Abstract
This cross-sectional study examines the associations between bright light therapy, sleep regularity, and depression symptoms among adults in the US.
Introduction
Bright light therapy (BLT) may treat depression symptoms,1 but how light exerts mood-boosting effects is still under investigation. Here, we evaluate sleep regularity in the association between bright light and depression symptoms.
Methods
This cross-sectional study of 2011-2014 National Health and Nutrition Examination Survey (NHANES) data, representative of the noninstitutionalized US population, included nonpregnant participants 18 years or older with valid light and actigraphy (wrist-worn GT3X+; ActiGraph) for bright light (time above lux threshold [TALT1000]) and sleep regularity index (SRI)2 measures (eMethods in Supplement 1) and follows STROBE guidelines. Depression symptoms were measured with the Patient Health Questionnaire-9 (PHQ-9; scores range from 0 to 30, with higher scores indicating greater frequency of depression symptoms). This analysis used publicly available, deidentified data from NHANES, which was approved by the Ethics Review Board of the Centers for Disease Control and Prevention. All participants provided written informed consent.
This analysis was conducted between March 2023 and May 2024. Associations between TALT1000, SRI, and PHQ-9 score were tested in weighted linear, quasi-Poisson, and logistic regression models adjusted for age, sex, race and ethnicity, season, employment, physical activity, sleep duration, body mass index (BMI), and serum cotinine (eMethods in Supplement 1). Sex differences were tested with weighted t tests. In exploratory analyses, associations between TALT1000 and vitamin D, vitamin D and SRI, and TALT1000 and SRI with vitamin D as a potential mediator were tested in weighted linear regression models. P < .05 was considered statistically significant. Analyses were performed in R, version 4.1.1 (R Foundation).
Results
There were 6639 participants included in the analysis (mean [SD] age, 49.41 [17.41] years; 52.2% female; self-reported 8.3% Mexican American, 4.4% non-Hispanic Asian, 11.3% non-Hispanic Black, 67.7% non-Hispanic White, 5.7% other Hispanic, and 2.6% other or multiracial). Male participants had greater TALT1000 and fewer depression symptoms than female participants (Table 1). Greater TALT1000 was associated with lower depression symptoms (β = −0.19; 95% CI, −0.34 to −0.04) but not more severe depression. TALT1000 was no longer associated with depression score after adjusting for SRI (β = −0.11; 95% CI, −0.26 to 0.04; P = .13). Longer duration in bright light was associated with more regular sleep (β = 1.60; 95% CI, 0.95-2.25). Likewise, more regular sleep was associated with a lower depression score (β = −0.05; 95% CI, −0.06 to −0.03) and lower odds of mild or more severe depression (OR = 0.98; 95% CI, 0.97-0.99) (Table 2). In exploratory analyses, more vitamin D was associated with greater TALT1000 and more regular sleep but not depression symptoms.
Table 1. Variable Mean Values by Total Sample and by Sexa.
| Variable | All (n = 6639) | No. missing | Male (n = 3208) | Female (n = 3431) | P valueb |
|---|---|---|---|---|---|
| Age, mean (SD), y | 49.41 (17.41) | 0 | 48.52 (17.37) | 50.22 (17.41) | <.001 |
| Sex, No. (%) | |||||
| Female | 3431 (52.2) | 0 | NA | NA | NA |
| Male | 3208 (47.8) | ||||
| Race and ethnicity, No. (%) | |||||
| Mexican American | 797 (8.3) | 0 | 389 (8.9) | 408 (7.7) | .03 |
| Non-Hispanic Asian | 720 (4.4) | 344 (4.2) | 376 (4.6) | ||
| Non-Hispanic Black | 1600 (11.3) | 783 (10.7) | 817 (11.9) | ||
| Non-Hispanic White | 2707 (67.7) | 1327 (68.1) | 1380 (67.4) | ||
| Other Hispanic | 626 (5.7) | 267 (5.2) | 359 (6.1) | ||
| Otherc or multiracial | 189 (2.6) | 98 (3.0) | 91 (2.3) | ||
| BMI, mean (SD) | 29.16 (6.97) | 70 | 28.86 (6.05) | 29.42 (7.71) | .02 |
| Serum cotinine, mean (SD), ng/mL | 50.80 (121.99) | 314 | 61.04 (137.65) | 41.36 (104.65) | <.001 |
| Season, May 1 to October 31, No. (%) | 3401 (54.4) | 0 | 1608 (53.9) | 1793 (54.9) | .42 |
| Employment status, yes, No. (%) | 3352 (57.9) | 2 | 1773 (64.9) | 1579 (51.5) | <.001 |
| Daily Monitor–Independent Movement Summary physical activity value, mean (SD) | 13 246 (3787) | 0 | 12 863 (3839) | 13 596 (3705) | <.001 |
| Sleep duration, mean (SD), h | 7.90 (1.52) | 0 | 7.84 (1.60) | 7.96 (1.44) | .005 |
| TALT1000, mean (SD), h | 1.12 (1.03) | 0 | 1.39 (1.20) | 0.87 (0.77) | <.001 |
| SRI, mean (SD) | 58.32 (15.96) | 0 | 57.51 (16.25) | 59.05 (15.66) | .009 |
| Vitamin D, nmol/L, mean (SD) | 71.60 (29.11) | 282 | 67.67 (24.23) | 75.22 (32.55) | <.001 |
| PHQ-9, mean (SD) | 3.31 (4.62) | 438 | 2.74 (4.19) | 3.84 (4.93) | <.001 |
| PHQ-9 score ≥5, No. (%) | 1638 (25.0) | 438 | 647 (20.4) | 991 (29.3) | <.001 |
| PHQ-9 score ≥10, No. (%) | 675 (9.8) | 438 | 250 (7.4) | 425 (12.1) | <.001 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable; PHQ-9, Patient Health Questionnaire-9; SRI, sleep regularity index; TALT1000, time above lux threshold.
SI conversion factor: To convert cotinine to nmol/L, multiply by 5.675.
Sample-weighted values are provided for mean (SD) and number (%), in addition to number of participants for categorical variables.
Sex differences were tested with 2-sample t tests or χ2 tests.
Cannot be specified further.
Table 2. Regression Model Associations Between Variables and Outcomes of National Health and Nutrition Examination Survey 2011-2014 Participants Aged 18 Years or Older (N = 6639).
| Regression type | Variable | Outcome | Estimate (95% CI)a | P value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Crude | Model 1 | Model 2 | Model 3 | Crude | Model 1 | Model 2 | Model 3 | |||
| Linear regression | TALT1000, h | PHQ-9 score | −0.386 (−0.533 to −0.240) | −0.303 (−0.473 to −0.133) | −0.271 (−0.432 to −0.111) | −0.190 (−0.344 to −0.036) | <.001 | .001 | .002 | .02 |
| Quasi-Poisson regression | TALT1000, h | PHQ-9 score | −0.131 (−0.185 to −0.077) | −0.107 (−0.169 to −0.045) | −0.100 (−0.158 to −0.042) | −0.068 (−0.126 to −0.009) | <.001 | .002 | .002 | .03 |
| Logistic regression | TALT1000, h | PHQ-9 score ≥5 | 0.821 (0.751 to 0.898) | 0.846 (0.767 to 0.933) | 0.853 (0.768 to 0.947) | 0.888 (0.792 to 0.995) | <.001 | .002 | .005 | .04 |
| Logistic regression | TALT1000, h | PHQ-9 score ≥10 | 0.842 (0.753 to 0.942) | 0.869 (0.759 to 0.995) | 0.880 (0.778 to 0.995) | 0.948 (0.830 to 1.081) | .004 | .04 | .04 | .39 |
| Linear regression | TALT1000, h | SRI | 2.947 (2.412 to 3.482) | 2.996 (2.354 to 3.637) | 2.906 (2.220 to 3.591) | 1.604 (0.954 to 2.253) | <.001 | <.001 | <.001 | <.001 |
| Linear regression | SRI, 1-unit increase | PHQ-9 score | −0.065 (−0.078 to −0.053) | −0.070 (−0.084 to −0.057) | −0.060 (−0.073 to −0.048) | −0.047 (−0.062 to −0.032) | <.001 | <.001 | <.001 | <.001 |
| Quasi-Poisson regression | SRI, 1-unit increase | PHQ-9 score | −0.018 (−0.021 to −0.015) | −0.019 (−0.022 to −0.016) | −0.016 (−0.019 to −0.013) | −0.013 (−0.017 to −0.009) | <.001 | <.001 | <.001 | <.001 |
| Logistic regression | SRI, 1-unit increase | PHQ-9 score ≥5 | 0.972 (0.967 to 0.977) | 0.969 (0.963 to 0.975) | 0.973 (0.967 to 0.978) | 0.978 (0.970 to 0.985) | <.001 | <.001 | <.001 | <.001 |
| Logistic regression | SRI, 1-unit increase | PHQ-9 score ≥10 | 0.970 (0.963 to 0.976) | 0.966 (0.960 to 0.973) | 0.972 (0.965 to 0.979) | 0.978 (0.969 to 0.986) | <.001 | <.001 | <.001 | <.001 |
| Linear regression | TALT1000, h | Vitamin D | 3.092 (1.936 to 4.248) | 3.712 (2.625 to 4.799) | 3.093 (1.948 to 4.237) | 2.849 (1.747 to 3.951) | <.001 | <.001 | <.001 | <.001 |
| Linear regression | Vitamin D | SRI | 0.070 (0.046 to 0.094) | 0.078 (0.050 to 0.105) | 0.073 (0.047 to 0.099) | 0.046 (0.025 to 0.068) | <.001 | <.001 | <.001 | <.001 |
| Linear regression | Vitamin D | PHQ-9 score | −0.008 (−0.014 to −0.001) | −0.011 (−0.018 to −0.003) | −0.009 (−0.016 to −0.003) | −0.004 (−0.010 to 0.002) | .03 | .009 | .01 | .20 |
| Quasi-Poisson regression | Vitamin D | PHQ-9 score | −0.002 (−0.005 to −0.0002) | −0.003 (−0.006 to −0.001) | −0.003 (−0.005 to −0.001) | −0.001 (−0.003 to 0.001) | .03 | .01 | .01 | .22 |
Abbreviations: PHQ-9, Patient Health Questionnaire-9; SRI, sleep regularity index; TALT1000, time above lux threshold.
Model 1 adjusted for age (years, continuous with natural spline df = 6), sex, and race and ethnicity. Model 2 adjusted for all the covariates included in model 1 in addition to season and employment status. Model 3 adjusted for all the covariates included in model 2 in addition to physical activity, mean sleep duration, body mass index, and blood cotinine level.
Discussion
In this study, greater sleep regularity partly explained the association between greater bright light exposure and lower depression symptoms. More vitamin D was associated with more regular sleep timing but not depression. Our findings align with prior population-based research, supporting reduced depression among those with greater daytime light exposure.3 Bright light therapy can be effective at improving numerous mood outcomes,4 possibly through effects on sleep and the circadian system.5 Sleep regularity may be an important modifier of BLT2 and relate to chronotype and phase angle of entrainment.6 However, sleep regularity is not often considered in BLT. The phase-shifting effect of light may be limited in people with irregular sleep schedules, requiring a larger dose or altered timing of light exposure. Irregular sleepers may benefit from a light intervention.2
This study has strengths and limitations. It used a nationally representative sample of adult participants with objective measures of individual light exposure, actigraphy, and vitamin D. This is a cross-sectional analysis, so causality cannot be ascertained, and associations may be bidirectional. Although we are unable to rule out that mood may also influence time spent outdoors and/or bright light avoidance, our results support the need for further prospective analyses to test the causality of these factors. Future studies of BLT should consider the role of sleep regularity.
eMethods.
eReferences
Data Sharing Statement
References
- 1.Sit DK, McGowan J, Wiltrout C, et al. Adjunctive bright light therapy for bipolar depression: a randomized double-blind placebo-controlled trial. Am J Psychiatry. 2018;175(2):131-139. doi: 10.1176/appi.ajp.2017.16101200 [DOI] [PubMed] [Google Scholar]
- 2.Phillips AJK, Clerx WM, O’Brien CS, et al. Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. Sci Rep. 2017;7(1):3216. doi: 10.1038/s41598-017-03171-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Burns AC, Windred DP, Rutter MK, et al. Day and night light exposure are associated with psychiatric disorders: an objective light study in >85,000 people. Nat Ment Health. 2023;1:853-862. doi: 10.1038/s44220-023-00135-8 [DOI] [Google Scholar]
- 4.van Maanen A, Meijer AM, van der Heijden KB, Oort FJ. The effects of light therapy on sleep problems: a systematic review and meta-analysis. Sleep Med Rev. 2016;29:52-62. doi: 10.1016/j.smrv.2015.08.009 [DOI] [PubMed] [Google Scholar]
- 5.Czeisler CA, Kronauer RE, Mooney JJ, Anderson JL, Allan JS. Biologic rhythm disorders, depression, and phototherapy: a new hypothesis. Psychiatr Clin North Am. 1987;10(4):687-709. doi: 10.1016/S0193-953X(18)30543-4 [DOI] [PubMed] [Google Scholar]
- 6.Papatsimpa C, Schlangen LJM, Smolders KCHJ, Linnartz JMG, de Kort YAW. The interindividual variability of sleep timing and circadian phase in humans is influenced by daytime and evening light conditions. Sci Rep. 2021;11(1):13709. doi: 10.1038/s41598-021-92863-z [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
eMethods.
eReferences
Data Sharing Statement