Higher Daytime Light Exposure Predicts Lower Risk of Gastrointestinal Cancer Incidence and Mortality

Xionge Mei; Wei Wang; Nana Zheng; Tong Luo; Jiaying Xu; Ngan Yin Chan; Jing Wang; Yannis Yan Liang; Sizhi Ai; Yaping Liu; Xiao Tan; Christian Benedict; Yun Kwok Wing; Jihui Zhang; Hongliang Feng

doi:10.1002/ijc.70517

. 2026 May 3;159(6):1401–1410. doi: 10.1002/ijc.70517

Higher Daytime Light Exposure Predicts Lower Risk of Gastrointestinal Cancer Incidence and Mortality

Xionge Mei ^1,², Wei Wang ^1,^2,³, Nana Zheng ^1,², Tong Luo ^1,², Jiaying Xu ^1,², Ngan Yin Chan ^4,⁵, Jing Wang ^1,², Yannis Yan Liang ^1,², Sizhi Ai ^1,², Yaping Liu ^1,², Xiao Tan ^6,⁷, Christian Benedict ⁸, Yun Kwok Wing ^4,⁵, Jihui Zhang ^1,^2,^✉, Hongliang Feng ^1,^2,^3,^✉

PMCID: PMC13397146 PMID: 42071269

ABSTRACT

Daytime light exposure is essential for health, primarily by synchronizing tissue and cellular‐level clocks with the external light/dark cycle. Insufficient exposure may disrupt circadian alignment and contribute to adverse outcomes, including cancer. Given the high prevalence of gastrointestinal cancer, we conducted a prospective cohort study of 89,069 participants with objectively measured daytime light intensity and duration. Cox proportional hazards models were used to evaluate associations with gastrointestinal cancer incidence and mortality. Over a median follow‐up of 8.8 years (804,111 person‐years), 1692 gastrointestinal cancer cases were recorded, of which 891 were fatal. Higher mean levels of daytime light (≥ 1916 lux between 7:30–20:30, based on the 80% cut‐off) were associated with lower risk of gastrointestinal cancer incidence (Hazard Ratio [HR]: 0.87, 95% confidence interval [CI]: 0.76–0.99, p = 0.04) and mortality (HR = 0.76, 95% CI: 0.63–0.93, p = 0.008), particularly for pancreatic cancer incidence (HR = 0.58, 95% CI: 0.40–0.85, p = 0.005) and mortality (HR = 0.47, 95% CI: 0.30–0.73, p = 0.001). Daytime light exposure ≥ 2.4 h (≥ 5000 lux between 7:30 and 20:30, which is often used as a chronotherapeutic threshold) was associated with a lower risk of pancreatic cancer incidence and mortality. Predictive ability of daytime light metrics exceeded that of sleep quality, diet, depression, and alcohol consumption. Higher daytime light exposure was associated with lower risks of gastrointestinal cancer incidence and mortality, especially for pancreatic cancer, indicating a potential protective effect that warrants further investigation in prevention and prognostic contexts.

Keywords: bright light, daytime light, gastrointestinal cancer, mortality, pancreatic cancer, UK Biobank

What's new?

Insufficient daytime light exposure may disrupt circadian rhythms and increase cancer risk. In this study, the impact of daytime light exposure on gastrointestinal cancer incidence and mortality was assessed using data on UK Biobank participants who were followed over a median of 8.8 years. Analyses show that increased daytime light intensity and duration are linked to reduced gastrointestinal cancer risk and mortality. The strongest protective effects were observed for pancreatic cancer. Notably, daytime light exposure outperformed several established predictors, including sleep quality and diet. Bright daylight is a potentially modifiable factor for cancer prevention and prognosis, warranting further investigation.

Higher level of daytime light exposure were significantly associated with a reduced risk of gastrointestinal cancers. In this cohort of 89,069 participants, 1692 incident cases and 891 deaths were recorded. The protective effect was most pronounced for pancreatic cancer incidence and total gastrointestinal cancer incidence. SHAP value analysis further identified daytime light as a potential predictor of clinical outcones, demonstrating its significance alongside established risk factors such as BMI and smoking.

graphic file with name IJC-159-1401-g002.jpg

Abbreviations

BMI: body mass index
CI: confidence interval
FDR: false discovery rate
HR: hazard ratio
ICD: international classification of diseases
MVPA: moderate‐to‐vigorous physical activity
NHS: The National Health Service
WHO: World Health Organization

1. Introduction

Gastrointestinal cancer is the fifth most frequently diagnosed malignancy and fourth leading cause of cancer‐related mortality worldwide. Both the incidence and mortality rates of gastrointestinal cancer have shown an increasing trend [1, 2]. Consequently, identifying modifiable factors is urgently needed to establish effective prevention strategies. Daytime light exposure is a crucial environmental factor that maintains key physiological functions, such as tissue‐ and cellular‐level circadian rhythms, hormone secretion, immune function, and cell proliferation [3], all of which are implicated in the incidence and mortality of gastrointestinal cancer [4]. However, in modern society, people spend most of their time indoors, resulting in a common deficiency of daytime light exposure [5]. Moreover, patients with gastrointestinal cancer may experience reduced daytime light due to hospitalization, chemotherapy‐induced fatigue, and decreased outdoor activity [6]. The potential effect of daytime light exposure on gastrointestinal cancer has gradually attracted attention.

High levels of daytime light exposure have been shown to reduce adipose tissue [7], which is a risk factor for gastrointestinal cancer [8]. In addition, higher levels of daytime light exposure increase melatonin rhythm regularity and improve sleep quality [9], both of which contribute to better symptom management and enhance the quality of life in patients with gastrointestinal cancer [10]. In addition, bright light therapy (typically ≥ 5000 lux for ≥ 0.5 h/day over at least 4 weeks) significantly improves sleep quality and reduces fatigue in cancer patients [11] Additionally, higher daytime bright light exposure is associated with better gastrointestinal activity [12]. These physiological improvements have the potential to prolong the overall survival of individuals diagnosed with gastrointestinal cancer [13]. These findings highlight the wide‐ranging physiological benefits of daytime light exposure and support its potential role in gastrointestinal cancer prevention and treatment.

In view of this emerging evidence, the present cohort study (n = 89,069) aimed to evaluate the association between objectively measured daytime light exposure, including both the average intensity and duration of bright light, and the incidence and mortality of gastrointestinal cancer. In addition, machine learning analyses were performed to compare the predictive ability of daytime light intensity and duration with other well‐established predictors.

2. Methods

2.1. Study Design and Participants

The UK Biobank (RRID:SCR_012815) is a large prospective study with 502,490 participants recruited across England, Scotland, and Wales, between 2006 and 2010 [14]. Accelerometry and light data were collected for a subset of 103,635 participants between 2013 and 2015. A total of 14,567 participants with raw accelerometer data were excluded for the following reasons: (1) Data problem indicators, (2) Inadequate quality wear time, (3) Poor calibration quality based on the participant's own data, (4) Interrupted recording periods, (5) Data recording errors, (6) Long‐term dim or bright light exposure. The details of the participant selection for the current study are presented in Figure S1. The final analysis included 89,069 participants with valid daylight data. Participants who accepted the invitation to join the UK Biobank cohort provided written informed consent and were reimbursed for travel expenses.

2.2. Exposures

Raw light data were extracted from. cwa files, timezone‐adjusted, converted to lux using the Axivity AX3 device formula, zero‐corrected, calibrated, and resampled into 5‐s epochs using median values, details are shown in Supplementary Methods [15]. Missing values were imputed based on matched acceleration data or average valid data from similar time intervals over 7 days. Participants with > 28.5% (2*24 h/7*24 h) missing data or consistently implausible light levels were excluded. Calibrated lux values were then averaged into 48 half‐hour clock time bins across the 24‐h cycle. The presence of non‐wear was determined using GGIR, a validated package for estimating the sleep–wake state from accelerometer data [16]. The daytime light (7:30–20:30) groupings were extracted using factor analysis based on loading ≥ 0.5 (varimax rotation, cumulative proportion variance explained = 0.56). The daylight factors demonstrated internal consistency (Cronbach's α = 0.98 and α = 0.93, respectively) and exhibited a weak positive correlation (rs = 0.10, p < 0.0001) [15]. The light exposure predictors were defined through a factor analysis of 48 half‐hourly light bins, identifying a significant daytime light factor from 7:30 to 20:30 [17]. This division into daytime and nighttime periods was further supported by the hourly risk patterns of cancer incidence and mortality across the 24‐h cycle (Figure S2). Considering the widely used thresholds of light intensity for bright light therapy from a previous study [18, 19], the duration of daytime light exposure was calculated with a light intensity ≥ 5000 lux.

2.3. Outcomes

Specific types of cancer were classified according to the International Classification of Diseases, Tenth Revision (ICD‐10) (Field ID: 40006), as well as the interpolated year in which the cancer was first diagnosed (Field ID: 20006). Incident cancer cases were identified through data linkage to the national cancer and mortality registries. The National Health Service (NHS) Digital provided these data for participants in England and Wales, and the NHS Central Registries provided these data for participants in Scotland. Cancer death was defined as cancer being the primary cause of death. An exhaustive examination of all gastrointestinal cancers (C15‐C26) was conducted, including colorectal cancer (C18‐C20) and pancreatic cancer (C25); the detailed list for each site‐specific cancer is presented in (Table S1). Additionally, only colorectal and pancreatic cancers were reported separately because these sites had sufficient case numbers (≥ 200) to provide stable estimates; other gastrointestinal cancer sites had fewer cases and are included in the overall gastrointestinal cancer category. Cox regression analyses were right‐censored at the date of death or the last available follow‐up date (January 25, 2024), whichever occurred first.

2.4. Covariates

Covariates were obtained using three distinct methodologies: accelerometers, self‐reporting questionnaires, and registry records. The age of the participants was assessed when the accelerometer was worn (from 2013.06.01 to 2015.12.23). Self‐report questionnaires at baseline were used to determine sex (female/male), ethnicity (white/non‐white), region of recruitment (England/Wales/Scotland), and Townsend deprivation index (continuous). The latter is a measurement of socioeconomic status and is based on the postcode of residence and includes data on unemployment, car and home ownership, and household overcrowding. Body mass index (BMI) categories were calculated from height and weight measurements obtained during the baseline center visits. The categories were normal weight or underweight (BMI < 25 kg/m²), overweight (BMI: 25–30 kg/m²), and obesity (BMI: ≥ 30 kg/m²). Education level (higher education, any other qualification, or no qualification), smoking status (never smoked, former smoker, or current smoker), frequency of alcohol consumption (no current, up to twice a week, three or more times per week), and diet‐related factors (Cooked vegetable intake diet, Salad/raw vegetable intake diet, fresh fruit intake diet, dried fruit intake diet, oily fish intake diet, non‐oily fish intake diet, processed meat intake diet, beef intake diet, lamb/mutton intake diet, and pork intake diet) were obtained from the touchscreen questions. Previous diagnoses of hypertension, diabetes, and depression were obtained from self‐report questionnaires, hospital records, and death registry. Family history of cancer, which refers to first‐degree relatives and includes any type of cancer, chronotype (derived from responses to a chronotype question that participants answered on a touch‐screen computer), photoperiod (derived as the interval between sunrise and sunset on the date of light‐tracking at 53.4808° N, 2.2426° W (Manchester, UK), using ‘getSunlightTimes()’ in the ‘suncalc’ package in R), sleep efficiency, and moderate‐to‐vigorous physical activity (recorded by an accelerometer; details on MVPA data processing are provided in our previous work) were included [20]. The values of certain covariates, including BMI, education level, smoking status, alcohol consumption, healthy diet score, and diabetes history, were obtained from touchscreen questionnaires at the time point closest to accelerometer wear (Figure S3 and Table S2).

2.5. Statistical Analysis

The number of incidence and mortality events was sufficient, with at least ten events per variable. Missing data were minimal, with less than 5% missing data for each covariate, and were dispersed throughout the dataset. Multiple imputation, using the “mice” R package (version 3.13.0, RRID:SCR_001905), was employed to address missing values and reduce inferential bias. Exposures, covariates, mortality outcomes, and time to outcomes were included in the imputation process. The baseline characteristics of the full and complete sample are shown in (Figure S4).

The associations between daytime light exposure and gastrointestinal cancer incidence and mortality were assessed using a Cox proportional hazards regression model (using the Kaplan–Meier “survival” package version 3.2‐11 in R, RRID:SCR_001905). This analysis yielded HRs and 95% CIs calculated for different adjustment models. Model 1 adjusted for age and sex; Model 2 further adjusted ethnicity, deprivation index, recruitment center, education, BMI, healthy diet score, alcohol consumption, smoking status, and month of wear; Model 3 further adjusted photoperiod, chronotype, MVPA, sleep duration, and family history of cancer. The Spearman correlation coefficient matrix for these covariates is shown in Figure S5. The cumulative risk curves for gastrointestinal cancer, pancreatic cancer, and colorectal cancer are shown in (Figures S6–S8). This method defines an 0%–80% reference category and 80%–100% exposure categories as comparison groups.

A series of sensitivity analyses was performed. First, we used a complete case analysis that included participants without missing covariate data. Second, we calculated Fine‐Gray subdistribution hazards (using the “cmprsk” package v2.2‐10 in R, RRID:SCR_001905), incorporating other‐cause death as a competing risk for cause‐specific incidence and mortality. Third, we further adjusted for gastrointestinal cancer‐related disorders, including gastrointestinal reflux disease, gastric ulcer, gastric duodenal ulcer, duodenal ulcer, gastritis and duodenitis, acute appendicitis, irritable bowel syndrome, and nonalcoholic fatty liver disease [21]. Fourth, we excluded participants who wore an accelerometer for fewer than 6 days. Fifth, we restricted the analyses to participants without a history of shift‐work. Sixth, Vitamin D levels were indicated by serum 25(OH)D concentrations (field ID: 30890). We included participants with both light data and serum 25(OH)D data in this sensitivity analysis. The concentration of 25(OH)D (10–240 nmol/L) did not fall outside the valid range of the assay (10–375 nmol/L) [22].

Multiplicative and additive interaction analyses (using the “interactionR” package v0.1.3.9000 in R) were performed on age, sex, smoking status, alcohol consumption, depression, and obesity (meeting the WHO recommendation) [23]. In order to clarification on whether the proportional hazards assumption was formally tested, the proportional hazards assumption using Schoenfeld residuals for all models were performed (Figures S9–S11). Furthermore, the XGBoost machine (RRID:SCR_021361) learning method was used to rank the predictive ability of daytime light intensity and duration and traditional risk factors, such as moderate‐to‐vigorous physical activity (MVPA), sleep efficiency, sleep duration (longer sleep duration and shorter sleep duration), smoking status, alcohol consumption, obesity (BMI ≥ 30 kg/m²), healthy diet score, depression, and family history of cancer. XGBoost calculates variable importance based on performance improvements at each splitting point, measured by the Gini index, followed by gradient boosting. Optimal model performance was achieved through 10‐fold cross‐validation. SHapley Additive exPlanations (SHAP, RRID:SCR_021362) values were also used to assess the relative importance of daytime light exposure for gastrointestinal cancer incidence and mortality. We implemented the XGBoost algorithm with 10‐fold cross‐validation to assess the predictive importance of daytime light exposure and established risk factors for gastrointestinal cancer outcomes. Model hyperparameters were set as follows: learning rate = 0.1, maximum tree depth = 5, number of estimators = 500, and subsample = 0.8. The outcomes analyzed included overall gastrointestinal cancer incidence and mortality, as well as colorectal and pancreatic cancer separately. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy, and calibration metrics. SHAP values were subsequently computed to rank feature importance for each outcome.

All statistical tests were two‐tailed, and p ≤ 0.05 was considered significant. To mitigate the occurrence of false positives across all analyses, and in the event that any type I error rate was included, all p values in the fully adjusted models were corrected using the false discovery rate (FDR) method. All statistical analyses were performed using the R software (version 4.0.3, RRID:SCR_001905).

3. Results

A total of 89,069 participants who wore accelerometers were included in the study. Two key metrics, average daytime light exposure (lux) and duration, were derived from accelerometer data. The participants were followed up for a median period of 8.8 years (804,111 person‐years) until January 24, 2024. In total, 1692 developed gastrointestinal cancer and 891 died from gastrointestinal cancer, with deaths counted from the entire cohort. Lower daytime light exposure is characterized by a daytime light intensity threshold of 1916 l×. The baseline characteristics of the study participants are shown in Table 1. The age of the participants ranged from 44 to 79 years. The low‐ and high‐light exposure groups exhibited comparable characteristics across a range of key variables including ethnicity, socioeconomic status, geographic region, education, smoking, alcohol use, and family history of cancer.

TABLE 1.

Characteristic of participants.

Characteristic	Lower daytime light exposur (n = 71,255)	Higher daytime light exposure (n = 17,814)	p
Age at accelerometer mean (SD), years	62.1 (7.92)	63.0 (7.63)	< 0.001
Sex
Male	30,480 (42.8)	8318 (46.7)	< 0.001
Female	40,775 (57.2)	9496 (53.3)	< 0.001
Ethnicity
White	68,789 (96.5)	17,497 (98.2)	< 0.001
Other	2466 (3.5)	317 (1.8)	< 0.001
Townsend deprivation index, median [IQR]	−2.35 (3.78)	−2.73 (3.10)	< 0.001
Recruitment regions
England	63,870 (89.6)	16,086 (90.3)	< 0.001
Wales	2580 (3.6)	736 (4.1)
Scotland	4805 (6.7)	992 (5.6)
Education level
Higher education	31,408 (44.1)	7529 (42.3)	< 0.001
Any other qualification	33,980 (47.7)	8763 (49.2)
No qualification	5867 (8.2)	1522 (8.5)
Body mass index categories
Normal/Underweight (≤ 25 kg/m²)	28,168 (39.5)	7227 (40.6)	< 0.001
Overweight (25–30 kg/m²)	28,908 (40.6)	7482 (42.0)	< 0.001
Obese (> 30 kg/m²)	14,179 (19.9)	3105 (17.4)
Mean healthy diet score (SD)	3.68 (1.17)	3.72 (1.17)	0.02
Smoking status
Never smoked	41,180 (57.8)	10,072 (56.5)	< 0.001
Previous smoker	25,457 (35.7)	6705 (37.6)
Current smoker	4618 (6.5)	1037 (5.8)
Alcohol consumption
None current	4437 (6.2)	934 (5.2)	< 0.001
Up to twice a week	33,709 (47.3)	7725 (43.4)
Three or more times a week	33,109 (46.5)	9155(51.4)
Start time of wear month, mean (SD)	6.90 (3.64)	6.46 (2.17)	< 0.001
Photoperiod, mean (SD)	11.5 (3.01)	15.3 (1.71)	< 0.001
MVPA, mean (SD)	148.8 (142.8)	185.3 (169.5)	< 0.001
Sleep duration, mean (SD)	7.28 (0.91)	7.27 (0.86)	0.65
Family history of cancer
Yes	21,622 (30.3)	5505 (30.1)	0.72
No	49,632 (69.7)	12,309 (69.9)

Open in a new tab

Abbreviations: IQR, interquartile range; MVPA, moderate‐to‐vigorous physical activity; SD, standard deviation.

3.1. Associations Between Daytime Light Exposures and Gastrointestinal Cancer Incidence and Mortality

In the fully adjusted Model 3, higher daytime light exposure exhibited a statistically significant correlation with a reduced incidence of gastrointestinal cancer (HR = 0.87, 95% CI: 0.76–0.99, p = 0.04) and mortality (HR = 0.76, 95% CI: 0.63–0.93, p = 0.008), the details are shown in Figure 1. Specifically, elevated daytime light exposure was associated with a reduced risk of pancreatic cancer incidence (HR = 0.58, 95% CI: 0.40–0.85, p = 0.005) and mortality (HR = 0.47, 95% CI: 0.30–0.73, p = 0.001). However, the correlations between daytime light exposure and colorectal cancer incidence (HR = 0.96, 95% CI: 0.80–1.16, p = 0.66) and mortality (HR = 0.93, 95% CI: 0.66–1.32, p = 0.70) were not statistically significant. Notably, results from Model 1 (age and sex) and Model 2 (age sex and others) were consistent with those from Model 3, indicating the robustness of the observed associations (Table S3).

The association between daytime light exposure and the incidence and mortality of gastrointestinal cancers. The association between daytime light exposure and gastrointestinal cancer incidence and mortality was adjusted for in a fully model analysis that included the following variables: Age, sex, ethnicity, Townsend deprivation index, recruitment center, education level, BMI, healthy diet score, alcohol consumption, smoking status, start time of wear month, photoperiod, MVPA, sleep e fficiency, chronotype and family history of cancer. A reference category of 0%–80% exposure was defined, with 80%–100% exposure serving as the comparison group. The threshold for average daylight intensity was 1916, and the threshold for average daylight duration was 2.4 h per day. BMI, body mass index; CI, confidence interval; HR, hazard ratio; MVPA, moderate‐to‐vigorous physical activity. ^aAll p‐values remained significant after multiple testing with the false discovery rate method. [Color figure can be viewed at wileyonlinelibrary.com]

To further explore the impact of extended bright light exposure, analyses were conducted using a threshold of ≥ 2.4 h/day at intensities ≥ 5000 l× with a reference category of 0%–80%, and 80%–100% exposure categories as comparison groups. The overall pattern of associations demonstrated consistency. Despite the lack of statistical significance in the association with overall gastrointestinal cancer incidence (HR = 0.89; 95% CI: 0.77–1.01; p = 0.08) and mortality (HR = 0.86; 95% CI: 0.70–1.05; p = 0.12). Specifically, extended exposure to high‐intensity light was significantly associated with a reduced risk of pancreatic cancer incidence (HR = 0.67; 95% CI: 0.46–0.97; p = 0.04) and mortality (HR = 0.57; 95% CI: 0.37–0.87; p = 0.009). No significant associations were observed for colorectal cancer outcomes (Table S3), suggesting a potential site‐specific effect of daytime light exposure.

3.2. Sensitivity Analyses

A series of sensitivity analyses was conducted to evaluate the robustness of the observed associations. In analyses restricted to participants without missing data, higher daytime light exposure was still significantly associated with a lower risk of both gastrointestinal and pancreatic cancer (Table S4). Furthermore, the results were consistent when employing Fine‐Gray subdistribution hazard models to account for competing risks and in a subsample that excluded individuals with relevant disorders or fewer than six valid accelerometer wear days (Tables S5–S7). The observed associations were significant only for pancreatic cancer after excluding participants with a history of shift work (Table S8), which suggests the presence of confounding factors or effect modification by circadian‐disruptive occupational factors. In sensitivity analyses additionally adjusting for serum vitamin D, the association with gastrointestinal cancer incidence was attenuated and did not reach statistical significance, whereas the inverse associations with pancreatic cancer incidence and mortality remained statistically significant (Table S9).

3.3. Interaction Analyses

To examine potential effect modifications, interaction analyses were performed across key subgroups, including age, sex, smoking status, alcohol consumption, obesity, and depression. The detailed results are presented in Figure S12 and Tables S10–S15. Across both the additive and multiplicative interaction models, no statistically significant interactions were observed between daytime light exposure and any of the covariates investigated. These findings suggest that the association between daytime light exposure and risk of gastrointestinal cancer did not differ substantially across subgroups.

3.4. Predictive Capacity of Daytime Light Exposure

We further applied the SHAP analysis to evaluate the relative predictive contribution of daytime light intensity and duration, along with 12 established risk factors for gastrointestinal cancer incidence and mortality. Among the evaluated factors, daytime light exposure and duration demonstrated stronger predictive efficacy for gastrointestinal cancer than several well‐recognized gastrointestinal cancer‐related risk factors, including BMI, smoking status, sleep efficiency, MVPA, and healthy diet score. Similarly, for pancreatic cancer incidence and mortality, light intensity and duration were ranked as the fifth and seventh most influential predictors, respectively. In the context of colorectal cancer, daytime light intensity and duration were identified as the fifth and seventh most significant predictors, respectively (Figure 2).

The importance of the predictive features is ranked by SHAP values. Shapley additive explanation (SHAP) values were derived from the XGBoost machine learning model to predict incidence and mortality. The X‐axis shows the outcomes and the Y‐axis shows the predictive ability of the factors. Deeper blue indicates the most important predictive ability and lighter blue indicates the least important predictive ability. BMI, body mass index; MVPA, moderate‐to‐vigorous physical activity. [Color figure can be viewed at wileyonlinelibrary.com]

4. Discussion

This large‐scale prospective cohort study provides robust evidence that higher levels of daytime light are associated with lower risk of gastrointestinal cancer incidence and mortality. These associations remained stable across multiple sensitivity analyses and were not modified by age, sex, smoking status, alcohol consumption, obesity, or depression. Notably, daytime light exposure showed robust and consistent predictive performance for gastrointestinal cancer outcomes within this cohort, comparable to or exceeding that of several established predictors, although differences in measurement precision across variables should be considered when interpreting these comparisons.

The current investigation addresses a critical knowledge gap, as there has been limited direct evidence linking daytime light exposure with gastrointestinal cancer outcomes. The present study showed that individuals with higher daytime light exposure (mean: 2559 lux) had a 13%–24% risk reduction in overall gastrointestinal cancer incidence and mortality, and a considerable 42%–53% risk reduction for pancreatic cancer incidence and mortality, compared to those with lower daytime light intensity (mean: 804 lux). These findings are supported by existing ecological research demonstrating an inverse association between solar radiation and prostate cancer risk [24]. In addition, the duration of urban photoperiod affects metabolism and obesity [25], which are risk factors for gastrointestinal cancer [26]. A study of 139 patients with pancreatic cancer indicated that those living in regions with higher sunlight exposure had better overall survival rates than those living in regions with lower sunlight exposure [27]. Additionally, a randomized controlled trial of 52 patients with gastrointestinal cancer demonstrated that exposure to bright daytime light improved sleep quality, reduced fatigue, and enhanced daily functioning. These improvements were associated with better overall survival [6].

Light therapy is a non‐pharmacological therapy that is currently being studied for cancer‐related symptoms and has been recognized by the FDA as a low‐risk intervention [11], which motivated us to further explore the association between bright daytime light duration and gastrointestinal cancer outcomes. It was observed in our study that a bright light exposure duration of ≥ 2.4 h/day (daytime light intensity ≥ 5000 lux) was associated with an 11%–33% reduction in incidence and a 14%–43% reduction in mortality of gastrointestinal cancer. According to a substantial corpus of supporting evidence, exposure to bright light (≥ 5000 lux) for 30–45 min over the course of several weeks can improve metabolic regulation and reduce body fat [28]. These factors contribute to a reduced risk of gastrointestinal cancer [29]. Furthermore, observational studies have shown that bright light therapy can improve chemotherapy response and prolong survival in patients with gastrointestinal cancer [30].

Multiple plausible mechanisms may underlie these associations. Exposure to bright daytime light regulates circadian rhythms through the suprachiasmatic nucleus, mitigating the nocturnal decline of melatonin. This, in turn, promotes sleep consolidation [31], and improved sleep quality has been associated with a reduced risk of cancer [32]. Moreover, melatonin has been shown to possess anti‐tumor effects, including inhibition of cancer cell proliferation, angiogenesis, and invasion, as well as enhancement of immune surveillance [33], all of which showed important protective and regulatory effects in the gastrointestinal tract [34]. Furthermore, patients with gastrointestinal cancer frequently experience persistent sleep disturbances due to inflammation, psychological distress, and treatment‐related side effects, all of which contribute to poor outcomes [35]. Therefore, for patients afflicted with gastrointestinal cancer, increased exposure to daylight may result in a marked improvement in overall survival [36]. In addition, exposure to daytime light has been demonstrated to increase energy levels, ameliorate emotional well‐being, and promote health‐conscious behaviors. This positive feedback loop has been shown to foster optimal sleep and survival outcomes in patients with gastrointestinal cancer [37, 38].

Although the timing of light exposure is biologically relevant to circadian regulation, our findings suggest that for long‐latency outcomes such as gastrointestinal cancer, averaged daytime and 24‐h light exposure may better capture habitual circadian entrainment than transient nocturnal light disturbance. Therefore, in our study, the absence of an association between nighttime light exposure and gastrointestinal cancer risk may reflect limitations of wrist‐worn sensors in capturing retinal light exposure, particularly short‐wavelength light from electronic devices, as well as exposure misclassification due to sensor occlusion. In addition, light exposure was assessed between 2014 and 2018, preceding the widespread nighttime use of personal electronic devices. Finally, daytime light exposure may biologically modulate susceptibility to nighttime light, consistent with the photic preload hypothesis. Further studies using ocular‐level light assessment and contemporary exposure patterns are warranted.

Notably, other established lifestyle risk factors, including diet, physical activity, alcohol consumption, and smoking, exhibited stronger associations with colorectal cancer than with pancreatic cancer in our cohort, which is consistent with previous studies [39]. Furthermore, Vitamin D is another widely discussed mechanism linking sunlight exposure to cancer risk through its roles in immune modulation, cell differentiation, and anti‐inflammatory signaling. In our cohort, serum vitamin D measurements were available for a subset of participants, allowing additional sensitivity analyses to evaluate whether vitamin D status explained or modified the observed associations. These analyses did not demonstrate statistically attenuation of effects after accounting for vitamin D levels, indicating that vitamin D alone is unlikely to fully account for the protective associations observed between daytime light exposure and gastrointestinal cancer outcomes in this population. Nevertheless, given that vitamin D data were not available for the entire cohort and were measured at a single time point, residual effects cannot be completely excluded. Therefore vitamin D remains a plausible contributory mechanism and should be considered alongside circadian, metabolic, and neuroendocrine pathways rather than as a sole explanatory factor.

In contrast, higher daytime light exposure showed a stronger inverse association with pancreatic cancer incidence than colorectal cancer, highlighting potential cancer‐site‐specific sensitivity to circadian and light‐related regulation. This may reflect the unique circadian sensitivity of pancreatic tissue which exhibits extensive circadian transcriptomic reprogramming in both pancreatic adenocarcinoma and aging, particularly in metabolic and stress‐related pathways. As circadian disruption is known to impair β‐cell function and metabolic homeostasis, enhanced daytime light exposure may preferentially stabilize circadian‐metabolic coupling in the pancreas, offering a plausible explanation for the stronger association observed in this cancer subtype [40, 41].

5. Strengths and Limitations

Our research indicated a correlation between higher levels of daytime light exposure and the incidence and mortality of gastrointestinal cancer, and this correlation might even surpass that of some traditional predictors. Increasing daytime light exposure may serve as a simple and cost‐effective measure to reduce the risk of gastrointestinal cancer in both the community and clinical settings. Additionally, this study proposes protective thresholds for daytime light exposure (intensity and duration), which may inform future interventional clinical studies. These findings suggest a targeted approach to mitigate gastrointestinal cancer incidence and mortality by increasing daytime light levels.

This study has several limitations. First, the UK Biobank sample was healthier and less socioeconomically deprived than the general population, which may limit the generalizability of the prevalence estimates. However, exposure‐outcome relationships remain widely applicable [42]. Second, while 7‐day light monitoring has been shown to be a reliable method for capturing weekly patterns, it may not fully reflect long‐term behavior [43]. Third, light exposure was not measured at the ocular level, providing only a rough estimate of its effects on gastrointestinal cancer. Fourth, the limited temporal resolution of wearable light sensors constrained our ability to precisely characterize the timing of light exposure across the 24‐h cycle, particularly during the habitual rest phase. Fifth, despite our efforts to exclude biased data, the device's inability to discern true darkness from merely being covered might result in underestimation. Last, the relatively large HR for pancreatic cancer may reflect residual confounding because of the limited number of cases. The reported lux values were validated 8 years later and showed strong accuracy. Finally, while some covariates were collected 9.7 years before baseline, their stability over time suggests that this does not weaken the findings [44].

6. Conclusion

High levels of daytime light exposure were associated with a reduced risk of gastrointestinal cancer incidence and mortality, particularly pancreatic cancer. These findings underscore the potential role of daytime light as a modifiable environmental factor for the prevention and prognosis of gastrointestinal cancer.

Author Contributions

Xionge Mei: conceptualization, formal analysis, data curation, methodology, writing – original draft. Wei Wang: conceptualization, writing – original draft, methodology, formal analysis, data curation. Nana Zheng: methodology, visualization. Tong Luo: validation, software. Jiaying Xu: conceptualization, software. Ngan Yin Chan: conceptualization, resources. Jing Wang: validation, writing – review and editing. Yannis Yan Liang: conceptualization, writing – review and editing. Sizhi Ai: conceptualization, methodology. Yaping Liu: conceptualization, methodology. Xiao Tan: conceptualization, methodology, resources. Christian Benedict: writing – review and editing, resources. Yun Kwok Wing: writing – review and editing, resources, supervision. Jihui Zhang: supervision, funding acquisition, writing – review and editing. Hongliang Feng: conceptualization, writing – review and editing, methodology, software, resources.

Funding

This work was supported by grants from the National Science and Technology Innovation 2030 of China‐Major Projects (Grant No. 2022ZD0214100), the National Natural Science Foundation of China (Grant No. 82101558, 82171476, and 82341240), Guangzhou Municipal School (College)‐Enterprise Joint Funding Project (Grant No. 2024A03J0214 and 2025A03J3354), the National Key R&D Program of China (Grant No. 2021YFC2501500), Guangzhou Key Clinical Specialty (Clinical Medical Research Institute), and the Guangzhou Municipal Key Discipline in Medicine (2025–2027). The funder had no role in the design and conduct of the study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Flowchart of participant enrolment.

Figure S2: Hazard ratios [95% CI] of gastrointestinal cancers incidence and mortality for light exposures across 24 h.

Figure S3: Timeline of covariates collection.

Figure S4: The missing data pattern of the covariates.

Figure S5: Correlation matrix of the covariates.

Figure S6: Cumulative risk curves of daytime light exposure and gastrointestinal cancer risk and mortality.

Figure S7: Cumulative risk curves of daytime light exposure and pancreas cancer risk and mortality.

Figure S8: Cumulative risk curves of daytime light exposure and colorectal cancer risk and mortality.

Figure S9: SH curves of daytime light exposure and gastrointestinal cancer risk and mortality.

Figure S10: SH curves of daytime light exposure and pancreas cancer risk and mortality.

Figure S11: SH curves of daytime light exposure and colorectal cancer risk and mortality.

Figure S12: Interaction analysis of daylight exposure and gastrointestinal cancers risk and mortality.

Table S1: Coding of site‐specific cancer among UK Biobank participants.

Table S2: Source of covariates and outcomes.

Table S3: The associations of daytime light exposures and gastrointestinal cancer risk and cancer mortality.

Table S4: Sensitivity analysis on the association of daytime light exposures with gastrointestinal cancer risk and cancer mortality by complete cases.

Table S5: Sensitivity analysis on the association of daytime light exposures with gastrointestinal cancer risk and cancer mortality by competing risk.

Table S6: Sensitivity analysis on the association of daytime light exposures with gastrointestinal cancer risk by excluding participants with relevant disorder.

Table S7: Sensitivity analysis on the association of daytime light exposures with gastrointestinal cancer risk by excluding participants with wearing days < 6 days.

Table S8: Sensitivity analysis on the association of daytime light exposures with gastrointestinal cancer risk by excluding participants with shift work.

Table S9: Sensitivity analysis of the association between daytime light exposure and gastrointestinal cancer risk with additional adjustment for serum vitamin D.

Table S10: Interaction effects of daytime light exposure and age categories on gastrointestinal risk and mortality.

Table S11: Interaction effects of daylight exposure and sex categories on gastrointestinal cancer risk and mortality.

Table S12: Interaction effects of daylight exposure and smoking categories on gastrointestinal cancer risk and mortality.

Table S13: Interaction effects of daylight exposure and alcohol categories on gastrointestinal cancer risk and mortality.

Table S14: Interaction effects of daylight exposure and depression categories on gastrointestinal cancer risk and mortality.

Table S15: Interaction effects of daylight exposure and obesity categories on gastrointestinal cancer risk and mortality.

IJC-159-1401-s001.pdf^{(6MB, pdf)}

Contributor Information

Jihui Zhang, Email: jihui.zhang@cuhk.edu.hk.

Hongliang Feng, Email: hlfeng@gzhmu.edu.cn, Email: hlfeng@link.cuhk.edu.hk.

Data Availability Statement

All source code is publicly available on GitHub (https://github.com/xiongemei/UK‐Biobank). This work has been conducted using the UK Biobank Resource under Application Number 58082. The UK Biobank is an open access resource, and bona fide researchers can apply to use the UK Biobank dataset by registering and applying at http://ukbiobank.ac.uk/register‐apply/. Further information is available from the corresponding author upon request.

References

1. Joshi S. S. and Badgwell B. D., “Current Treatment and Recent Progress in Gastric Cancer,” CA: A Cancer Journal for Clinicians 71, no. 3 (2021): 264–279, 10.3322/caac.21657. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Lin Y., Zheng Y., Wang H. L., and Wu J., “Global Patterns and Trends in Gastric Cancer Incidence Rates (1988‐2012) and Predictions to 2030,” Gastroenterology 161, no. 1 (2021): 116–127, 10.1053/j.gastro.2021.03.023. [DOI] [PubMed] [Google Scholar]
3. Fortin B. M., Mahieu A. L., Fellows R. C., et al., “The Diverse Roles of the Circadian Clock in Cancer,” Nature Cancer 26 (2025): 1–5, 10.1038/s43018-025-00981-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Chen H. M., Cheung D. S., Huang C. S., Wu Y. C., and Hsu P. K., “Mediation and Dose‐Response Relationship Among Physical Activity, Daylight Exposure, and Rest‐Activity Circadian Rhythm in Patients With Esophageal and Gastric Cancer,” Cancer Nursing 48 (2022): 10–97, 10.1097/NCC.0000000000001306. [DOI] [PubMed] [Google Scholar]
5. Brown T. M., Brainard G. C., Cajochen C., et al., “Recommendations for Daytime, Evening, and Nighttime Indoor Light Exposure to Best Support Physiology, Sleep, and Wakefulness in Healthy Adults,” PLoS Biology 20, no. 3 (2022): e3001571, 10.1371/journal.pbio.3001571. [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Chen H. M., Liu J. H., Huang C. S., Dai M. F., and Chien L. I., “Moderating Effect of Daylight Exposure on the Relationship Between Physical Activity and Sleep in Patients With Upper Digestive Cancer: An Exploratory Study,” Seminars in Oncology Nursing 40, no. 4 (2024): e151659, 10.1016/j.soncn.2024.151659. [DOI] [PubMed] [Google Scholar]
7. Obayashi K., Saeki K., and Kurumatani N., “Ambient Light Exposure and Changes in Obesity Parameters: A Longitudinal Study of the HEIJO‐KYO Cohort,” Journal of Clinical Endocrinology and Metabolism 101, no. 9 (2016): 3539–3547, 10.1210/jc.2015-4123. [DOI] [PubMed] [Google Scholar]
8. Murphy N., Jenab M., and Gunter M. J., “Adiposity and Gastrointestinal Cancers: Epidemiology, Mechanisms and Future Directions,” Nature Reviews Gastroenterology & Hepatology 15, no. 11 (2018): 659–670, 10.1038/s41575-018-0038-1. [DOI] [PubMed] [Google Scholar]
9. Kamyab P., Kouchaki H., Motamed M., Boroujeni S. T., Akbari H., and Tabrizi R., “Sleep Disturbance and Gastrointestinal Cancer Risk: A Literature Review,” Journal of Investigative Medicine 71, no. 3 (2023): 163–172, 10.1177/10815589221140595. [DOI] [PubMed] [Google Scholar]
10. Xiao P., Ding S., Duan Y., et al., “Effect of Light Therapy on Cancer‐Related Fatigue: A Systematic Review and Meta‐Analysis,” Journal of Pain and Symptom Management 63, no. 2 (2022): e188–e202, 10.1016/j.jpainsymman.2021.09.010. [DOI] [PubMed] [Google Scholar]
11. Lai F., Luo Z., Zhang J., Xia W., and Tian L., “Bright Light Therapy Has a Positive Effect on Sleep Quality in Patients With Cancer: A Meta‐Analysis,” Sleep Medicine Reviews 75 (2024): 101925, 10.1016/j.smrv.2024.101925. [DOI] [PubMed] [Google Scholar]
12. Bishehsari F., Post Z., Swanson G. R., and Keshavarzian A., “Circadian Rhythms in Gastroenterology: The Biological Clock's Impact on Gut Health,” Gastroenterology 169 (2025): 8–1396, 10.1053/j.gastro.2025.06.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Ulrich C. M., Himbert C., Holowatyj A. N., and Hursting S. D., “Energy Balance and Gastrointestinal Cancer: Risk, Interventions, Outcomes and Mechanisms,” Nature Reviews Gastroenterology & Hepatology 15, no. 11 (2018): 683–698, 10.1038/s41575-018-0053-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
14. UK Biobank , “UK Biobank,” (2022), https://www.ukbiobank.ac.uk/.
15. Burns A. C., Windred D. P., Rutter M. K., et al., “Day and Night Light Exposure Are Associated With Psychiatric Disorders: An Objective Light Study in > 85,000 People,” Nature Mental Health 1, no. 11 (2023): 853–862, 10.1038/s44220-023-00135-8. [DOI] [Google Scholar]
16. Migueles J. H., Rowlands A. V., Huber F., Sabia S., and van Hees V. T., “GGIR: A Research Community‐Driven Open Source R Package for Generating Physical Activity and Sleep Outcomes From Multi‐Day Raw Accelerometer Data,” Journal for the Measurement of Physical Behaviour 2, no. 3 (2019): 188–196, 10.1123/jmpb.2018-0063. [DOI] [Google Scholar]
17. Windred D. P., Burns A. C., Lane J. M., et al., “Brighter Nights and Darker Days Predict Higher Mortality Risk: A Prospective Analysis of Personal Light Exposure in > 88,000 Individuals,” Proceedings of the National Academy of Sciences 121, no. 43 (2024): e2405924121, 10.1073/pnas.2405924121. [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Sit D. K., McGowan J., Wiltrout C., et al., “Adjunctive Bright Light Therapy for Bipolar Depression: A Randomized Double‐Blind Placebo‐Controlled Trial,” American Journal of Psychiatry 175, no. 2 (2018): 131–139, 10.1176/appi.ajp.2017.16101200. [DOI] [PubMed] [Google Scholar]
19. Chen G., Chen P., Yang Z., et al., “Increased Functional Connectivity Between the Midbrain and Frontal Cortex Following Bright Light Therapy in Subthreshold Depression: A Randomized Clinical Trial,” American Psychologist 79, no. 3 (2024): 437–450, 10.1037/amp0001218. [DOI] [PubMed] [Google Scholar]
20. Feng H., Yang L., Liang Y. Y., et al., “Associations of Timing of Physical Activity With All‐Cause and Cause‐Specific Mortality in a Prospective Cohort Study,” Nature Communications 14, no. 1 (2023): 930, 10.1038/s41467-023-36546-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Yuan S., Sun Y., Chen J., Li X., and Larsson S. C., “Long‐Term Risk of Venous Thromboembolism Among Patients With Gastrointestinal Non‐Neoplastic and Neoplastic Diseases: A Prospective Cohort Study of 484,211 Individuals,” American Journal of Hematology 99, no. 2 (2024): 172–181, 10.1002/ajh.27106. [DOI] [PubMed] [Google Scholar]
22. Revez J. A., Lin T., Qiao Z., et al., “Genome‐Wide Association Study Identifies 143 Loci Associated With 25 Hydroxyvitamin D Concentration,” Nature Communications 11, no. 1 (2020): 1647, 10.1038/s41467-020-15421-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Bull F. C., Al‐Ansari S. S., Biddle S., et al., “World Health Organization 2020 Guidelines on Physical Activity and Sedentary Behaviour,” British Journal of Sports Medicine 54, no. 24 (2020): 1451–1462, 10.1136/bjsports-2020-10295. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Sirhan‐Atalla M., Gabinet N. M., and Portnov B. A., “Disaggregating the Effects of Daytime and Nighttime Light Exposures on Obesity, Overweight, Prostate and Breast Cancer Morbidity Worldwide,” Chronobiology International 40, no. 4 (2023): 483–514, 10.1080/07420528.2023.2187230. [DOI] [PubMed] [Google Scholar]
25. Wyse C. A., Biello S. M., and Gill J. M., “The Bright‐Nights and Dim‐Days of the Urban Photoperiod: Implications for Circadian Rhythmicity, Metabolism and Obesity,” Annals of Medicine 46, no. 5 (2014): 253–263, 10.3109/07853890.2014.913422. [DOI] [PubMed] [Google Scholar]
26. O'Sullivan J., Lysaght J., Donohoe C. L., and Reynolds J. V., “Obesity and Gastrointestinal Cancer: The Interrelationship of Adipose and Tumour Microenvironments,” Nature Reviews Gastroenterology & Hepatology 15, no. 11 (2018): 699–714, 10.1038/s41575-018-0069-7. [DOI] [PubMed] [Google Scholar]
27. Eryılmaz M. K., Mutlu H., Gündüz Ş., et al., “More Sunlight Exposure May Improve the Overall Survival in Patients With Pancreas Cancer,” Journal of Oncological Sciences 2, no. 2–3 (2016): 73–76, 10.1016/j.jons.2016.08.002. [DOI] [Google Scholar]
28. Danilenko K. V., Mustafina S. V., and Pechenkina E. A., “Bright Light for Weight Loss: Results of a Controlled Crossover Trial,” Obesity Facts 6, no. 1 (2013): 28–38, 10.1159/000348549. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Melanson E. L., Ritchie H. K., Dear T. B., et al., “Daytime Bright Light Exposure, Metabolism, and Individual Differences in Wake and Sleep Energy Expenditure During Circadian Entrainment and Misalignment,” Neurobiol Sleep Circadian Rhythms 4 (2018): 49–56, 10.1016/j.nbscr.2017.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Strøm L., Danielsen J. T., Amidi A., et al., “Sleep During Oncological Treatment‐A Systematic Review and Meta‐Analysis of Associations With Treatment Response, Time to Progression and Survival,” Frontiers in Neuroscience 16 (2022): 817837, 10.3389/fnins.2022.817837. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Kim R., Fang Y., Lee M., et al., “Seasonal Timing and Interindividual Differences in Shiftwork Adaptation,” Npj Digital Medicine 28, no. 1 (2025): 300, 10.1038/s41746-025-01678-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Song C., Zhang R., Wang C., et al., “Sleep Quality and Risk of Cancer: Findings From the English Longitudinal Study of Aging,” Sleep 44, no. 3 (2021): zsaa192, 10.1093/sleep/zsaa192. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Huang X., Tao Q., and Ren C., “A Comprehensive Overview of the Neural Mechanisms of Light Therapy,” Neuroscience Bulletin 40, no. 3 (2024): 350–362, 10.1007/s12264-023-01089-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Han J., Xiong Z., Liu Y., et al., “Circadian Rhythm in Gastrointestinal Cancer: Clinical Applications and Future Perspectives,” Annals of Medicine 31, no. 1 (2025): 2529571, 10.1080/07853890.2025.2529571. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Lee Y., “Roles of Circadian Clocks in Cancer Pathogenesis and Treatment,” Experimental & Molecular Medicine 53, no. 10 (2021): 1529–1538, 10.1038/s12276-021-00681-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Lu L., Mullins C. S., Schafmayer C., Zeißig S., and Linnebacher M., “A Global Assessment of Recent Trends in Gastrointestinal Cancer and Lifestyle‐Associated Risk Factors,” Cancer Communications 41, no. 11 (2021): 1137–1151, 10.1002/cac2.12220. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Weiss C., The Effects of Morning Bright Light Therapy on Fatigue, Sleep Disturbances and Circadian Activity Rhythms in Lung Cancer Survivors (State University of New York at, 2018), 153. [Google Scholar]
38. Starreveld D. E., Daniels L. A., Kieffer J. M., et al., “Light Therapy for Cancer‐Related Fatigue in (Non‐) Hodgkin Lymphoma Survivors: Results of a Randomized Controlled Trial,” Cancers 13, no. 19 (2021): 4948, 10.3390/cancers13194948. [DOI] [PMC free article] [PubMed] [Google Scholar]
39. O'Sullivan D. E., Sutherland R. L., Town S., et al., “Risk Factors for Early‐Onset Colorectal Cancer: A Systematic Review and Meta‐Analysis,” Clinical Gastroenterology and Hepatology 20, no. 6 (2022): 1229–1240, 10.1016/j.cgh.2021.01.037. [DOI] [PubMed] [Google Scholar]
40. Sharma D., Adnan D., Abdel‐Reheem M. K., Anafi R. C., Leary D. D., and Bishehsari F., “Circadian Transcriptome of Pancreatic Adenocarcinoma Unravels Chronotherapeutic Targets,” JCI Insight 9, no. 9 (2024): e177697, 10.1172/jci.insight.177697. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Sharma D., Wessel C. R., Mahdavinia M., Preuss F., and Bishehsari F., “Reorganization of Pancreas Circadian Transcriptome With Aging,” Reorganization of Pancreas Circadian Transcriptome With Aging. Aging 15, no. 16 (2023): 7909–7921, 10.18632/aging.204929. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Fry A., Littlejohns T. J., Sudlow C., et al., “Comparison of Sociodemographic and Health‐Related Characteristics of UK Biobank Participants With Those of the General Population,” American Journal of Epidemiology 186, no. 9 (2017): 1026–1034, 10.1093/aje/kwx246. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Matthews C. E., Hagströmer M., Pober D. M., and Bowles H. R., “Best Practices for Using Physical Activity Monitors in Population‐Based Research,” Medicine and Science in Sports and Exercise 44, no. 1 Suppl 1 (2012): S68–S76, 10.1249/MSS.0b013e3182399e5b. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Strain T., Wijndaele K., Dempsey P. C., et al., “Wearable‐Device‐Measured Physical Activity and Future Health Risk,” Nature Medicine 26, no. 9 (2020): 1385–1391, 10.1038/s41591-020-1012-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials