Abstract
Objective
To investigate the relationships between sunlight exposure and the risk of three major rheumatic diseases (systemic lupus erythematosus [SLE], rheumatoid arthritis [RA], and gout) in a large-scale prospective cohort.
Methods
Data were derived from the UK Biobank (a nationwide prospective cohort in the United Kingdom), including 363,873 participants followed up from 2006 to 2022. Sunlight exposure was assessed by self-reported average daily outdoor time in summer and winter. Incident SLE, RA, and gout were identified using International Classification of Diseases, Tenth Revision (ICD-10) codes from linked hospital records. Multivariate Cox proportional hazards models, restricted cubic splines, subgroup analyses, and sensitivity analyses were performed to evaluate associations.
Results
During the 13-year follow-up, 335 SLE, 5,591 RA, and 7,485 gout cases were ascertained. Compared with sunlight exposure of 2.5 h/day, both lower and higher sunlight exposure were associated with increased risks of SLE (< 2.5 h/day: HR = 1.71, 95% CI:1.13, 2.59; >2.5 h/day: HR = 1.75, 95% CI:1.17, 2.63). We also observed increased risk associations of higher sunlight exposure with RA (HR = 1.15, 95% CI:1.06, 1.26) and gout (HR = 1.09, 95% CI:1.02, 1.18). Subgroup analyses suggested an age-related difference in the association between sunlight exposure and gout after Bonferroni correction (adjusted p value = 0.014).
Conclusion
Sunlight exposure was associated with disease-specific risks of SLE, RA, and gout.
Keywords: gout, outdoor light exposure, prospective cohort, restricted cubic spline, rheumatoid arthritis, systemic lupus erythematosus
Highlights
Sunlight exposure was associated with distinct risk patterns across SLE, RA, and gout in a large-scale prospective cohort.
Both lower and higher sunlight exposure were associated with increased SLE risk, whereas the associations with RA and gout were modest and should be interpreted cautiously.
Introduction
Sunlight exposure represents one of the most pervasive environmental stimuli encountered throughout human life, yet its long-term effects on immune-mediated diseases remain incompletely understood (1). Beyond its classical role as a source of cutaneous photodamage, advances in photoimmunology have revealed that ultraviolet radiation (UVR) acts as an important regulator of immune homeostasis through interactions involving cellular stress responses, antigen processing, and inflammatory signaling. Importantly, the immunological effects of UVR are highly context dependent (2, 3). Depending on exposure characteristics and host susceptibility, UVR may induce immune regulatory responses under physiological conditions while promoting inflammatory and autoimmune processes in susceptible individuals (4, 5). Therefore, whether the consequences of sunlight exposure are shared across immune-mediated diseases or shaped by disease-specific pathogenic mechanisms remains an important unresolved question.
Among rheumatic diseases, SLE is one of the conditions most strongly associated with sunlight exposure (6). Photosensitivity is a well-recognized clinical hallmark of SLE, and UVR has been proposed to contribute to disease pathogenesis through mechanisms involving keratinocyte apoptosis, autoantigen release, and type I interferon activation (7–9). However, previous studies have predominantly focused on UV-induced photosensitivity, cutaneous manifestations, and disease activity among patients with established SLE (10, 11). Although ultraviolet exposure has been implicated as an environmental factor contributing to SLE susceptibility, evidence from prospective population-based studies evaluating habitual sunlight exposure in relation to incident SLE remains limited.
Beyond SLE, the role of sunlight exposure in other rheumatic diseases has received comparatively less attention. Rheumatic diseases encompass diverse immunopathological processes, and the biological effects of the same environmental exposure may vary depending on the underlying disease mechanisms. Rheumatoid arthritis (RA) and gout represent two distinct inflammatory models with different immunological foundations: RA is characterized by autoimmune-mediated adaptive immune dysregulation (12), whereas gout is primarily driven by innate immune activation triggered by monosodium urate crystal deposition (13). However, whether sunlight exposure influences the development of these diseases, and whether such effects differ from those observed in SLE, remains largely unknown.
Therefore, prospective studies evaluating multiple rheumatic diseases within a unified population framework are needed. Using data from the UK Biobank, we conducted a prospective cohort study to investigate the associations between sunlight exposure and incident SLE, RA, and gout. By comparing diseases with distinct immunological backgrounds within the same cohort, we aimed to determine whether sunlight exposure exhibits disease-specific associations across rheumatic disorders.
Methods
Study design
Data for this study were derived from the UK Biobank, a nationwide prospective cohort that enrolled approximately 500,000 adults from across the United Kingdom between 2006 and 2010. The cohort primarily comprised individuals aged 40 years and older at baseline, who were followed longitudinally through linked health records until death, withdrawal, or the end of follow-up. Extensive information was systematically collected through touchscreen questionnaires, nurse-led interviews, physical examinations, biological specimen collection, imaging assessments, and linkage to national health registries, thereby providing a rich resource for population-based research. Ethical approval was granted by the North West Multi-Centre Research Ethics Committee (number: 21/NW/0157), and all participants provided written informed consent.
Assessment of sunlight exposure
Sunlight exposure was assessed based on self-reported daily time spent outdoors during typical summer and winter days. Data were collected at baseline (2006–2010) through touchscreen questionnaires, corresponding to fields 1050 (summer) and 1060 (winter) in the UK Biobank database. Participants were asked the question: “In a typical day in summer (winter), how many hours do you spend outdoors?” If the amount of time spent outdoors in summer (winter) varied substantially, participants were instructed to report the average daily time. For example, if a participant spent 1 h per day on weekdays and 4 h per day on weekends, the total weekly duration would be 13 h (5 + 8), corresponding to an average of approximately 2 h per day. The touchscreen questionnaire included no items on occupational or recreational sun exposure and no personal ultraviolet dosimetry was performed, so the exposure analyzed here represents self-reported time spent outdoors rather than a measured ultraviolet dose.
During data preprocessing, responses of missing or ambiguous values in summer (53,438) or winter (112,247) were removed, respectively. The average daily sunlight exposure duration was calculated as the mean of the reported summer and winter durations. To minimize the influence of extreme values on the analysis, participants with average daily sunlight exposure above the 95th percentile or below the 5th percentile were excluded.
Covariates
A series of covariates were selected into our analyses, including age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), Body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No). When performing regression analyses, missing values in continuous variables were imputed using the mean, whereas missing values in categorical variables were imputed using the mode. Use of sun protection was ascertained with a single baseline question and cut as a binary variable (yes versus no), which captures neither its type nor its frequency or adequacy.
Outcome definition
Incident rheumatoid arthritis, systemic lupus erythematosus, and gout were ascertained from linked hospital admission records in the UK Biobank using International Classification of Diseases, Tenth Revision (ICD-10) codes. According to ICD-10, systemic lupus erythematosus was defined using 4 codes; rheumatoid arthritis was identified using 15 codes; gout was ascertained using 38 codes (Supplementary Table S1). Dates of death were obtained from the Office for National Statistics mortality register. Follow-up extended from baseline to the earliest of outcome diagnosis, death, or loss to follow-up and was censored on October 31, 2022. A detailed flow of participant inclusion and exclusion is presented in Figure 1.
Figure 1.
Participant selection flow for outdoor light exposure and rheumatic diseases. (a) Systemic lupus erythematosus (SLE); (b) rheumatoid arthritis (RA); (c) gout.
Statistical analysis
Baseline characteristics by sunlight exposure levels were summarized as means (SD) for continuous variables and counts (percentages) for categorical variables, with group differences tested using analysis of variance or χ2 tests as appropriate. Sunlight exposure duration was trichotomized at the median (< 2.5, 2.5 and > 2.5 h/day). Multivariate Cox proportional hazard regression models were used to examine the associations of sunlight exposure with systemic lupus erythematosus, rheumatoid arthritis and gout. The estimates using hazard ratios (HRs) and the corresponding 95% confidence intervals (95% CIs) compared with regular sunlight exposure as 2.5 h (median values). To assess the potential impact of unmeasured confounding, we calculated E-values, which quantify the minimum strength of association that an unmeasured confounder would need to fully explain away the observed exposure–outcome association. Potential non-linear associations between sunlight exposure and risk of diseases were evaluated using restricted cubic splines with three knots to assess potential nonlinearity based on minimum Akaike information criterion. Subgroup analyses were conducted to examine the consistency of the associations across background or lifestyle strata. Participants were stratified by age (< 65 years and ≥ 65 years), sex (male and female), BMI (< 24.0, 24.0 ~ 27.9, and ≥ 28.0), smoking (yes and no), alcohol consumption (yes and no), physical activity (yes and no), and sleep duration (≤ 7, > 7), and interaction terms between the exposure and subgroup variables were included in the models to formally test for effect modification. Stratified hazard ratios and 95% confidence intervals were estimated within each subgroup. Statistical significance for interaction was assessed using the likelihood ratio test, comparing models with and without interaction terms.
Certain sensitivity analyses were conducted. First, we also conducted Cox models to investigate the associations of sunlight exposure with rheumatic diseases during summer (median: 4 h) or winter (median: 2 h) season, respectively. Second, their associations were examined using inverse probability weighting. Inverse probability weighting was used to address potential baseline confounding. Stabilized weights were derived from logistic models of exposure on covariates, creating a pseudo-population with balanced confounders across exposure groups. Weighted analyses provided effect estimates under the assumption of no unmeasured confounding. Third, as a novel causal inference approach, Cox models with inverse probability weighting were used to estimate the causal effect of average sunlight exposure on disease onset using average treatment effect (ATE) and average treatment effect on the treated (ATT). Weighted Cox models provided marginal effect estimates with hazard ratios and 95% confidence intervals. Fourth, we excluded non-white population, or participants with self-reported poor health or missing covariates. Fifth, we examined the association between average outdoor activity duration and serum 25-hydroxyvitamin D [25(OH)D] concentrations. We further assessed whether the association between average outdoor activity duration and rheumatic disease risk differed across strata of serum 25(OH)D concentrations. In addition, we compared the incidence proportions of SLE, RA, and gout between White and non-White participants.
Results
A total of 363,873 participants with available sunlight exposure data were included in the analyses. Of these, 138,171 had outdoor sunlight exposure of less than 2.5 h per day, 46,257 had approximately 2.5 h per day, and 179,445 had more than 2.5 h per day. For age distribution, 288,557 participants (79.3%) were younger than 65 years and 75,316 (20.7%) were 65 years or older. For sex distribution, 193,627 participants (53.2%) were female and 170,246 (46.8%) were male. Most participants were White (346,126 [95.1%]). Compared with participants reporting shorter outdoor time, those reporting more than 2.5 h/day were older, more likely to be male, and less likely to have a college education. We found most participants had a white skin color (272,220 in all population) and used sun exposure protection (326,601 in all population) (Table 1).
Table 1.
Baseline characteristics from 363,873 participants.
| Variable name | Category levels | Time spent in outdoor light (average) | p | |||
|---|---|---|---|---|---|---|
| Overall | < 2.5 h | 2.5 h | > 2.5 h | |||
| N = 363,873 | N = 138,171 | N = 46,257 | N = 179,445 | |||
| Age, years, N (%) | <65 | 288,557 (79.30) | 119,614 (86.57) | 37,074 (80.15) | 131,869 (73.49) | <0.001 |
| ≥ 65 | 75,316 (20.70) | 18,557 (13.43) | 9,183 (19.85) | 47,576 (26.51) | ||
| Sex, N (%) | Female | 193,627 (53.21) | 78,031 (56.47) | 26,656 (57.63) | 88,940 (49.56) | <0.001 |
| Male | 170,246 (46.79) | 60,140 (43.53) | 19,601 (42.37) | 90,505 (50.44) | ||
| Ethnicity, N (%) | Non-White | 16,649 (4.58) | 6,731 (4.87) | 1922 (4.16) | 7,996 (4.46) | <0.001 |
| White | 346,126 (95.12) | 131,047 (94.84) | 44,184 (95.52) | 170,895 (95.24) | ||
| Missing | 1,098 (0.30) | 393 (0.28) | 151 (0.33) | 554 (0.31) | ||
| Educational levels, N (%) | College | 116,672 (32.06) | 56,886 (41.17) | 15,817 (34.19) | 43,969 (24.50) | <0.001 |
| Other levels | 182,828 (50.25) | 66,357 (48.03) | 23,556 (50.92) | 92,915 (51.78) | ||
| Unknown | 64,373 (17.69) | 14,928 (10.80) | 6,884 (14.88) | 42,561 (23.72) | ||
| Body mass index, kg/m2, N (%) | < 24 | 120,458 (33.10) | 49,435 (35.78) | 16,026 (34.65) | 54,997 (30.65) | <0.001 |
| 24–27.9 | 156,454 (43.00) | 57,101 (41.33) | 19,553 (42.27) | 79,800 (44.47) | ||
| ≥ 28 | 85,239 (23.43) | 30,983 (22.42) | 10,469 (22.63) | 43,787 (24.40) | ||
| Missing | 1722 (0.47) | 652 (0.47) | 209 (0.45) | 861 (0.48) | ||
| Physical activity, N (%) | Yes | 231,576 (63.64) | 78,818 (57.04) | 28,762 (62.18) | 123,996 (69.10) | <0.001 |
| No | 127,198 (34.96) | 58,023 (41.99) | 16,943 (36.63) | 52,232 (29.11) | ||
| Missing | 5,099 (1.40) | 1,330 (0.96) | 552 (1.19) | 3,217 (1.79) | ||
| Drinking status, N (%) | Yes | 349,019 (95.92) | 132,438 (95.85) | 44,558 (96.33) | 172,023 (95.86) | <0.001 |
| No | 14,553 (4.00) | 5,630 (4.07) | 1,663 (3.60) | 7,260 (4.05) | ||
| Missing | 301 (0.08) | 103 (0.07) | 36 (0.08) | 162 (0.09) | ||
| Smoking status, N (%) | Yes | 166,492 (45.76) | 57,847 (41.87) | 20,674 (44.69) | 87,971 (49.02) | <0.001 |
| No | 196,208 (53.92) | 79,997 (57.90) | 25,465 (55.05) | 90,746 (50.57) | ||
| Missing | 1,173 (0.32) | 327 (0.24) | 118 (0.26) | 728 (0.41) | ||
| Household income, £, N (%) | < 31,000 | 154,560 (42.48) | 47,548 (34.41) | 18,803 (40.65) | 88,209 (49.16) | <0.001 |
| ≥ 31,000 | 160,445 (44.09) | 75,923 (54.95) | 21,489 (46.46) | 63,033 (35.13) | ||
| Missing | 48,868 (13.43) | 14,700 (10.64) | 5,965 (12.90) | 28,203 (15.72) | ||
| Dieting score, N (%) | 0–3 | 252,213 (69.31) | 98,034 (70.95) | 32,086 (69.36) | 122,093 (68.04) | <0.001 |
| 4–7 | 111,646 (30.68) | 40,133 (29.05) | 14,168 (30.63) | 57,345 (31.96) | ||
| Missing | 14 (0.00) | 4 (0.00) | 3 (0.01) | 7 (0.00) | ||
| Sleep duration, hours, N (%) | ≤7 | 226,762 (62.32) | 91,448 (66.18) | 28,942 (62.57) | 106,372 (59.28) | <0.001 |
| >7 | 136,117 (37.41) | 46,421 (33.60) | 17,208 (37.20) | 72,488 (40.40) | ||
| Missing | 994 (0.27) | 302 (0.22) | 107 (0.23) | 585 (0.33) | ||
| Ambient PM2.5 levels, μg/m3, mean (SD) | 9.98 (1.01) | 10.01 (1.02) | 9.95 (0.99) | 9.95 (1.01) | <0.001 | |
| Skin color, N (%) | Other | 87,602 (24.07) | 30,713 (22.23) | 11,030 (23.85) | 45,859 (25.56) | <0.001 |
| White | 272,220 (74.81) | 106,229 (76.88) | 34,723 (75.07) | 131,268 (73.15) | ||
| Missing | 4,051 (1.11) | 1,229 (0.89) | 504 (1.09) | 2,318 (1.29) | ||
| Use of sun exposure protection, N (%) | No | 35,201 (9.67) | 12,728 (9.21) | 3,835 (8.29) | 18,638 (10.39) | <0.001 |
| Yes | 326,601 (89.76) | 124,375 (90.02) | 42,226 (91.29) | 160,000 (89.16) | ||
| Missing | 2071 (0.57) | 1,068 (0.77) | 196 (0.42) | 807 (0.45) | ||
During the follow-up between 2006 and 2022, a total of 335 incident cases of systemic lupus erythematosus, 5,591 cases of rheumatoid arthritis, and 7,485 cases of gout were documented (Table 2). Compared to sunlight exposure at 2.5 h, the HRs and 95% CIs under 2.5 h were 1.71 (95% CIs: 1.13, 2.59) for systemic lupus erythematosus, 1.05 (95% CIs: 0.96, 1.14) for rheumatoid arthritis and 0.96 (95% CIs: 0.88, 1.04) for gout; the HRs and 95% CIs above 2.5 h were 1.75 (95% CIs: 1.17, 2.63) for systemic lupus erythematosus, 1.15 (95% CIs:1.06, 1.26) for rheumatoid arthritis and 1.09 (95% CIs: 1.02, 1.18) for gout. The observed E-values ranged from 1.25 to 2.81 for participants with less than 2.5 h of outdoor sunlight exposure and from 1.40 to 2.89 for those with more than 2.5 h, indicating that only unmeasured confounders of substantial strength could fully explain these associations.
Table 2.
Associations of sunlight exposure with systemic lupus erythematosus, rheumatoid arthritis, and gout in the UK.
| Outdoor duration (hours) | Cases/Person-years | HR (95% CI) | p value | E value |
|---|---|---|---|---|
| Systemic lupus erythematosus, SLE | ||||
| 2.5 | 27/616357 | 1 | NA | NA |
| < 2.5 | 133/1847551 | 1.71 (1.13–2.59) | 0.011 | 2.81 |
| > 2.5 | 175/2370470 | 1.75 (1.17–2.63) | 0.007 | 2.89 |
| Rheumatoid arthritis, RA | ||||
| 2.5 | 645/610485 | 1 | NA | NA |
| < 2.5 | 1836/1831941 | 1.05 (0.96–1.14) | 0.333 | 1.28 |
| > 2.5 | 3110/2343037 | 1.15 (1.06–1.26) | 0.001 | 1.57 |
| Gout | ||||
| 2.5 | 827/611821 | 1 | NA | NA |
| < 2.5 | 2194/1836420 | 0.96 (0.88–1.04) | 0.304 | 1.25 |
| > 2.5 | 4464/2345562 | 1.09 (1.02–1.18) | 0.018 | 1.40 |
Models were adjusted for age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), Body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No). Bold values indicate statistically significant associations (p < 0.05).
We used restricted cubic splines to fit the nonlinearity of sunlight exposure time and rheumatic diseases (Figure 2). No significant nonlinearity was detected for SLE, RA, or gout (p for nonlinearity > 0.05).
Figure 2.
Restricted cubic spline analysis of outdoor light exposure and rheumatic disease risk. (a) Systemic lupus erythematosus (SLE); (b) rheumatoid arthritis (RA); (c) gout. Restricted cubic spline regression was used. Models were adjusted for age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No).
In the subgroup analyses, several nominal interactions were observed before correction (smoking in SLE, p for interaction = 0.025, Table 3; age and BMI in RA, p = 0.043 and p < 0.01, Table 4). However, after Bonferroni correction these were no longer significant and should be regarded as exploratory. Only the interaction for age in gout remained significant after correction (adjusted p = 0.014), with a higher risk of gout associated with longer sunlight exposure (> 2.5 h) among participants aged ≥ 65 years (Table 5).
Table 3.
Subgroup analysis of sunlight exposure and systemic lupus erythematosus.
| Systemic lupus erythematosus, SLE | Cases | Time spent in outdoor light (average) | |||||
|---|---|---|---|---|---|---|---|
| 2.5 h | < 2.5 h | > 2.5 h | p for interaction (Adjusted p value) | ||||
| Reference | HR | p | HR | p | |||
| Age, years | 0.133 (0.931) | ||||||
| < 65 | 267 | 1 | 1.86 (1.14–3.03) | 0.013 | 2.09 (1.29–3.38) | 0.003 | |
| ≥ 65 | 68 | 1 | 1.46 (0.65–3.27) | 0.357 | 1.00 (0.46–2.15) | 0.991 | |
| Sex | 0.236 (1.000) | ||||||
| Male | 60 | 1 | 1.87 (0.54–6.44) | 0.321 | 2.66 (0.82–8.66) | 0.103 | |
| Female | 275 | 1 | 1.69 (1.09–2.62) | 0.020 | 1.60 (1.04–2.48) | 0.034 | |
| Body mass index (kg/m2) | 0.578 (1.000) | ||||||
| <24 | 128 | 1 | 2.17 (1.03–4.58) | 0.042 | 2.57 (1.23–5.36) | 0.012 | |
| 24.0–27.9 | 117 | 1 | 1.43 (0.74–2.79) | 0.287 | 1.45 (0.76–2.75) | 0.262 | |
| ≥ 28.0 | 90 | 1 | 1.66 (0.78–3.57) | 0.191 | 1.38 (0.65–2.95) | 0.404 | |
| Smoking | 0.025 (0.175) | ||||||
| Yes | 171 | 1 | 1.07 (0.63–1.83) | 0.807 | 1.41 (0.85–2.33) | 0.183 | |
| No | 164 | 1 | 2.98 (1.49–5.94) | 0.002 | 2.40 (1.20–4.80) | 0.013 | |
| Drinking | 0.941 (1.000) | ||||||
| Yes | 316 | 1 | 1.69 (1.11–2.58) | 0.015 | 1.72 (1.13–2.60) | 0.011 | |
| No | 19 | 1 | 2.49 (0.31–20.01) | 0.39 | 2.50 (0.32–19.62) | 0.384 | |
| Physical activity | 0.922 (1.000) | ||||||
| Yes | 205 | 1 | 1.64 (0.96–2.78) | 0.069 | 1.65 (0.99–2.74) | 0.056 | |
| No | 130 | 1 | 1.85 (0.94–3.61) | 0.073 | 1.93 (0.99–3.78) | 0.055 | |
| Sleep duration, hours | 0.483 (1.000) | ||||||
| ≤7 | 206 | 1 | 1.69 (0.99–2.89) | 0.056 | 1.93 (1.14–3.27) | 0.014 | |
| > 7 | 129 | 1 | 1.78 (0.93–3.41) | 0.084 | 1.50 (0.79–2.86) | 0.213 | |
Models were adjusted for age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), Body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No) except for the stratifying variable. Bold values indicate statistically significant associations (p < 0.05).
Table 4.
Subgroup analysis of sunlight exposure and rheumatoid arthritis.
| Rheumatoid arthritis, RA | Cases | Time spent in outdoor light (average) | p for interaction | ||||
|---|---|---|---|---|---|---|---|
| 2.5 h | < 2.5 h | > 2.5 h | |||||
| Reference | HR | p value | HR | p value | (Adjusted p value) | ||
| Age, years | 0.043 (0.301) | ||||||
| < 65 | 3,852 | 1 | 1.07 (0.96–1.19) | 0.236 | 1.22 (1.10–1.35) | < 0.001 | |
| ≥ 65 | 1739 | 1 | 1.02 (0.87–1.21) | 0.773 | 1.03 (0.88–1.19) | 0.738 | |
| Sex | 0.327 (1.000) | ||||||
| Male | 1888 | 1 | 1.05 (0.89–1.25) | 0.566 | 1.18 (1.01–1.38) | 0.038 | |
| Female | 3,703 | 1 | 1.04 (0.94–1.16) | 0.445 | 1.13 (1.02–1.26) | 0.016 | |
| Body mass index (kg/m2) | 0.008 (0.056) | ||||||
| <24 | 1,504 | 1 | 0.85 (0.72–1.00) | 0.055 | 1.03 (0.88–1.20) | 0.742 | |
| 24.0–27.9 | 2,302 | 1 | 1.18 (1.02–1.37) | 0.024 | 1.26 (1.10–1.45) | 0.001 | |
| ≥ 28.0 | 1785 | 1 | 1.08 (0.92–1.26) | 0.369 | 1.15 (0.99–1.34) | 0.076 | |
| Smoking | 0.224 (1.000) | ||||||
| Yes | 3,050 | 1 | 1.10 (0.97–1.25) | 0.147 | 1.22 (1.08–1.37) | 0.001 | |
| No | 2,541 | 1 | 0.99 (0.87–1.13) | 0.904 | 1.08 (0.96–1.23) | 0.203 | |
| Drinking | 0.347 (1.000) | ||||||
| Yes | 5,236 | 1 | 1.04 (0.95–1.14) | 0.385 | 1.16 (1.06–1.27) | 0.001 | |
| No | 355 | 1 | 1.10 (0.77–1.57) | 0.619 | 1.05 (0.74–1.49) | 0.785 | |
| Physical activity | 0.212 (1.000) | ||||||
| Yes | 3,466 | 1 | 1.01 (0.89–1.13) | 0.927 | 1.16 (1.04–1.30) | 0.007 | |
| No | 2,125 | 1 | 1.10 (0.95–1.26) | 0.201 | 1.13 (0.99–1.30) | 0.070 | |
| Sleep duration, hours | 0.133 (0.931) | ||||||
| ≤7 | 3,459 | 1 | 0.98 (0.87–1.09) | 0.700 | 1.10 (0.99–1.23) | 0.072 | |
| > 7 | 2,132 | 1 | 1.18 (1.01–1.38) | 0.032 | 1.25 (1.08–1.44) | 0.002 | |
Models were adjusted for age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), Body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No) except for the stratifying variable. Bold values indicate statistically significant associations (p < 0.05).
Table 5.
Subgroup analysis of sunlight exposure and gout.
| Gout | Cases | Time spent in outdoor light (average) | p for interaction (Adjusted p value) | ||||
|---|---|---|---|---|---|---|---|
| 2.5 h | < 2.5 h | > 2.5 h | |||||
| Reference | HR | p value | HR | p value | |||
| Age, y | 0.002 (0.014) | ||||||
| < 65 | 4,760 | 1 | 0.98 (0.85–1.12) | 0.731 | 1.03 (0.91–1.17) | 0.591 | |
| ≥ 65 | 2,725 | 1 | 0.97 (0.88–1.07) | 0.511 | 1.14 (1.04–1.25) | 0.007 | |
| Sex | 0.055 (0.385) | ||||||
| Male | 6,264 | 1 | 0.99 (0.91–1.09) | 0.855 | 1.14 (1.04–1.23) | 0.003 | |
| Female | 1,221 | 1 | 0.83 (0.70–1.00) | 0.045 | 0.97 (0.82–1.14) | 0.688 | |
| Body mass index | 0.096 (0.672) | ||||||
| <24 | 792 | 1 | 0.90 (0.71–1.14) | 0.393 | 1.02 (0.82–1.28) | 0.830 | |
| 24.0–27.9 | 3,292 | 1 | 0.95 (0.84–1.08) | 0.465 | 1.15 (1.03–1.29) | 0.017 | |
| ≥ 28.0 | 3,401 | 1 | 0.98 (0.87–1.10) | 0.683 | 1.06 (0.95–1.18) | 0.309 | |
| Smoking | 0.904 (1.000) | ||||||
| Yes | 4,466 | 1 | 0.96 (0.86–1.06) | 0.410 | 1.10 (1.00–1.21) | 0.060 | |
| No | 3,019 | 1 | 0.96 (0.85–1.09) | 0.543 | 1.09 (0.97–1.22) | 0.161 | |
| Drinking | 0.959 (1.000) | ||||||
| Yes | 7,287 | 1 | 0.96 (0.88–1.04) | 0.286 | 1.09 (1.01–1.18) | 0.026 | |
| No | 198 | 1 | 1.05 (0.63–1.75) | 0.863 | 1.28 (0.79–2.09) | 0.316 | |
| Physical activity | 0.566 (1.000) | ||||||
| Yes | 4,608 | 1 | 0.99 (0.89–1.11) | 0.879 | 1.13 (1.03–1.25) | 0.012 | |
| No | 2,877 | 1 | 0.92 (0.81–1.03) | 0.155 | 1.04 (0.93–1.17) | 0.499 | |
| Sleep duration, h | 0.081 (0.567) | ||||||
| ≤7 | 4,462 | 1 | 1.00 (0.90–1.11) | 0.980 | 1.16 (1.05–1.28) | 0.003 | |
| > 7 | 3,023 | 1 | 0.91 (0.80–1.03) | 0.124 | 1.00 (0.89–1.13) | 0.953 | |
Models were adjusted for age (< 65 years and ≥ 65 years), sex (Male and Female), ethnicity (White and Non-white), educational levels (College and Other levels), Body mass index (BMI) (kg/m2) (< 24, 24 ~ 27.9, ≥ 28), alcohol consumption (Yes and No), smoking status (Yes and No), household income (< £ 31,000 and ≥ £ 31,000), physical activity (Yes and No), dieting score (ranged from 0–7), sleep duration (≤ 7 h and > 7 h), ambient PM2.5 levels (μg/m3), skin color (White and Other), use of sun exposure protection (Yes and No) except for the stratifying variable. Bold values indicate statistically significant associations (p < 0.05).
A series of sensitivity analyses were conducted. We observed the associations of sunlight exposure with systemic lupus erythematosus and gout in summer (Supplementary Table S2), as well as with gout in winter (Supplementary Table S3). The associations of sunlight exposure with systemic lupus erythematosus, rheumatoid arthritis, and gout remained consistent after using inverse probability weighting (Supplementary Table S4), only including white population (Supplementary Table S5), excluding poor health (Supplementary Table S6) and missing covariates (Supplementary Table S7). Compared with 2.5 h of outdoor exposure, < 2.5 h was associated with lower odds, while >2.5 h was associated with higher odds of above-mean serum 25(OH)D concentrations (Supplementary Table S8). In analyses stratified by serum 25(OH)D levels, we observed a significant interaction for SLE after Bonferroni correction, although the stratum-specific estimates were imprecise. No significant interaction was observed for RA or gout after correction (Supplementary Table S9). The incidence proportions of SLE and RA were numerically higher among non-White participants than among White participants, whereas the incidence proportion of gout was similar between the two groups (Supplementary Table S10).
Discussion
In this large prospective cohort, sunlight exposure showed distinct associations with incident SLE, RA, and gout. Both shorter and longer outdoor time were associated with higher SLE risk relative to 2.5 h/day, whereas longer outdoor time was associated with modestly increased risks of RA and gout. These differences suggest that sunlight-related exposures may interact with disease-specific pathogenic backgrounds, although the modest effect sizes for RA and gout and the observational nature of the study warrant cautious interpretation.
Among rheumatic diseases, SLE represents the condition most strongly associated with sunlight exposure (6). Photosensitivity is a well-recognized feature of SLE, and ultraviolet radiation has been implicated in disease pathogenesis through keratinocyte apoptosis, autoantigen release, and type I interferon activation (8, 10). Although ultraviolet exposure has been widely implicated in SLE susceptibility, previous studies have mainly focused on geographic variation, environmental associations, or disease manifestations among patients with established SLE (3, 14, 15). However, prospective evidence linking habitual sunlight exposure to incident SLE remains limited. In our study, both shorter and longer outdoor time were associated with higher SLE risk, suggesting that the effects of sunlight exposure may vary according to exposure patterns and host susceptibility. This finding may reflect the context-dependent immunological effects of ultraviolet radiation.
Previous studies have suggested that the relationship between sunlight exposure and RA risk is less well established and more heterogeneous than that observed in SLE. Prospective evidence has indicated a potential inverse association between ambient UV-B exposure and RA risk, which may be partly attributable to vitamin D-mediated immunoregulation (16, 17). However, this association has not been consistently reproduced across populations, and population-based analyses have not identified a clear relationship between sunlight exposure and ACPA positivity (18). In our study, longer outdoor time was associated with a modestly increased risk of incident RA, suggesting that the biological impact of sunlight exposure on RA development may reflect a complex interplay among vitamin D-related pathways, UV-mediated immune responses, and other environmental determinants.
Compared with SLE and RA, the role of environmental factors in gout has received relatively less attention, with previous studies mainly focusing on metabolic and toxic exposures (19–21). Gout is characterized by urate metabolism dysregulation and monosodium urate crystal-induced innate immune activation (22). Although environmental exposures such as lead have been linked to gout risk (23, 24), the potential contribution of sunlight exposure to gout development remains poorly understood. In our study, longer outdoor time was associated with increased gout risk, particularly among older participants, suggesting that age-related metabolic susceptibility may influence the relationship between sunlight exposure and gout risk.
These cut-points also warrant external context. The median of 2.5 h/day (4 h/day in summer and 2 h/day in winter) is comparable with other analyses of the same UK Biobank item (25) and is not low by general-population standards, since US respondents reported spending only about 7.6% of the day, roughly 1.8 h, outdoors (26). Time outdoors is nevertheless not equivalent to ultraviolet dose: the United Kingdom lies between about 50° N and 58° N, and vitamin D-effective ultraviolet-B doses differ up to five-fold across Europe (27), so 2.5 h here delivers considerably less ultraviolet radiation than the same duration at lower latitudes. Disease burden varies geographically in parallel, with SLE reported more frequently in high-income countries (28) and gout driven largely by aging and metabolic factors (29). Consistent with these being distinct exposure metrics, studies of ambient ultraviolet-B reported no association with SLE (15) and an inverse association with RA (16). Our median-based cut-points are therefore population-specific and should not be transferred to higher-ultraviolet settings.
The methodological strengths of this study include its large-scale cohort spanning 2006 to 2022 and the use of multiple statistical approaches. Furthermore, this research provides a systematic comparison of three distinct rheumatic diseases within a unified cohort, offering unique insights into disease-specific variations. Several methodological limitations warrant consideration. The most important is the reliance on self-reported sunlight exposure, which is a major limitation in its own right and not merely a matter of unresolved wavelengths. Recall of a “typical” summer and winter day at a single baseline visit is subject to recall error, rounding and social-desirability bias, and cannot capture time of day, day-to-day variability, or shade and clothing use; self-reported exposure agrees only fair-to-moderately with personal ultraviolet dosimetry, with correlations of approximately 0.3 to 0.5 (30, 31). Non-differential error of this kind would attenuate hazard ratios toward the null and may contribute to the modest estimates for RA and gout (32), whereas reporting related to prodromal symptoms could bias estimates in either direction. UK Biobank also collected no items on occupational or recreational sun exposure, and neither validated instruments such as the Sun Exposure and Protection Index (33) nor occupational job-exposure matrices (34) were available to us; future cohorts should embed these and, where feasible, calibrate them against personal dosimetry. A second major limitation is that sun protection was ascertained with a single baseline question and modeled as a binary variable, whereas it is a time-varying habit that is difficult to quantify. A yes/no indicator distinguishes neither the modality of protection nor its frequency or adequacy of application; sun protection factors are determined at an application thickness of 2 mg/cm2, whereas real-world application is typically only 0.4 to 1.0 mg/cm2, so the protection achieved is well below the labeled value (35). Protective behavior may also change once symptoms appear. Residual confounding by sun protection therefore probably persists, and our E-values (1.25 to 2.89), while informative, do not remove this concern. The predominantly White UK population and distinctive spectral characteristics at high latitudes limit generalizability to other geographic and ethnic contexts. Additionally, despite extensive covariate adjustment, residual confounding from unmeasured factors such as genetic susceptibility and skin barrier function variations may persist. Our findings carry important implications for public health practice.
Despite these limitations, our study suggests that sunlight exposure should not be viewed simply as universally harmful or beneficial. Rather, its health effects may depend on the underlying disease context. In SLE, where photosensitivity is a well-recognized clinical feature, our findings support the continued emphasis on moderate sunlight exposure and appropriate sun protection. Short outdoor activities may be considered in the early morning or late afternoon, when UV intensity is relatively low and red or near-infrared light is more abundant. In contrast, for patients with RA or gout, the modest associations we observed do not support excessive restriction of outdoor activity. A more balanced approach, combining regular outdoor activity with reasonable sun protection, may be more appropriate. Future studies using objective light-monitoring tools could help distinguish the effects of different wavelengths and exposure durations, and could provide a stronger basis for individualized sunlight exposure guidance in patients with rheumatic diseases.
Conclusion
In this large prospective cohort, sunlight exposure was associated with the risks of SLE, RA, and gout in a disease-specific manner. These findings suggest that sunlight exposure should not be considered a uniform environmental factor across rheumatic diseases, but rather one whose health implications may depend on the underlying disease context. Further studies using objective exposure assessment and more diverse populations are needed to clarify these associations and to support more tailored guidance on sunlight exposure and sun protection.
Acknowledgments
We also thank the UK Biobank participants and the UK Biobank team for their work in collecting, processing and disseminating the data.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This study was supported by NATCM Initiative for Strengthening TCM Evidence-Based Research, NATCM-STER, the Major and Difficult Diseases Integrated Traditional Chinese and Western Medicine Clinical Collaboration Project (ZDYN-2024-A-151), the Chongqing Key Special Project for Technological Innovation and Application Development in Traditional Chinese Medicine for Prevention and Treatment of Major Difficult Chronic Diseases (Approval No.CSTB2024TIAD-KPX0032 and No.CSTB2024TIAD-KPX0040), and the Chongqing Medical Scientific Research Tackling Project(Joint Tackling Project of Chongqing Health Commission and Science and Technology Bureau) (No. 2025GGXM003 and 2026ZYZD001) and the Science and Technology Research Program of Chongqing Municipal Education Commission (No. KJQN202415126).
Footnotes
Edited by: Shuo-Yan Gau, Charité University Medicine Berlin, Germany
Reviewed by: Ana Valle, Brigham and Women's Hospital, United States
Rana Salieva, Osh State University, Kyrgyzstan
Data availability statement
The data used in the present study are available from the UK Biobank with restrictions applied. Access to the UK Biobank data can be requested by means of a standard protocol (https://www.ukbiobank.ac.uk/register-apply/). Requests to access these datasets should be directed to https://www.ukbiobank.ac.uk/register-apply/.
Ethics statement
The UK Biobank study received ethical approval from the North West Multi-centre Research Ethics Committee, and all participants provided written informed consent at recruitment. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
YP: Formal analysis, Writing – original draft. JC: Writing – original draft. LZ: Writing – review & editing. YaZ: Writing – original draft. YiZ: Writing – review & editing. LD: Writing – review & editing. YG: Writing – review & editing. LC: Writing – review & editing. CL: Writing – review & editing. BW: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2026.1873382/full#supplementary-material
References
- 1.Bernard JJ, Gallo RL, Krutmann J. Photoimmunology: how ultraviolet radiation affects the immune system. Nat Rev Immunol. (2019) 19:688–701. doi: 10.1038/s41577-019-0185-9, [DOI] [PubMed] [Google Scholar]
- 2.Ullrich SE, Byrne SN. The immunologic revolution: Photoimmunology. J Invest Dermatol. (2012) 132:896–905. doi: 10.1038/jid.2011.405, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ponsonby A-L, McMichael A, van der Mei I. Ultraviolet radiation and autoimmune disease: insights from epidemiological research. Toxicology. (2002) 181-182:71–8. doi: 10.1016/s0300-483x(02)00257-3, [DOI] [PubMed] [Google Scholar]
- 4.Hart PH. Sun-Immune Connection. Nat Rev Immunol. (2019) 19:661. doi: 10.1038/s41577-019-0216-6, [DOI] [PubMed] [Google Scholar]
- 5.Halliday GM, Damian DL, Rana S, Byrne SN. The suppressive effects of ultraviolet radiation on immunity in the skin and internal organs: implications for autoimmunity. J Dermatol Sci. (2012) 66:176–82. doi: 10.1016/j.jdermsci.2011.12.009, [DOI] [PubMed] [Google Scholar]
- 6.Estadt SN, Maz MP, Musai J, Kahlenberg JM. Mechanisms of photosensitivity in autoimmunity. J Invest Dermatol. (2022) 142:849–56. doi: 10.1016/j.jid.2021.05.007, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chen J, Liao S, Pang W, Guo F, Yang L, Liu H-F, et al. Life factors acting on systemic lupus erythematosus. Front Immunol. (2022) 13:986239. doi: 10.3389/fimmu.2022.986239, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wolf SJ, Estadt SN, Gudjonsson JE, Kahlenberg JM. Human and murine evidence for mechanisms driving autoimmune photosensitivity. Front Immunol. (2018) 9:2430. doi: 10.3389/fimmu.2018.02430, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.He J, Guo Y, Chen J, Xu J, Zhu X. Exploring the correlation between UVB sensitivity and SLE activity: insights into UVB-driven pathogenesis in lupus erythematosus. J Autoimmun. (2025) 153:103393. doi: 10.1016/j.jaut.2025.103393, [DOI] [PubMed] [Google Scholar]
- 10.Kuhn A, Wenzel J, Weyd H. Photosensitivity, apoptosis, and cytokines in the pathogenesis of lupus erythematosus: a critical review. Clin Rev Allergy Immunol. (2014) 47:148–62. doi: 10.1007/s12016-013-8403-x, [DOI] [PubMed] [Google Scholar]
- 11.Woo JMP, Parks CG, Jacobsen S, Costenbader KH, Bernatsky S. The role of environmental exposures and gene-environment interactions in the etiology of systemic lupus erythematous. J Intern Med. (2022) 291:755–78. doi: 10.1111/joim.13448, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.McInnes IB, Schett G. The pathogenesis of rheumatoid arthritis. N Engl J Med. (2011) 365:2205–19. doi: 10.1056/NEJMra1004965, [DOI] [PubMed] [Google Scholar]
- 13.Dalbeth N, Gosling AL, Gaffo A, Abhishek A. Gout. Lancet. (2021) 397:1843–55. doi: 10.1016/S0140-6736(21)00569-9, [DOI] [PubMed] [Google Scholar]
- 14.Kumar M, Yip L, Wang F, Marty S-E, Fathman CG. Autoimmune disease: genetic susceptibility, environmental triggers, and immune dysregulation. Where can we develop therapies? Front Immunol. (2025) 16:1626082. doi: 10.3389/fimmu.2025.1626082, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Barbhaiya M, Hart JE, Malspeis S, Tedeschi SK, VoPham T, Sparks JA, et al. Association of Ultraviolet B Radiation and Risk of systemic lupus erythematosus among women in the nurses’ health studies. Arthritis Care Res. (2023) 75:1409–15. doi: 10.1002/acr.24974, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Arkema EV, Hart JE, Bertrand KA, Laden F, Grodstein F, Rosner BA, et al. Exposure to ultraviolet-B and risk of developing rheumatoid arthritis among women in the nurses’ health study. Ann Rheum Dis. (2013) 72:506–11. doi: 10.1136/annrheumdis-2012-202302, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ishikawa LLW, Colavite PM, Fraga-Silva TF d C, Mimura LAN, França TGD, Zorzella-Pezavento SFG, et al. Vitamin D deficiency and rheumatoid arthritis. Clin Rev Allergy Immunol. (2017) 52:373–88. doi: 10.1007/s12016-016-8577-0, [DOI] [PubMed] [Google Scholar]
- 18.Zhao N, Smargiassi A, Colmegna I, Hudson M, Fritzler M, Bernatsky S. Sunlight exposure, sun-protective behavior, and anti-citrullinated protein antibody positivity: a general population-based study in Quebec, Canada. Arthritis Care Res. (2022) 74:236–42. doi: 10.1002/acr.24448, [DOI] [PubMed] [Google Scholar]
- 19.Neogi T, Chen C, Niu J, Chaisson C, Hunter DJ, Choi H, et al. Relation of temperature and humidity to the risk of recurrent gout attacks. Am J Epidemiol. (2014) 180:372–7. doi: 10.1093/aje/kwu147, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Guan ML, Katsumoto TR, Toprover M, Rasheed N, Anand S, Pillinger MH, et al. Climate change, environmental risk factors, and gout. Arthritis Care Res. (2026) 78:620–9. doi: 10.1002/acr.25673, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sigurdardottir V, Jacobsson L, Schiöler L, Svärd A, Dehlin M, Toren K. Occupational exposure to inorganic dust and risk of gout: a population-based study. RMD Open. (2020) 6:e001178. doi: 10.1136/rmdopen-2020-001178, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.FitzGerald JD. Gout. Ann Intern Med. (2025) 178:ITC33–48. doi: 10.7326/ANNALS-24-03951, [DOI] [PubMed] [Google Scholar]
- 23.Wan X, Sun Y, Mao H, Han X, Zhou J, Duan W, et al. Exposure to environmental toxicants and gout risk in adults: unveiling the associations and potential targets. Ecotoxicol Environ Saf. (2025) 302:118580. doi: 10.1016/j.ecoenv.2025.118580, [DOI] [PubMed] [Google Scholar]
- 24.Helget LN, Mikuls TR. Environmental triggers of hyperuricemia and gout. Rheum Dis Clin N Am. (2022) 48:891–906. doi: 10.1016/j.rdc.2022.06.009, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ma L-Z, Ma Y-H, Ou Y-N, Chen S-D, Yang L, Dong Q, et al. Time spent in outdoor light is associated with the risk of dementia: a prospective cohort study of 362094 participants. BMC Med. (2022) 20:132. doi: 10.1186/s12916-022-02331-2, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Klepeis NE, Nelson WC, Ott WR, Robinson JP, Tsang AM, Switzer P, et al. The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. J Expo Anal Environ Epidemiol. (2001) 11:231–52. doi: 10.1038/sj.jea.7500165, [DOI] [PubMed] [Google Scholar]
- 27.Khanna T, Shraim R, Zarkovic M, van Weele M, van Geffen J, Zgaga L. Comprehensive analysis of seasonal and geographical variation in UVB radiation relevant for vitamin D production in Europe. Nutrients. (2022) 14:5189. doi: 10.3390/nu14235189, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Barber MRW, Drenkard C, Falasinnu T, Hoi A, Mak A, Kow NY, et al. Global epidemiology of systemic lupus erythematosus. Nat Rev Rheumatol. (2021) 17:515–32. doi: 10.1038/s41584-021-00668-1, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.GBD 2021 Gout Collaborators. Global, regional, and national burden of gout, 1990-2020, and projections to 2050: a systematic analysis of the global burden of disease study 2021. Lancet Rheumatol. (2024) 6:e507–17. doi: 10.1016/S2665-9913(24)00117-6, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Glanz K, Gies P, O’Riordan DL, Elliott T, Nehl E, McCarty F, et al. Validity of self-reported solar UVR exposure compared with objectively measured UVR exposure. Cancer Epidemiol Biomarkers Prev. (2010) 19:3005–12. doi: 10.1158/1055-9965.EPI-10-0709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Køster B, Søndergaard J, Nielsen JB, Allen M, Olsen A, Bentzen J. The validated Sun exposure questionnaire: Association of Objective and Subjective Measures of Sun exposure in a Danish population-based sample. Br J Dermatol. (2017) 176:446–56. doi: 10.1111/bjd.14861, [DOI] [PubMed] [Google Scholar]
- 32.Hutcheon JA, Chiolero A, Hanley JA. Random measurement error and regression dilution Bias. BMJ. (2010) 340:c2289. doi: 10.1136/bmj.c2289, [DOI] [PubMed] [Google Scholar]
- 33.Detert H, Hedlund S, Anderson CD, Rodvall Y, Festin K, Whiteman DC, et al. Validation of Sun exposure and protection index (SEPI) for estimation of Sun habits. Cancer Epidemiol. (2015) 39:986–93. doi: 10.1016/j.canep.2015.10.022, [DOI] [PubMed] [Google Scholar]
- 34.Würtz ET, Pugdahl K, Fenger-Grøn M, Berglind IA, Cherrie MPC, Dahlman-Höglund A, et al. A quantitative solar ultraviolet radiation job-exposure matrix for Europe. Ann Work Expo Health. (2025) 69:415–28. doi: 10.1093/annweh/wxaf011, [DOI] [PubMed] [Google Scholar]
- 35.Petersen B, Wulf HC. Application of sunscreen--theory and reality. Photodermatol Photoimmunol Photomed. (2014) 30:96–101. doi: 10.1111/phpp.12099, [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data used in the present study are available from the UK Biobank with restrictions applied. Access to the UK Biobank data can be requested by means of a standard protocol (https://www.ukbiobank.ac.uk/register-apply/). Requests to access these datasets should be directed to https://www.ukbiobank.ac.uk/register-apply/.