14,42,43 In the present study, we did not observe a significant association between sedentary time and depression; however, improving vision to facilitate normal activity levels would also negate this risk if it were observed in future studies. 

Subsequent subgroup analyses revealed that the temporality of activity was associated with refractive status. Notably, individuals with better visual acuity tended to engage in MVPA during the early evening hours. For depressive disorders, the protective effect of MVPA was evident across different time periods, with the strongest protective effect observed in the afternoon, whereas individuals who reported longer afternoon napping had a higher risk of depressive disorders. It was noted previously that physically active daytime patterns were associated with greater white matter microstructure,⁴⁴ which supports cognition and mental health.⁴⁵ Extended sleep durations throughout the day may disrupt nighttime sleep and circadian rhythms.^46,47 This finding is in keeping with previous studies, which observed shorter nighttime sleep durations in hyperopic older populations, and this factor presumably impacts sleep quality⁴⁸ and such timed behaviors interestingly linked refractive errors and psychiatric disorders.^49,50 These findings provide valuable insights into the design of targeted routines for people with depression, and centers lifestyle behaviors and activity patterns throughout the day as an evidence-based way of improving depression. 

Moreover, the further mediation analysis indicated a notable protective effect of high MVPA time against the occurrence of depression, especially in individuals with moderate to high hyperopia. This finding suggests that individuals with hyperopia are more likely to benefit from MVPA interventions, leading to a lower risk of depression. A meta-analysis of adults indicated that larger effects were found for interventions in major depression disorder, using aerobic exercise, at moderate and vigorous intensities.⁵¹ We hypothesize that this effect may be because, on the one hand, individuals with hyperopia are older, and in the elderly population, MVPA is actually of greater relative intensity owing to the deconditioning that occurs during aging. Therefore, their activity is relatively more likely to reach the level of MVPA, which is also an intervention for depression.⁵² In contrast, hyperopia is often accompanied by more severe visual symptoms, such as blurred vision, asthenopia, and bifrontal headaches exacerbated by near work,^53,54 complicating the feasibility of MVPA in hyperopic elderly people. Moreover, excessive daytime sleep resulted in a relative decrease in MVPA time and decreased nighttime sleep, thereby leading to an increased risk of depression. 

Study Strengths and Limitations

The strengths of this study include its large sample size, use of a prospective cohort, determination of activity via an objective measure, comprehensive outcome (including three types of depression), and adjustment for various potential confounders. Despite this, some limitations should be acknowledged. First, the cross-sectional nature of these data limits our ability to establish causal relationships. Longitudinal studies and objective measurements of refractive errors and daily behaviors would provide further insights into the temporal relationship between MSE, daily patterns, and depressive disorders. Second, device-based methods are limited in horizontal locomotion and are unable to distinguish between types of physical activity and sedentary behavior. Future research combining accelerometer-measured and self-reported methods is needed to verify our findings. Finally, the UK Biobank is representative of a largely White, high-income country, which limits these findings from being generalized to multiethnic populations and countries of dissimilar demographic status. In consideration, it would be ideal for these findings to be examined across ethnicities and geographic regions with marked differences. 

Conclusions

This study suggests older adults with higher MSE, especially hyperopia, are associated with lower levels of MVPA, longer sleep durations, and a higher risk of depression. MVPA and sleep duration played significant roles in the relationship between refractive errors and depression, and understandably may be limited by vision disturbances. Future interventions to restore normal vision while also targeting MVPA and normal sleeping patterns may cater more holistically toward an aging population to decrease the risk and burden of depression. 

Acknowledgments

The authors grateful to the UK Biobank participants. This project corresponds to UK Biobank application ID#86091. Data from the UK Biobank dataset are available at https://biobank.ndph.ox.ac.uk/ by application. 

Funded by the National Natural Science Foundation of China (82171075, 82301260, 82271125), the Medical Scientific Research Foundation of Guangdong Province, China (A2021378), the Science and Technology Program of Guangzhou, China (20220610092, 202103000045), the Outstanding Young Talent Trainee Program of Guangdong Provincial People's Hospital (KJ012019087), the launch fund of Guangdong Provincial People's Hospital for NSFC (8217040546, 8220040257, 8217040449, 8227040339), Personalized Medical Incubator Project, The fund for Precision Medicine Research and Industry Development in SIMQ (2023–31), Guangdong Basic and Applied Basic Research Foundation (2023B1515120028). The funders had no role in the study design, data collection, data analysis, data interpretation, or report writing. 

Authors' Contributions: Study concept and design: Du ZJ, Zhang XY, Yu HH. Acquisition, analysis, or interpretation: Du ZJ, Wang S, Bulloch G, Zhang F, Wang YX, Lai CR, Zhuo ZY, Huang Y, Shang XW, Fang Y, Zhu ZT, Hu YJ. Drafting of the manuscript: Du ZJ, Zhang XY, Bulloch G. Critical revision of the manuscript for important intellectual content: Zhu ZT, Hu YJ, Yu HH. Statistical Analysis: Du ZJ, Zhang XY, Shang XW. Obtained Funding: Yu HH. Administrative, technical, or material support: Zhang XY, Yu HH. Study Supervision: Zhang XY, Yu HH. 

Patient Involvement: No patients were involved in setting the research question or the outcome measures, nor were they involved in developing plans for design or implementation of the study. No patients were asked to advise on interpretation or writing up of results. There are no plans to disseminate the results of the research to study participants or the relevant patient community. 

Disclosure: Z. Du, None; S. Wang, None; G. Bulloch, None; F. Zhang, None; Y. Wang, None; C. Lai, None; Z. Zhuo, None; Y. Huang, None; X. Shang, None; Y. Fang, None; Z. Zhu, None; Y. Hu, None; X. Zhang, None; H. Yu, None