ArticlePDF Available

Effect of Sunlight Exposure on Cognitive Function Among Depressed and Non-Depressed Participants: A REGARDS Cross-Sectional Study

Authors:

Abstract and Figures

Possible physiological causes for the effect of sunlight on mood are through the suprachiasmatic nuclei and evidenced by serotonin and melatonin regulation and its associations with depression. Cognitive function involved in these same pathways may potentially be affected by sunlight exposure. We evaluated whether the amount of sunlight exposure (i.e. insolation) affects cognitive function and examined the effect of season on this relationship. We obtained insolation data for residential regions of 16,800 participants from a national cohort study of blacks and whites, aged 45+. Cognitive impairment was assessed using a validated six-item screener questionnaire and depression status was assessed using the Center for Epidemiologic Studies Depression Scale. Logistic regression was used to find whether same-day or two-week average sunlight exposure was related to cognitive function and whether this relationship differed by depression status. Among depressed participants, a dose-response relationship was found between sunlight exposure and cognitive function, with lower levels of sunlight associated with impaired cognitive status (odds ratio = 2.58; 95% CI 1.43-6.69). While both season and sunlight were correlated with cognitive function, a significant relation remained between each of them and cognitive impairment after controlling for their joint effects. The study found an association between decreased exposure to sunlight and increased probability of cognitive impairment using a novel data source. We are the first to examine the effects of two-week exposure to sunlight on cognition, as well as the first to look at sunlight's effects on cognition in a large cohort study.
Content may be subject to copyright.
BioMed Central
Page 1 of 14
(page number not for citation purposes)
Environmental Health
Open Access
Research
Effect of sunlight exposure on cognitive function among depressed
and non-depressed participants: a REGARDS cross-sectional study
Shia T Kent*1, Leslie A McClure2, William L Crosson3, Donna K Arnett1,
Virginia G Wadley4 and Nalini Sathiakumar1
Address: 1Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, 1665 University Blvd, Birmingham,
Alabama, USA, 2Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, 1665 University Blvd, Birmingham,
Alabama, USA, 3National Space Science and Technology Center, NASA Marshall Space Flight Center, 320 Sparkman Drive, Huntsville, Alabama,
USA and 4Department of Medicine, University of Alabama at Birmingham, 1808 7th Avenue South Birmingham, Alabama, USA
Email: Shia T Kent* - shia@uab.edu; Leslie A McClure - lmcclure@uab.edu; William L Crosson - bill.crosson@nasa.gov;
Donna K Arnett - Arnett@ms.soph.uab.edu; Virginia G Wadley - vwadley@uab.edu; Nalini Sathiakumar - NSathiakumar@ms.soph.uab.edu
* Corresponding author
Abstract
Background: Possible physiological causes for the effect of sunlight on mood are through the
suprachiasmatic nuclei and evidenced by serotonin and melatonin regulation and its associations
with depression. Cognitive function involved in these same pathways may potentially be affected by
sunlight exposure. We evaluated whether the amount of sunlight exposure (i.e. insolation) affects
cognitive function and examined the effect of season on this relationship.
Methods: We obtained insolation data for residential regions of 16,800 participants from a
national cohort study of blacks and whites, aged 45+. Cognitive impairment was assessed using a
validated six-item screener questionnaire and depression status was assessed using the Center for
Epidemiologic Studies Depression Scale. Logistic regression was used to find whether same-day or
two-week average sunlight exposure was related to cognitive function and whether this
relationship differed by depression status.
Results: Among depressed participants, a dose-response relationship was found between sunlight
exposure and cognitive function, with lower levels of sunlight associated with impaired cognitive
status (odds ratio = 2.58; 95% CI 1.43–6.69). While both season and sunlight were correlated with
cognitive function, a significant relation remained between each of them and cognitive impairment
after controlling for their joint effects.
Conclusion: The study found an association between decreased exposure to sunlight and
increased probability of cognitive impairment using a novel data source. We are the first to examine
the effects of two-week exposure to sunlight on cognition, as well as the first to look at sunlight's
effects on cognition in a large cohort study.
Published: 28 July 2009
Environmental Health 2009, 8:34 doi:10.1186/1476-069X-8-34
Received: 8 January 2009
Accepted: 28 July 2009
This article is available from: http://www.ehjournal.net/content/8/1/34
© 2009 Kent et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 2 of 14
(page number not for citation purposes)
Introduction
It is widely accepted that climate and season affect psycho-
logical characteristics [1,2]. Recent research has shown
that serotonin and melatonin regulation, mechanisms
that are involved in the relationship between sunlight and
light therapy on mood, are also involved in cognition,
which suggests that cognitive function may also be influ-
enced by light [3-5]. Melatonin, serotonin and other
mechanisms involved in circadian rhythms are associated
with cognitive functioning, and are regulated by the
suprachiasmatic nuclei (SCN), which are susceptible to
the effects of differing intensities and patterns of environ-
mental illumination [6]. However, the effect of sunlight
and light therapy on cognitive function has not been ade-
quately studied. This study aimed to explore if sunlight
exposure, measured by insolation (the rate of solar radia-
tion received in an area), is associated with cognitive
impairment. In addition, examined the role of season in
this relationship. This study was conducted using baseline
data from a large prospective study, the REasons for Geo-
graphic And Racial Differences in Stroke (REGARDS)
Study, and National Aeronautics and Space Administra-
tion (NASA) satellite and ground data. We hypothesized
that lower levels of sunlight exposure at participants' resi-
dences would be associated with increased rates of cogni-
tive impairment. This study was the first to examine the
effects of two-week exposure to natural sunlight on cogni-
tion, as well as the first to look at solar effects on cognition
in a large cohort study.
Methods
Participants
The present study consisted of participants from the
REGARDS study, which has been described in detail else-
where [7]. In brief, REGARDS is a longitudinal study
being conducted to determine relationships between var-
ious risk factors and the incidence of stroke. The partici-
pants are aged 45 and older and sampled from the 48
conterminous United States. Study participants were over-
sampled from the "Stroke Belt", a high stroke mortality
region consisting of the 8 southeastern states of Arkansas,
Louisiana, Tennessee, Mississippi, Alabama, Georgia,
North Carolina, and South Carolina. The sample popula-
tion was particularly oversampled from the "Stroke
Buckle", a region with even higher stroke mortality along
the coastal plains of Georgia, North Carolina, and South
Carolina. Within each region the planned recruitment
included half white and half African-Americans (actual:
41% African-American, 59% white). Planned recruitment
within each race-region strata was half male and half
female (actual total recruitment: 45% male, 55% female).
At baseline, a telephone interview was conducted that
recorded the patient's medical history, demographic data,
socioeconomic status, stroke-free status, depression, and
cognitive screening. An in-home exam was conducted
recording height, weight, and blood pressure. All partici-
pants provided written informed consent, and the study
was approved by the Institutional Review Board for
Human Subjects at the University of Alabama at Birming-
ham, as well as all other participating institutions.
Sunlight Exposure (Insolation) Assessment
Sunlight exposure was measured using data values pre-
pared and provided by NASA's Marshall Space Flight
Center. Solar radiation values were obtained for 2003 to
2006 from the North American Regional Reanalysis
(NARR), an assimilated data product produced by the
National Center for Environmental Prediction (NCEP), a
division of the U.S. National Weather Service. The prod-
uct, including information from satellites and ground
observations, was compiled on a grid with a 32 km reso-
lution over North America and matched to each partici-
pant by the geocoded residence obtained from the
REGARDS database. Solar radiation in Watts/meters2 (W/
m2), a measure of the instantaneous solar energy reaching
the Earth's surface, was assessed 8 times a day at 3-hour
intervals for each residence starting at 1:00 AM Pacific
Standard Time (PST). A daily integral of solar radiation
was calculated; this is referred to as insolation and has
units of kilojoules per meters2 per day (KJ/m2/day). As a
point of reference, under clear skies in late spring or early
summer, a typical daily insolation value in the central U.S.
is approximately 25,000–30,000 KJ/m2. In late fall or
early winter, a typical daily value is approximately 8,000–
10,000 KJ/m2.
The current residence from the original recruitment file
plus updated information from participant at time of
scheduling in-home exam was used to geocode each par-
ticipant. Geocoding of the participants was performed
using SAS/GIS batch geocoding. Information obtained
from SAS/GIS with 80% accuracy or greater was utilized.
Using a subset of the data, we validated the results from
the SAS/GIS procedure against a commercially available
program http://www.geocode.com using the Haversine
formula, and found there to be high agreement between
the two algorithms [8]. For those with a SAS/GIS accuracy
of 80% or greater, the difference between the latitudes
given between the two programs had a mean of 0.23 kil-
ometers and a maximum of 0.95 kilometers.
Cognitive Assessment
A six-item screener questionnaire was used to evaluate
global cognitive status by assessing short-term recall and
temporal orientation [9-11]. As REGARDS has done in
other studies, the score of this screener was dichotomized
into an outcome of cognitively impaired or intact. A score
of four or fewer correct responses out of the six questions
indicated cognitive impairment. Callahan et al. 2002 val-
idated the screener in both a community-based popula-
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 3 of 14
(page number not for citation purposes)
Table 1: Demographic, medical, and lifestyle characteristics by cognitive status
Characteristics All Subjects
(N = 16800)
Missing Intact Cognitive Status
(N = 15421; 92%)
Impaired Cognitive Status
(N = 1379; 8%)
p-value
N (%) N N (%) N (%)
Demographics
Male 6657 (40) 0 6033 (39) 624 (45) <0.0001
Education 19
Less than High School 2081 (12) 1719 (11) 362 (26)
High School 4404 (26) 3978 (26) 426 (31) <0.0001
Some College 4525 (27) 4230 (27) 305 (22)
College Diploma 5761 (34) 5480 (36) 281 (20)
Age 5
Less than 55 years 2077 (12) 1982 (13) 95 (7)
55 to 59 years 2908 (17) 2756 (18) 152 (11)
60 to 64 years 3197 (19) 3009 (20) 188 (14) <0.0001
65 to 69 years 3117 (19) 2870 (19) 247 (18)
70 to 74 years 2418 (14) 2172 (14) 246 (18)
75 to 79 years 1765 (11) 1536 (10) 229 (17)
80 or more years 1313 (8) 1091 (7) 222 (16)
Income 2171
Less than $20,000/year 3097 (21) 2698 (20) 399 (35)
$20,000–35,000/year 4017 (27) 3633 (27) 384 (33) <0.0001
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 4 of 14
(page number not for citation purposes)
$35,000-$75,000/year 4968 (34) 4678 (35) 290 (25)
Over $75,000/year 2547 (17) 2467 (18) 80 (7)
Region 0
Non Belt/Buckle 7850 (47) 7211 (47) 639 (46) 0.13
Stroke Belt 6095 (36) 5567 (36) 528 (38)
Stoke Buckle 2855 (17) 2643 (17) 212 (15)
Population Density 0
Urban 13532 (81) 12365 (80) 1167 (85)
Mixed 1678 (10) 1571 (10) 107 (8) 0.0003
Rural 1590 (9) 1485 (10) 105 (8)
Season 0
Spring 3610 (22) 3260 (21) 350 (25) <0.0001
Summer 6439 (38) 5998 (39) 441 (32)
Fall 3410 (20) 3181 (21) 229 (17)
Winter 3341 (20) 2982 (19) 359 (26)
Medical Factors
BMI 238
Underweight 212 (1) 189 (1) 23 (2)
Normal 3973 (24) 3651 (24) 322 (24) 0.26
Overweight 5970 (36) 5462 (36) 508 (38)
Table 1: Demographic, medical, and lifestyle characteristics by cognitive status (Continued)
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 5 of 14
(page number not for citation purposes)
Obese 6407 (39) 5906 (39) 501 (37)
Diabetic 3527 (22) 623 3150 (21) 377 (29) <0.0001
Hypertensive 9854 (59) 73 8943 (58) 911 (67) <0.0001
High Cholesterol 1740 (10) 80 1595 (10) 145 (11) 0.85
Depressed 1877 (11) 155 1624 (11) 253 (18) <0.0001
PCS-12 (mean, stddev) 46.1 (10.6) 0 46.3 (10.5) 44.0 (11.0) <.0001
Lifestyle Factors
Never Used Alcohol 5129 (31) 0 4606 (30) 523 (40) <0.0001
No weekly exercise 5948 (36) 234 5359 (35) 589 (44) <0.0001
Smoking 67
Current 2445 (15) 2227 (15) 218 (16) 0.18
Past 6609 (40) 6055 (39) 554 (40)
Never 7679 (46) 7078 (46) 601 (44)
stddev = standard deviation; BMI = Body Mass Index; PCS = Physical Components Summary
P-values for categorical variables provided from a chi squared test statistic and for continuous variables provided from a scatterthwaite t-test statistic calculated for each variable by cognitive
status.
P-values in bold indicate values that are significant at α = 0.05.
Table 1: Demographic, medical, and lifestyle characteristics by cognitive status (Continued)
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 6 of 14
(page number not for citation purposes)
tion of 344 black adults aged 65 or older and a population
of 651 subjects who were referred to the Alzheimer's dis-
ease Center (16% black). Results from the community-
based sample found that for a six-item screener score of 4
or fewer, using clinically confirmed cognitive impairment
as the gold standard, the sensitivity was 74% and specifi-
city was 80% [10]. The instrument was based on and val-
idated against the Mini-mental State examination [11]. It
was also validated against other cognitive measures and
diagnoses of both dementia and non-dementia cognitive
impairment [10].
Participant Selection
A total of 19,853 participants without previous stroke
were enrolled in the study at the time of this analysis
(December 1, 2006). Of these, 3,020 patients with poor
geocoding (less than 80% accuracy) and a further 33
patients who were missing cognitive scores were excluded,
leaving 16,800 participants. Due to data missing in any of
the potential confounders, 3,253 of the 16,800 partici-
pants were excluded during model selection. Once the
model selection was completed, only 2,326 of the 16,800
participants were excluded due to missing covariates
selected for the final multivariable model. Chi-squared
and t-tests were used to measure differences of the 5,379
excluded and 14,474 included subjects in the final model.
Statistical Analyses
We analyzed insolation from the day the six-item screener
was administered and the preceding two weeks. Insola-
tion measurements were analyzed as continuous variables
Table 2: Crude logistic univariate relationships of depression
with two-week insolation
Characteristics OR (95% CI)
Primary Variable of Interest
2 week solar radiation
(by 5,000 KJ/m2/day)
<10,000 J/m21.14 (0.91–1.42)
10,000–15,000 J/m20.94 (0.81–1.09)
15,000–20,000 J/m20.87 (0.75–1.01)
20,000–25,000 J/m20.87 (0.76–1.00)
>25,000 J/m21.00 (Referent)
CMH chisq p = 0.6054
OR = odds ratio; CI = confidence interval
CMH chisq = Cohran-Mantel-Haenzel chi-squared
Table 3: Crude logistic univariate relationships of cognitive
impairment with predictors and final covariates
Characteristics OR (95% CI)
Primary Variable of Interest
Same day solar radiation
(by 5,000 KJ/m2/day)
<10,000 J/m21.15 (0.96–1.38)
10,000–15,000 J/m21.02 (0.86–1.21)
15,000–20,000 J/m20.96 (0.82–1.13)
20,000–25,000 J/m21.03 (0.88–1.21)
>25,000 J/m21.00 (Referent)
CMH chisq p = 0.4346
2 week solar radiation
(by 5,000 KJ/m2/day)
<10,000 J/m21.36 (1.08–1.70)
10,000–15,000 J/m21.13 (0.95–1.33)
15,000–20,000 J/m21.09 (0.92–1.30)
20,000–25,000 J/m21.01 (0.86–1.19)
>25,000 J/m21.00 (Referent)
CMH chisq p = 0.0075
Covariates
Male 1.29 (1.15–1.44)
Education
Less than High School 4.56 (3.57–5.83)
High School 3.26 (2.55–4.17)
Some College 1.91 (1.49–2.46)
College Diploma 1.00 (Referent)
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 7 of 14
(page number not for citation purposes)
and categorical variables (by 5,000 KJ/m2/day incre-
ments).
Due to prior evidence regarding relationships with cogni-
tive function, we considered the following as potential
confounders: sex, geographic region (stroke belt, stroke
buckle, or non-stroke belt), population density (urban,
suburban, and rural), income (less than $20,000, $20,000
to $34,999, $35,000 to $74,900, or $75,000 and more),
education (less than high school, high school diploma,
some college, or college diploma), race (black or white),
smoking (current, past, or never), alcohol use (never used
or ever used), Body Mass Index (BMI) (underweight, nor-
mal, overweight, or obese), hypertension status (systolic
blood pressure 140, diastolic blood pressure 90 or self-
reported use of hypertension medications), high choles-
Predicted Probabilities and Odds Ratios of Cognitive Impair-ment by Depression StatusFigure 1
Predicted Probabilities and Odds Ratios of Cognitive
Impairment by Depression Status.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
<10,000 10,000-
15,000
15,000-
20,000
20,000-
25,000
>25,000
Insola tion (J/m
2
)
Proability of Cognitive Impairment
Depress ed
Non-depress ed
OR
d
=2.59
(95% CI:1.43-4. 70)
OR
d
=0.94
(95% CI:0.59-1.51)
OR
d
=0.83
(95% CI:0.51-1. 37)
OR
d
=1.14
(95% CI:0.74-1. 76) OR
d
=1.00
(reference)
OR
n
=1.00
(reference)
OR
n
=1.14
(95% CI:0.85-1. 53)
OR
n
=1.10
(95% CI:0.90-1. 35)
OR
n
=1.13
(95% CI:0.91-1. 40) OR
n
=0.89
(95% CI:0.73-1. 09)
Spring 1.46 (1.26–1.69)
Depressed 1.90 (1.65–2.20)
PCS-12 (by 10 unit increase) 0.82 (0.78–0.86)
Never Used Alcohol 1.44 (1.28–1.61)
OR = odds ratio; CI = confidence interval; PCS = Physical
Components Summary
CMH chisq = Cohran-Mantel-Haenzel chi-squared
ORs in bold have CI's which do not overlap a null value
Table 3: Crude logistic univariate relationships of cognitive
impairment with predictors and final covariates (Continued)
CMH chisq p < .0001
Income
Less than $20,000 per year 4.47 (3.46–5.77)
$20,000 to $35,000 per year 3.21 (2.49–4.15)
$35,000 to $75,000 per year 1.88 (1.45–2.45)
$75,0000 or more per year 1.00 (Referent)
CMH chisq p < .0001
Age
Less than 55 years 1.00 (Referent)
55 to 59 years 1.15 (0.89–1.50)
60 to 64 years 1.30 (1.01–1.68)
65 to 69 years 1.80 (1.41–2.29)
70 to 74 years 2.36 (1.85–3.02)
75 to 79 years 3.11 (2.43–3.99)
80 or more years 4.25 (3.30–5.46)
CMH chisq p < .0001
Region
Non Belt/Buckle 1.00 (Referent)
Stroke Belt 1.07 (0.95–1.21)
Stoke Buckle 0.91 (0.77–1.06)
Season
Summer 1.00 (Referent)
Fall 0.98 (0.83–1.16)
Winter 1.64 (1.42–1.90)
Table 3: Crude logistic univariate relationships of cognitive
impairment with predictors and final covariates (Continued)
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 8 of 14
(page number not for citation purposes)
terol (cholesterol >240), diabetes status (fasting glucose
126, non-fasting glucose 200, or self-reported diabetes
medications), exercise (weekly or less than weekly),
depression status based on the Center for Epidemiologic
Studies Depression Scale (CES-D) scale, physical function
as measured by the 100 point scale Physical Components
Summary (PCS) in the 12-item Short Form (SF-12), sea-
son of phone interview (spring, summer, fall, or winter),
and age in years (45–54, 55–59, 60–64, 65–69, 70–74,
75–79, or 80 or more) [12,13].
T-tests, chi-squared tests, and correlation tests were used
to determine preliminary relationships between insola-
tion, cognition, and the covariates. Cochran-Mantel-
Haenzel (CMH) chi-squared tests were used to determine
if ordinally categorized insolation had relationships with
categorical predictors. Logistic regressions were used to
model the association between insolation and cognition.
Backwards elimination was used to build the final multi-
variable model. Covariates whose relationships with cog-
nitive function carried p-values over 0.10 were not
considered for inclusions in the multivariable model. Due
to the REGARDS sampling methods, the variables race,
region, and sex were included in the multivariable model
regardless of statistical significance. All interactions of the
remaining covariates with insolation in the final models
were considered. For any significant interactions the pre-
dicted probabilities and odds ratios (ORs) of cognitive
impairment with their 95% confidence intervals were cal-
culated. Since other factors related to seasonality besides
insolation may be related to cognitive function, such as
temperature, activity level, allergies, and stress, the final
model was run both with and without season as a covari-
ate [1,14-17].
Finally, the cognitive screener was divided into two com-
ponents, the three points that measure short-term recall
and the three points that measure temporal orientation.
Each of these components was used to explore individual
relationships with insolation. Univariate relationships
were analyzed using chi-square tests and multivariable
relationships were evaluated by taking the final logistic
regression model obtained above and replacing the sum-
mary measure of cognitive impairment with each of the
individual components of the six-item screener. For these
analyses, the three point component measures were
dichotomized, with one missing point indicating a deficit
in either orientation or recall. The measures were also ana-
lyzed continuously, so that each component would be
equal to the number of points obtained (0, 1, 2, or 3).
Results
Continuous two-week insolation (p = 0.005) but not
same-day insolation (p = 0.332) differed significantly by
cognitive status (data not shown). Table 1 presents the
baseline characteristics overall and by cognitive status.
Sex, education, age, income, population density, season,
diabetes status, hypertension status, depression status,
PCS, alcohol usage, and weekly exercise all differed signif-
icantly by cognitive status (all p-values < 0.05; Table 1).
CMH chi-squared tests indicated that gender, age, region,
population density, season, BMI, PCS-12, and weekly
exercise had dose-response relationships with ordinally
categorized insolation, but education, diabetes, hyperten-
sion, high cholesterol, smoking (data not shown), and
depression (Table 2) did not. Dose-response relationships
of income (p = 0.0956) and alcohol use (p = 0.0650) with
insolation were marginal (data not shown).
Table 3 shows the univariate analyses of categorized inso-
lation testing for dose-response relationships with cogni-
tive function. Two week insolation categorized by 5,000
KJ/m2/day showed does-response effects (p = 0.0075).
Participants in the lowest category of insolation compared
to those in the highest insolation category had 1.36 times
(95% CI = 1.08–1.70) the odds of cognitive impairment.
When this measure of insolation was modeled as an ordi-
nal variable, each successively lower insolation level com-
pared to the adjacent higher insolation level had 1.06
(95% CI 1.02–1.11) fold odds of cognitive impairment.
This study also confirmed prior study findings that there
is a dose-response relationship between cognitive impair-
ment and age, income, and education status (Table 3).
Depressed participants showed an increased odds of cog-
nitive impairment (OR = 1.90; 95% CI 1.65–2.20). In
addition, univariate analyses showed that the seasons of
winter (OR = 1.64; 95% CI 1.42–1.90) and spring (OR =
1.46; 95% CI 1.26–1.69) compared to fall gave increased
odds of cognitive impairment.
Due to non-significant (p > 0.10) relationships of high
cholesterol, BMI, and smoking with cognitive impairment
in crude analyses, these variables were not considered for
model-building. Further, population density, diabetes,
hypertension, and weekly exercise were dropped from the
model for non-significance (p > 0.10) during backwards
selection. The final multivariable model included sex,
race, education, income, age, region, depression, PCS-12
and alcohol as covariates. Because depression had a signif-
icant interaction with sunlight exposure in this model (p
= 0.008), the final model included this interaction term
and predicted probabilities of cognitive impairment were
calculated according to sunlight exposure category and
depression status. Figure 1 shows that the predicted prob-
abilities of cognitive impairment for depressed partici-
pants are consistently higher than the predicted
probabilities of impairment for non-depressed partici-
pants. Figure 1 also shows that depressed participants
receiving less than 10,000 KJ/m2/day of sunlight exposure
compared to depressed participants in other solar expo-
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 9 of 14
(page number not for citation purposes)
Table 4: Final covariates of excluded and modeled participants
Characteristics Participants in the Final Model Excluded Participants p-value
N (%) N (%)
Total 14,474 (73) 5378 (27)
Male 5944 (41) 1887 (35) <.0001
Education
Less than High School 1670 (12) 798 (15)
High School 3751 (26) 1525 (28) <.0001
Some College 3963 (27) 1407 (26)
College Diploma 5090 (35) 1628 (30)
Income
Less than $20,000 per year 3071 (21) 623 (22)
$20,000 to $35,000 per year 3991 (28) 797 (28) 0.42
$35,000 to $75,000 per year 4904 (34) 914 (32)
$75,0000 or more per year 2508 (17) 484 (17)
Age
Less than 55 years 1885 (13) 585 (11)
55 to 59 years 2583 (18) 848 (16)
60 to 64 years 2759 (19) 1052 (20) <.0001
65 to 69 years 2654 (18) 1038 (19)
70 to 74 years 2063 (14) 758 (14)
75 to 79 years 1443 (10) 653 (12)
80 or more years 1087 (8) 436 (8)
Black 6291 (44) 2046 (38) <.0001
Region
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 10 of 14
(page number not for citation purposes)
sure categories had a significantly higher predicted proba-
bility of cognitive impairment. When season was added to
the multivariable model, it was significantly related to
cognitive function (p < 0.01). Both spring (OR = 1.20;
95% CI 1.01–1.42) and winter (OR = 1.33; 95% CI 1.07–
1.67) seasons compared to summer season showed
increased odds of cognitive impairment. The relationships
between sunlight exposure, depression, and cognitive
impairment were unchanged when season was added to
the model, giving identical predicted probabilities as in
Figure 1.
Relationships between temporal orientation and short-
term recall components of the six-item screener did not
reveal any significant univariate or multivariable relation-
ships with insolation, indicating that no single compo-
nent was likely responsible for the relationship (data not
shown).
Conclusion
We found that among participants with depression, low
exposure to sunlight was associated with a significantly
higher predicted probability of cognitive impairment.
This relationship remained significant after adjustment
for season. Among participants without depression, inso-
lation did not have a significant effect on cognitive func-
tion.
This study adds to the body of literature that shows that
environment and lifestyle profoundly affect those who are
prone to Seasonal Affective Disorder (SAD) and other
types of depression. Studies based on violent homicides,
suicides, and aggressive behaviors have repeatedly dem-
onstrated seasonal characteristics, typically with peaks in
the spring. These peaks have been associated with sunlight
and other climatic variables [18]. Those with SAD have
mental states that vary with season, with regular depres-
sions occurring in the winter and remissions in the spring
or summer. It is established that these SAD episodes are
associated with the shorter daylight hours occurring in
winter [19].
The fact that sunlight exposure was associated with cogni-
tion in depressed participants supports our hypothesis
that the physiological mechanisms which give rise to sea-
sonal depression may also be involved with sunlight's
effect on cognitive function. Leonard and Myint, 2006
laid out a paradigm showing how lack of environmental
Non Belt/Buckle 6802 (47) 2158 (40)
Stroke Belt 5228 (36) 2074 (39) <.0001
Stoke Buckle 2444 (17) 1145 (21)
Season
Summer 5542 (38) 2080 (39)
Fall 2957 (20) 1116 (21) 0.76
Winter 2891 (20) 1040 (19)
Spring 3084 (21) 1142 (21)
Never Used Alcohol 4291 (30) 1842 (34) <.0001
Depressed 1605 (11) 653 (13) 0.0047
PCS-12 (mean, stddev) 46.2 (10.6) 45.9 (10.5) 0.0449
stddev = standard deviation; PCS = Physical Components Summary
P-values for categorical variables provided from a chi squared test statistic and for continuous variables provided from a pooled t-test statistic
calculated for each variable by inlcusion status.
Table 4: Final covariates of excluded and modeled participants (Continued)
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 11 of 14
(page number not for citation purposes)
illumination and other stresses might lead to altered sero-
tonin levels, neurodegeneration, depression, cognitive
deficits, and ultimately dementia [20]. Both seasonal and
non-seasonal depression have been shown to have rela-
tionships with environmental illumination [19,21,22].
Theories regarding the body's seasonal cycles, which affect
depression and may also affect cognition, are mostly
based on the regulation of the body's circadian rhythms
by the hypothalamic suprachiasmatic nuclei (SCN)
[6,23]. The SCN are modulated by various factors such as
body temperature and physical activity, but are in particu-
lar modulated by light received by retinal sensors at opti-
mal wavelengths close to sunlight's dominant wavelength
of 477 nanometers [23]. The SCN regulate the body's
sleep cycle, body temperature, blood pressure, digestion,
immune system, and various hormonal systems. Dysfunc-
tional circadian rhythms and sleep disorders, which can
occur from inadequate environmental light, have been
associated with cognitive deficits [24]. One of the SCN's
regulatory functions are their inhibition of the pineal
gland from turning serotonin into melatonin during the
presence of daytime light [19]. Abnormalities and regula-
tion of both the melatonin and serotonin systems have
been found to vary according to sunlight and light therapy
in SAD [25,26], bipolar [5] and schizophrenic [27]
patients, and even among those without psychiatric diag-
noses [28]. Serotonin and melatonin have also been
implicated in many mental and cognitive disorders, such
as Alzheimer's disease, Parkinson's disease, and sleep dis-
orders [25,29].
Light has been shown to also affect brain blood flow. Cer-
ebral blood flow has specifically been found to improve
after phototherapy in pre-term infants [30] and SAD
patients [31], and has repeatedly been found to be associ-
ated with cognitive functions, such as memory. Inade-
quate cerebral blood flow has been found to be a likely
cause or result of decreasing cognitive functions among
those with cardiovascular diseases [32-34], as well as cor-
related with age-related diseases such as Alzheimer's [35]
and non-age related diseases such as Lyme disease [36].
The relationships that serotonin, melatonin, and cerebral
hemodynamics have with sunlight, depression, and cog-
nitive function suggest that persons prone to sunlight-
related mood disturbances may also be prone to sunlight-
related cognitive difficulties.
This study adds to the limited base of knowledge regard-
ing the relationship of weather variables with cognitive
function. Studies that have tested the effects of artificial
light on cognitive abilities have found that increased light
exposure leads to increased alertness and a variety of
changes in regional brain activity [37]. In addition, differ-
ent spectral wavelengths have been found to have differ-
ing effects on memory and other cognitive abilities [38].
However, unlike our study, these studies only dealt with
immediately acute effects and did not directly examine the
effects of natural sunlight. They also have poor generaliz-
ability due to using animals or small numbers of human
subjects from populations with particular occupations,
socio-economic statuses, or ethnicities. We found only
two studies that examined the relationship between cog-
nition and sunlight, both of which only dealt with the
effects of immediate, short-term exposure. Sinclair et al.
(1994) found that increased sunlight exposure was associ-
ated with an increase in heuristic processing, which
requires memory storage and relevant memory retrieval,
but a decrease in systematic processing, a more compli-
cated process requiring analysis and judgment [39]. Keller
et al. (2005) found weak positive correlations between
sunny days and performance on two measures of cogni-
tion, digit span and openness to new information [1]. A
major difference between our study and the previous stud-
ies is our method of obtaining the participant's exposure
to sunlight. The NASA satellite used to obtain the insola-
tion data in this study is able to record data eight times a
day as well as provide an accurate characterization of inso-
lation matched to each participant's geocoded home
address. This gives superior space and time precision com-
pared to ground sensors used by previous studies. Keller et
al. (2005) used barometric pressure as a surrogate for
measuring sunny, clear days. Other studies that have not
found significant associations between mood or cogni-
tion and sunlight in the general population [40] have
directly measured insolation using the nearest available
ground sensors, which are centered on metropolitan
areas. Satellite data allowed us to obtain multiple daily
measurements across urban, suburban, and rural areas.
Exposure misclassification exists as a possible limitation
of the study. Exposure misclassifications may have taken
place if during the two week exposure measurements par-
ticipants spent a large amount of time in a climate differ-
ent than the climate recorded by the satellite. This could
happen if participants spent large amounts of time
indoors or away from their reported home addresses.
Also, the daily insolation values were taken by the satellite
sensors recorded simultaneously throughout the four dif-
ferent time zones in the U.S. Thus, this point represented
different times in the day for different regions of the coun-
try. For example, the insolation values used to calculate
insolation for participants in the Eastern time zone corre-
spond to 3-hour sampling periods of 1:00, 4:00, 7:00 and
10:00 AM/PM standard time, while for the Mountain time
zone the sampling times are 2:00, 5:00, 8:00 and 11:00
AM/PM standard time. However, the relatively short
three-hour intervals at which the measurements were
taken captures the diurnal cycle well, and the misclassifi-
cation due to this issue is quite small. It should also be
noted that while the relationships found in this study may
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 12 of 14
(page number not for citation purposes)
not apply in younger people (as our study was restricted
to those 45 years or older), the participants of the study
were recruited from all over the country, with differing
demographics, medical factors, and lifestyle factors.
Due to the exclusion of a considerable proportion (27%)
of 19,853 enrolled REGARDS participants from the final
model as a result of missing values and poor geocoding,
we investigated if the excluded participants differed from
those with complete information. While sex, education,
region, alcohol, age, and depression status of the excluded
subjects were statistically different, the proportions of
these variables all differed by eight percentage points or
less (Table 4). Covariates with larger differences (over 2%)
show a disproportionate inclusion into the model of
males, those with college diplomas, blacks, non-belt resi-
dents, and those that have ever used alcohol. These varia-
bles all have known relationships with cognitive
impairment and would be the most likely causes of any
bias, which might have resulted in underestimating or
overestimating the effect of insolation on cognition.
There always remains the possibility of residual con-
founding. In addition to the imprecision or bias that may
be present in any measurement, we could not account for
specific psychiatric diagnoses or medicine consumption.
Also, environmental temperature may be related to cogni-
tive function, although temperature fluctuations are par-
tially controlled for by season, exercise, cardiovascular
factors, and other possible correlates of temperature [41-
44]. Eye function is another possible confounder. Specifi-
cally crystalline lens transmittance and papillary area have
been found affect circadian photoreception, although
controlling for age may reduce confounding from these
factors. [23]. The interview's time of the day may also have
an effect on cognition; however, the sampling method of
REGARDS should result in all participants having an
equal chance of being interviewed during a given time
resulting in similar time distributions at any given varia-
ble level [6].
This new finding that weather may not only affect mood,
but also cognition, has significant implications and needs
to be further elucidated in future studies. That insolation
had a relationship with cognitive function but not depres-
sion, and that the effect of insolation on cognition is
shown among depressed, but not non-depressed partici-
pants indicates that insolation may have a relationship
with cognition that is independent of, but modified by,
depression. It also suggests the possibility that light ther-
apy that is prescribed for SAD may also improve cognitive
function. Future studies involving light and other thera-
pies for SAD should include cognitive function as a varia-
ble in order to determine relationships with insolation,
mood, and cognitive function. Future studies are also
needed to demonstrate particular cognitive deficits. The
six-item screener was designed to test global cognitive sta-
tus for large numbers of participants in an easy and effi-
cient manner. While it has adequate sensitivity and
specificity as a screening procedure to identify those most
likely to have cognitive deficits, it cannot be used to make
any particular diagnosis and is limited in its sensitivity to
cognitive deficits of small magnitude. In the future, more
specific exams and diagnoses can be used to find the spe-
cific effects of sunlight on cognitive processes and dis-
eases. We also show that future research regarding
treatment and lifestyle should in particular focus on eld-
erly people, since the older a participant is, the more likely
the participant is to be cognitively impaired. In addition,
research and possibly programs regarding outreach and
health education might be targeted to depressives in lower
education groups, not only because they are known to
have lower access to healthcare in general, but also
because they are at a particularly high risk of cognitive
impairment. Many of the prior studies have looked at the
effects of weather on mood and cognition as seasonal, but
the results of this study demonstrate that the effect of sea-
son on cognition can be explained by sunlight and other
variables. This study also has an interesting finding
regarding those without an elevated level of depressive
symptoms. We did not find that sunlight meaningfully
affected the cognitive abilities of these individuals. How-
ever, this lack of a significant finding may be found due to
a number of inadequately controlled for indirect behav-
iors acting as confounders, since there is previous environ-
mental evidence for both season's effects on cognition
and environmental illumination's effects on mood and
cognition in general populations. Of particular impor-
tance, it may be true that those who are non-depressed
may spend more time outside, thus receiving a more ade-
quate supply of environmental illumination
[17,19,21,22,45-47].
Because cognitive impairment is also associated with
other psychological and neurological disorders, discover-
ing the environment's impact on cognitive functioning
within the context of these disorders may lead not only to
better understanding of the disorders, but also to the
development of targeted interventions to enhance every-
day functioning and quality of life.
Abbreviations
BMI: Body Mass Index; CESD: Center for Epidemiologic
Studies Depression Scale; CMH: Cochran-Mantel-Haen-
zel; KJ/m2/day: kilojoules per meters2 per day; NASA:
National Aeronautics and Space Administration; NCAP:
National Center for Environmental Prediction (NCEP);
NARR: North American Regional Reanalysis; REGARDS:
REasons for Geographic And Racial Differences in Stroke;
OR: odds ratio; PST: Pacific Standard Time; PCS: Physical
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 13 of 14
(page number not for citation purposes)
Components Summary; SAD: Seasonal Affective Disorder;
SF-12: 12-item Short Form; SCN: the suprachiasmatic
nuclei; W/m2: Watts/meters2.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SK performed the analysis and drafted the manuscript. LM
was a mentor for the methods, statistical analyses, manu-
script editing, and data procurement. WC was a consult-
ant for environmental science, manuscript editing, and
procured and managed data. DA was a mentor for meth-
ods and manuscript editing. VW was a consultant for cog-
nitive function in the REGARDS dataset and manuscript
editing. NS was a mentor for the methods, statistical anal-
yses, and manuscript editing.
Authors' informations
SK is a PhD student in the Department of Epidemiology at
the University of Alabama at Birmingham (UAB) and has
done work with Marshall Space Flight Center in Hunts-
ville. LM is an Assistant Professor in the Department of
Biostatistics at UAB and is also currently working with
Marshall Space Flight Center. BC is a scientist working for
the Universities Space Research Association and the Mar-
shall Space Flight Center and has extensive experience
using satellite data to characterize earth environment var-
iables. VW is an Assistant Professor, works in the Depart-
ment of Psychology in the School of Medicine at UAB, is
the Director of the Dementia Care Research Program,
Assistant Director for Translational Research on Aging and
Mobility, and has previously used the cohort used in this
study for cognitive research. DA is a Professor in and the
chair of the Department of Epidemiology at UAB and has
extensive experience in cardiovascular and genetic
research, an example being the PI of the Genetics of Left
Ventricular Hypertrophy: HyperGEN study. NS is an Asso-
ciate Professor and an environmental and occupational
epidemiologist and pediatrician whose research interests
include cancer and infectious diseases epidemiology. Her
current research activities include epidemiologic studies
relating pesticide exposure and suicide, and of workers in
the rubber industry, plastics industry, and chemical man-
ufacturing facilities.
Acknowledgements
This research project is supported by a cooperative agreement U01
NS041588 from the National Institute of Neurological Disorders and
Stroke, National Institutes of Health, Department of Health and Human
Services. The content is solely the responsibility of the authors and does
not necessarily represent the official views of the National Institute of Neu-
rological Disorders and Stroke or the National Institutes of Health. Repre-
sentatives of the funding agency have been involved in the review of the
manuscript but not directly involved in the collection, management, analysis
or interpretation of the data The authors acknowledge the participating
investigators and institutions for their valuable contributions: The Univer-
sity of Alabama at Birmingham, Birmingham, Alabama (Study PI, Statistical
and Data Coordinating Center, Survey Research Unit): George Howard
DrPH, Leslie McClure PhD, Virginia Howard PhD, Libby Wagner MA, Vir-
ginia Wadley PhD, Rodney Go PhD, Monika Safford MD, Ella Temple PhD,
Margaret Stewart MSPH, J. David Rhodes BSN; University of Vermont
(Central Laboratory): Mary Cushman MD; Wake Forest University (ECG
Reading Center): Ron Prineas MD, PhD; Alabama Neurological Institute
(Stroke Validation Center, Medical Monitoring): Camilo Gomez MD,
Susana Bowling MD; University of Arkansas for Medical Sciences (Survey
Methodology): LeaVonne Pulley PhD; University of Cincinnati (Clinical
Neuroepidemiology): Brett Kissela MD, Dawn Kleindorfer MD; Examina-
tion Management Services, Incorporated (In-Person Visits): Andra Graham;
Medical University of South Carolina (Migration Analysis Center): Daniel
Lackland DrPH; Indiana University School of Medicine (Neuropsychology
Center): Frederick Unverzagt PhD; National Institute of Neurological Dis-
orders and Stroke, National Institutes of Health (funding agency): Claudia
Moy PhD.
Additional funding, data, data processing, and consultation were provided
by an investigator-initiated grant-in-aid from NASA. NASA did not have any
role in the design and conduct of the study, the collection, management,
analysis, and interpretation of the data or the preparation or approval of
the manuscript. The manuscript was sent to NASA Marshall Space Flight
Center for review prior to submission for publication.
References
1. Keller MC, Fredrickson BL, Ybarra O, Cote S, Johnson K, Mikels J,
Conway A, Wager T: A warm heart and a clear head. The con-
tingent effects of weather on mood and cognition. Psychol Sci
2005, 16:724-731.
2. Sinclair RC, Mark MM, Clore GL: Mood-related persuasion
depends on (mis)attributions. Social Cognition 1994, 12:309-326.
3. Winkler D, Pjrek E, Iwaki R, Kasper S: Treatment of seasonal
affective disorder. Expert Rev Neurother 2006, 6:1039-1048.
4. McColl SL, Veitch JA: Full-spectrum fluorescent lighting: a
review of its effects on physiology and health. Psychol Med
2001, 31:949-964.
5. Srinivasan V, Smits M, Spence W, Lowe AD, Kayumov L, Pandi-Peru-
mal SR, Parry B, Cardinali DP: Melatonin in mood disorders.
World J Biol Psychiatry 2006, 7:138-151.
6. Van Someren EJ, Lek RF Riemersma-Van Der: Live to the rhythm,
slave to the rhythm. Sleep Med Rev 2007, 11:465-484.
7. Howard VJ, Cushman M, Pulley L, Gomez CR, Go RC, Prineas RJ,
Graham A, Moy CS, Howard G: The reasons for geographic and
racial differences in stroke study: objectives and design. Neu-
roepidemiology 2005, 25:135-143.
8. Sinnott RW: Virtues of the Haversine. 1984, 68:159.
9. Wadley VG, McClure LA, Howard VJ, Unverzagt FW, Go RC, Moy
CS, Crowther MR, Gomez CR, Howard G: Cognitive status,
stroke symptom reports, and modifiable risk factors among
individuals with no diagnosis of stroke or transient ischemic
attack in the REasons for Geographic and Racial Differences
in Stroke (REGARDS) Study. Stroke 2007, 38:1143-1147.
10. Callahan CM, Unverzagt FW, Hui SL, Perkins AJ, Hendrie HC: Six-
item screener to identify cognitive impairment among
potential subjects for clinical research. Med Care 2002,
40:771-781.
11. Folstein MF, Folstein SE, McHugh PR: "Mini-mental state". A
practical method for grading the cognitive state of patients
for the clinician. J Psychiatr Res 1975, 12:189-198.
12. Ware J Jr, Kosinski M, Keller SD: A 12-Item Short-Form Health
Survey: construction of scales and preliminary tests of relia-
bility and validity. Med Care 1996, 34:220-233.
13. Melchiot LA, Huba GJ, Brown VB, Reback CJ: A short depression
index for women. Educational and Psychological Measurement 1993,
53:1117-1125.
14. Palinkas LA, Makinen TM, Paakkonen T, Rintamaki H, Leppaluoto J,
Hassi J: Influence of seasonally adjusted exposure to cold and
Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral
Environmental Health 2009, 8:34 http://www.ehjournal.net/content/8/1/34
Page 14 of 14
(page number not for citation purposes)
darkness on cognitive performance in circumpolar residents.
Scand J Psychol 2005, 46:239-246.
15. Blaiss MS: Cognitive, social, and economic costs of allergic
rhinitis. Allergy Asthma Proc 2000, 21:7-13.
16. Marshall PS, Colon EA: Effects of allergy season on mood and
cognitive function. Ann Allergy 1993, 71:251-258.
17. Paakkonen T, Leppaluoto J, Makinen TM, Rintamaki H, Ruokonen A,
Hassi J, Palinkas LA: Seasonal levels of melatonin, thyroid hor-
mones, mood, and cognition near the Arctic Circle. Aviat
Space Environ Med 2008, 79:695-699.
18. Lambert G, Reid C, Kaye D, Jennings G, Esler M: Increased suicide
rate in the middle-aged and its association with hours of sun-
light. Am J Psychiatry 2003, 160:793-795.
19. Miller AL: Epidemiology, etiology, and natural treatment of
seasonal affective disorder. Altern Med Rev 2005, 10:5-13.
20. Leonard BE, Myint A: Changes in the immune system in depres-
sion and dementia: causal or coincidental effects? Dialogues
Clin Neurosci 2006, 8:163-174.
21. Espiritu RC, Kripke DF, Ancoli-Israel S, Mowen MA, Mason WJ, Fell
RL, Klauber MR, Kaplan OJ: Low illumination experienced by
San Diego adults: association with atypical depressive symp-
toms. Biol Psychiatry 1994, 35:403-407.
22. Haynes PL, Ancoli-Israel S, McQuaid J: Illuminating the impact of
habitual behaviors in depression. Chronobiol Int 2005,
22:279-297.
23. Turner PL, Mainster MA: Circadian photoreception: ageing and
the eye's important role in systemic health. Br J Ophthalmol
2008, 92:1439-1444.
24. Walker MP, Stickgold R: Sleep-dependent learning and memory
consolidation. Neuron 2004, 44:121-133.
25. Khait VD, Huang YY, Malone KM, Oquendo M, Brodsky B, Sher L,
Mann JJ: Is there circannual variation of human platelet 5-
HT(2A) binding in depression? J Affect Disord 2002, 71:249-258.
26. Leppamaki S, Partonen T, Vakkuri O, Lonnqvist J, Partinen M, Laudon
M: Effect of controlled-release melatonin on sleep quality,
mood, and quality of life in subjects with seasonal or
weather-associated changes in mood and behaviour. Eur Neu-
ropsychopharmacol 2003, 13:137-145.
27. Jakovljevic M, Muck-Seler D, Pivac N, Ljubicic D, Bujas M, Dodig G:
Seasonal influence on platelet 5-HT levels in patients with
recurrent major depression and schizophrenia. Biol Psychiatry
1997, 41:1028-1034.
28. Golden RN, Gaynes BN, Ekstrom RD, Hamer RM, Jacobsen FM, Sup-
pes T, Wisner KL, Nemeroff CB: The efficacy of light therapy in
the treatment of mood disorders: a review and meta-analy-
sis of the evidence. Am J Psychiatry 2005, 162:656-662.
29. Srinivasan V, Pandi-Perumal SR, Cardinali DP, Poeggeler B, Hardeland
R: Melatonin in Alzheimer's disease and other neurodegen-
erative disorders. Behav Brain Funct 2006, 2:15.
30. Dani C, Bertini G, Martelli E, Pezzati M, Filippi L, Prussi C, Tronchin
M, Rubaltelli FF: Effects of phototherapy on cerebral haemody-
namics in preterm infants: is fibre-optic different from con-
ventional phototherapy? Dev Med Child Neurol 2004, 46:114-118.
31. Matthew E, Vasile RG, Sachs G, Anderson J, Lafer B, Hill T: Regional
cerebral blood flow changes after light therapy in seasonal
affective disorder. Nucl Med Commun 1996, 17:475-479.
32. Jennings JR, Muldoon MF, Price J, Christie IC, Meltzer CC: Cere-
brovascular support for cognitive processing in hypertensive
patients is altered by blood pressure treatment. Hypertension
2008, 52:65-71.
33. Beason-Held LL, Moghekar A, Zonderman AB, Kraut MA, Resnick
SM: Longitudinal changes in cerebral blood flow in the older
hypertensive brain. Stroke 2007, 38:1766-1773.
34. Elias MF, Sullivan LM, Elias PK, D'Agostino RB Sr, Wolf PA, Seshadri
S, Au R, Benjamin EJ, Vasan RS: Left ventricular mass, blood pres-
sure, and lowered cognitive performance in the Framing-
ham offspring. Hypertension 2007, 49:439-445.
35. Scarmeas N, Zarahn E, Anderson KE, Habeck CG, Hilton J, Flynn J,
Marder KS, Bell KL, Sackeim HA, Van Heertum RL, et al.: Associa-
tion of life activities with cerebral blood flow in Alzheimer
disease: implications for the cognitive reserve hypothesis.
Arch Neurol 2003, 60:359-365.
36. Fallon BA, Keilp J, Prohovnik I, Heertum RV, Mann JJ: Regional cer-
ebral blood flow and cognitive deficits in chronic lyme dis-
ease. J Neuropsychiatry Clin Neurosci 2003, 15:326-332.
37. Vandewalle G, Balteau E, Phillips C, Degueldre C, Moreau V, Sterpen-
ich V, Albouy G, Darsaud A, Desseilles M, Dang-Vu TT, et al.: Day-
time light exposure dynamically enhances brain responses.
Curr Biol 2006, 16:1616-1621.
38. Vandewalle G, Gais S, Schabus M, Balteau E, Carrier J, Darsaud A,
Sterpenich V, Albouy G, Dijk DJ, Maquet P: Wavelength-depend-
ent modulation of brain responses to a working memory
task by daytime light exposure. Cereb Cortex 2007,
17:2788-2795.
39. Chen S, Duckworth K, Chaiken S: Motivated Heuristic and Sys-
tematic Processing. Psychological Inquiry 1999, 10:44-49.
40. Bulbena A, Pailhez G, Acena R, Cunillera J, Rius A, Garcia-Ribera C,
Gutierrez J, Rojo C: Panic anxiety, under the weather? Int J
Biometeorol 2005, 49:238-243.
41. Hansen A, Bi P, Nitschke M, Ryan P, Pisaniello D, Tucker G: The
effect of heat waves on mental health in a temperate Aus-
tralian city. Environ Health Perspect 2008, 116:1369-1375.
42. Makinen TM, Palinkas LA, Reeves DL, Paakkonen T, Rintamaki H, Lep-
paluoto J, Hassi J: Effect of repeated exposures to cold on cog-
nitive performance in humans. Physiol Behav 2006, 87:166-176.
43. Huang J: Prediction of air temperature for thermal comfort of
people in outdoor environments. Int J Biometeorol 2007,
51:375-382.
44. Alperovitch A, Lacombe JM, Hanon O, Dartigues JF, Ritchie K, Duci-
metiere P, Tzourio C: Relationship between blood pressure and
outdoor temperature in a large sample of elderly individuals:
the Three-City study. Arch Intern Med 2009, 169:75-80.
45. Booker JM, Roseman C: A seasonal pattern of hospital medica-
tion errors in Alaska. Psychiatry Res 1995, 57:251-257.
46. Brennen T, Martinussen M, Hansen BO, Hjemdal O: Arctic cogni-
tion: a study of cognitive performance in summer and winter
at 69 degrees N. Appl Cogn Psychol 1999, 13:561-580.
47. Mills PR, Tomkins SC, Schlangen LJ: The effect of high correlated
colour temperature office lighting on employee wellbeing
and work performance. J Circadian Rhythms 2007, 5:2.
... Scientific evidence suggests that some mental disorders could be attributed to climate factors [9]. However, previous studies evaluating links between solar radiation and mood disorders are mixed and results vary by geographic region [10][11][12][13][14][15][16][17]. Some studies found beneficial effects of solar radiation on mood disorders including depression [10,13]. ...
... Most studies focused on general mood disorders. Some found beneficial effects of sunlight, and some did not [10][11][12][13][14][15]. These inconsistencies might be due to the definition of outcomes such as the failure of differentiating psychological distress from other mood disorders [28]. ...
... Besides, we found a potential U shape effect for SRAD7 where we observed more apparent beneficial effects for Q2 and Q3 but not Q4 associated with both outcomes. Similarly, Kent et al. [11] identified a U-shaped effect of solar radiation on depression [11]. In their study, the OR of the lowest radiation (<10,000 J/m 2 ) = 1.14, and the ORs of the radiation (10,000-25,000 J/m 2 ) = 0.87-0.94 ...
Article
Full-text available
Background Climate factors such as solar radiation could contribute to mood disorders, but evidence of associations between exposure to solar radiation and mood disorders is mixed and varies by region. Objective To evaluate the association of solar radiation with depression and distress among residents living in U.S. Gulf states. Methods We enrolled home-visit participants in the Gulf Long-Term Follow-up Study who completed validated screening questionnaires for depression (Patient Health Questionnaire-9, N = 10,217) and distress (Kessler Psychological Distress Questionnaire, N = 8,765) for the previous 2 weeks. Solar radiation estimates from the Daymet database (1-km grid) were linked to residential addresses. Average solar radiation exposures in the seven (SRAD7), 14 (SRAD14), and 30 days (SRAD30) before the home visit were calculated and categorized into quartiles (Q1-Q4). We used generalized linear mixed models to estimate prevalence ratios (PR) and 95% confidence intervals (CI) for associations between solar radiation and depression/distress. Results Higher levels of SRAD7 were non-monotonically inversely associated with depression [PRVs.Q1 (95%CI): Q2 = 0.81 (0.68, 0.97), Q3 = 0.80 (0.65, 0.99), Q4 = 0.88 (0.69, 1.15)] and distress [PRVs.Q1 (95%CI): Q2 = 0.76 (0.58, 0.99), Q3 = 0.77 (0.57, 1.06), Q4 = 0.84 (0.58, 1.22)]. Elevated SRAD14 and SRAD30 appeared to be associated with decreasing PRs of distress. For example, for SRAD14, PRs were 0.86 (0.63–1.19), 0.80 (0.55–1.18), and 0.75 (0.48–1.17) for Q2-4 versus Q1. Associations with SRAD7 varied somewhat, though not significantly, by season with increasing PRs of distress in spring and summer and decreasing PRs of depression and distress in fall. Impact Statement Previous research suffered from exposure misclassification, which impacts the validity of their conclusions. By leveraging high-resolution datasets and Gulf Long-term Follow-up Cohort, our findings support an association between increased solar radiation and fewer symptoms of mood disorders.
... We calculate these times at a monthly granularity at the average locations of the users in our dataset within Italy (sunset and sunrise times differ by less than an hour between any points on the map). Sunlight impacts many physiological and cognitive processes 46,48 , synchronising the human biological rhythm across the population group. Inferred waking time, also indicated by dashed vertical lines, is defined per cluster and represents the 16 continuous hours of highest aggregated activity. ...
... Generally, our findings are in line with previous literature detailing the link between cyclical behavioural patterns and Twitter use 41,59-61 as well as with findings associating sunlight with cognitive function (and by extension critical thinking) 46 and with activity on Twitter 45,47 . Similar patterns of diurnal activity archetypes have been identified in other studies. ...
Article
Full-text available
Social media manipulation poses a significant threat to cognitive autonomy and unbiased opinion formation. Prior literature explored the relationship between online activity and emotional state, cognitive resources, sunlight and weather. However, a limited understanding exists regarding the role of time of day in content spread and the impact of user activity patterns on susceptibility to mis- and disinformation. This work uncovers a strong correlation between user activity time patterns and the tendency to spread potentially disinformative content. Through quantitative analysis of Twitter (now X) data, we examine how user activity throughout the day aligns with diurnal behavioural archetypes. Evening types exhibit a significantly higher inclination towards spreading potentially disinformative content, which is more likely at night-time. This knowledge can become crucial for developing targeted interventions and strategies that mitigate misinformation spread by addressing vulnerable periods and user groups more susceptible to manipulation.
... Most studies find exposure to air pollution and increased precipitation are linked to more mental health problems, while green spaces can alleviate negative emotions (F. Chen et al., 2023;Kent et al., 2009;Leslie & Cerin, 2008;Sturm & Cohen, 2014). Our analytical framework also includes those typical natural environmental attributes. ...
... Chen et al., 2023;Liu et al., 2018;Von Lindern et al., 2016). Also, precipitation increases the potential for depression, since it reduces the time people can expose to sunlight and outdoor activities, thereby inhibiting serotonin production (Kent et al., 2009). All significant control variables support our hypotheses in Table 1, except for racial composition. ...
Article
Full-text available
Mental health disorders have become a global problem, garnering considerable attention. However, the root causes of deteriorating mental health remain poorly understood, with existing literature predominantly concentrating on socioeconomic conditions and psychological factors. This study uses multi‐linear and geographically weighted regressions (GWR) to examine the associations between built and natural environmental attributes and the prevalence of depression in US counties. The findings reveal that job sprawl and land mixed use are highly correlated with a lower risk of depression. Additionally, the presence of green spaces, especially in urban area, is associated with improved mental health. Conversely, higher concentrations of air pollutants, such as PM2.5 and CO, along with increased precipitation, are linked to elevated depression rates. When considering spatial correlation through GWR, the impact of population density and social capital on mental health displays substantial spatial heterogeneity. Further analysis, focused on two high depression risk clustering regions (northwestern and southeastern counties), reveals nuanced determinants. In northwestern counties, depression rates are more influenced by factors like precipitation and socioeconomic conditions, including unemployment and income segregation. In southeastern counties, population demographic characteristics, particularly racial composition, are associated with high depression prevalence, followed by built environment factors. Interestingly, job growth and crime rates only emerge as significant factors in the context of high depression risks in southeastern counties. This study underscores the robust linkages and spatial variations between built and natural environments and mental health, emphasizing the need for effective depression treatment to incorporate these multifaceted factors.
... Cognitive function along with physiological factors of bodily distress takes place due to seasonality and air pollution but must be supported by temperature and hazardous pollutant-induced weather effects (Kent et al. 2009 There are mixed results on the winter and summer seasons' effect on reaction times, memory, physiological processes, vigilance, and reasoning (Dai et al. 2016). ...
Article
Full-text available
ims. The goal of this meteorological psychology study is to understand the behavioral implications on the professional lives of medical professionals in one of the most polluted regions of the world with a compromised air quality index. There is a paucity of literature to explore climate ethics as to how behavioral aspects of morals may alter related to climate change. Methods. The study is a cross-sectional correlational quantitative study formulated through purposive sampling. Research took place during the smog season of the world’s most polluted city. The research instruments of Climate anxiety, pain and distress, and cognitive failure were administered to a sample of 211 medical professionals. Confirmatory factor analysis was conducted and resulted in robust model fitness with indices between .80 and .90. Item loadings for three factors were statistically ideal of greater than. 30. Barron and Kenny’s approach was applied to conduct mediation through path analysis in which climate anxiety significantly mediates between pain distress and cognitive failure (β = .51; p < .0001). Multiple regression results using bootstrap confidence showed that climate anxiety (β = .993; p < .0001), and pain distress result in cognitive failure. The independent sample t-test was carried out to explore if nurses and doctors are different in terms of experiencing climate anxiety, pain distress, and cognitive failure. Results. There is a strong significant association among climate anxiety, pain distress, and cognitive failure with strong to moderate correlation values, p < .001. Confirmatory factor analysis was conducted and resulted in robust model fitness with indices between .80 and .90. Item loadings for three factors were statistically ideal at .30. 102 Nurses (M = 85.216; SD = 4.94) were more painfully distressed, prone to cognitive failure, and have climate anxiety as compared to 109 doctors (M = 71.211; SD = 9.23).
... As such, trees with these characteristics were increasingly preferred, with increase to PET. However, practically, these trees block more sunlight and can result in psychological stress (Kent et al., 2009). As a result, very few people preferred short trunks at sites with TH of less than 2 m. ...
... Numerous studies have shown that properly dosed light exposure can significantly improve concentration, cognitive abilities, and mood (Fisk et al., 2018), (Kent et al., 2009), (Siraji et al., 2022). Studies have also demonstrated that HCL can improve academic performance in students by positively affecting their ability to concentrate and their overall well-being, particularly in schools in Sweden, Denmark, Norway, and the Netherlands, where these technologies were first implemented (Y. ...
Conference Paper
Full-text available
The paper synthesizes the results of previously published studies on the impact of lighting design in the interiors of educational facilities, and provides a comprehensive overview of the current state of knowledge in the field of Human Centric Lighting (HCL) application, identifying deficiencies, gaps, and controversies in the existing literature. The topic is an analysis of how adapting classroom lighting to natural light cycles can enhance student learning and well-being. Key findings indicate promising evidence that Human Centric Lighting (HCL) systems, simulating natural day and night cycles by adjusting lighting parameters, can improve a ention, working memory, information processing speed, executive functions, and overall academic performance of students. However, gaps in understanding long-term effects and individual differences have been identified. The analysis provides guidelines for implementing HCL systems in the interior design of educational facilities and suggests future research directions, including the development of personalized solutions and the application of advanced technologies. The findings point to the need for wider implementation of HCL technologies in educational institutions as a standard part of the infrastructure, which can have far-reaching positive effects on the educational system. The discussion also identifies gaps in current research and suggests directions for future research, particularly in the context of individualized lighting and its integration with other elements of interior design.
... The recommendation for adequate sun exposure in fair-skinned people to synthesize vitamin D is unprotected sun exposure for 2-14 min, outside of the peak UV period of 10am − 3pm, 3-4 times per week [43,49]. The benefits of sun exposure are not limited to vitamin D, but it also helps to regulate blood pressure, heart rate [50], and cognitive function [51]. Epimerisation decreases with age [19,20]. ...
Article
Background Several studies have explored the association between temperature and cognitive function. However, few studies have examined the effect of extreme temperature on cognitive function. In this study, we aimed to quantify the long‐term effect of extreme temperature (e.g., heat waves, cold spells, and hot night excess (HNE)) on cognitive performance in middle‐aged and older people in China. Method We investigated 7915 aged >45 years people from the China Health and Retirement Longitudinal Study (CHARLS), surveyed in 2011 and 2015. A structured questionnaire was utilized to assess cognitive function, including four dimensions: episodic memory, attention, orientation, and visuo‐construction. Hourly ambient temperature from the ERA5‐Land datasets were used to calculate variables indicating extreme temperature. We performed difference‐in‐difference (DID) models to assess the potential causal relationship between extreme temperature and cognitive function. Results Non‐linear analyses suggested that both sustained increases in temperature and excessive variability in temperature increased the risk of cognitive decline. Meanwhile, we observed the extra risk of global cognitive function decline was 2.3% (95% Confidence interval (95% CI): 0.2%, 4.4%) for heat waves (one unit increase) and 5.9% (95% CI: 0.6%, 11.6%) for HNE (one unit increase), while the association for cold spells was insignificant. Two cognitive dimensions, episodic memory and visuo‐construction, were sensitive to these two heat‐related factors. Conclusion Extreme temperature was inversely related to cognitive performance in middle‐aged and older adults, which was substantial for heat waves and HNE particularly. The effect size varied by cognitive dimensions.
Article
Full-text available
Background Depression rates among adolescents have risen dramatically over the past decade. Therefore, preventing depression among adolescents is particularly important. Differences in lifestyle habits may play a role in depression. Purpose This study aimed to explore the influence of living habits on depression levels among rural middle school students in Northeast China and to provide a theoretical basis for developing interventions to reduce depression levels in middle school students. Methods A total of 296 middle school students aged 13–15 years from Benxi City, Northeast China completed the anthropometric measurements, Physical Activity Scale-3 (PARS-3), and the Self-Rating Depression Scale (SDS). Their average screen time in the most recent week, parents' education level, and monthly family income were collected through a questionnaire. Results Females had higher depression scores than males (41.0 ± 6.9 vs. 37.9 ± 8.0). Physical activity (β = −0.38, t = −7.06, P < 0.01), family income (β = −0.20, t = −4.07, P < 0.01), screen time (β = 0.16, t = 3.34, P < 0.01), age (β = 0.15, t = 3.16, P < 0.01), sex (β = −0.13, t = −2.74, P < 0.01), and sleep quality (β = −0.08, t = −1.87, P < 0.01) are important factors related to depression levels. Conclusion The preliminary analysis results showed that among middle school students in rural Northeast China, the depression level of females was significantly higher than that of males. Poor quality sleep, low levels of physical activity, low household income, and long screen time were positively associated with depression. Therefore, strengthening physical activity, improving sleep quality, and reducing screen time are of clinical relevance in preventing and reducing depression.
Article
Full-text available
Regression methods were used to select and score 12 items from the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) to reproduce the Physical Component Summary and Mental Component Summary scales in the general US population (n = 2,333). The resulting 12-item short-form (SF-12) achieved multiple R squares of 0.911 and 0.918 in predictions of the SF-36 Physical Component Summary and SF-36 Mental Component Summary scores, respectively. Scoring algorithms from the general population used to score 12-item versions of the two components (Physical Component Summary and Mental Component Summary) achieved R squares of 0.905 with the SF-36 Physical Component Summary and 0.938 with the SF-36 Mental Component Summary when cross-validated in the Medical Outcomes Study. Test-retest (2-week) correlations of 0.89 and 0.76 were observed for the 12-item Physical Component Summary and the 12-item Mental Component Summary, respectively, in the general US population (n = 232). Twenty cross-sectional and longitudinal tests of empirical validity previously published for the 36-item short-form scales and summary measures were replicated for the 12-item Physical Component Summary and the 12-item Mental Component Summary, including comparisons between patient groups known to differ or to change in terms of the presence and seriousness of physical and mental conditions, acute symptoms, age and aging, self-reported 1-year changes in health, and recovery from depression. In 14 validity tests involving physical criteria, relative validity estimates for the 12-item Physical Component Summary ranged from 0.43 to 0.93 (median = 0.67) in comparison with the best 36-item short-form scale. Relative validity estimates for the 12-item Mental Component Summary in 6 tests involving mental criteria ranged from 0.60 to 1.07 (median = 0.97) in relation to the best 36-item short-form scale. Average scores for the 2 summary measures, and those for most scales in the 8-scale profile based on the 12-item short-form, closely mirrored those for the 36-item short-form, although standard errors were nearly always larger for the 12-item short-form.
Article
Full-text available
Previous research has shown that respondents in sad moods are likely to elaborate persuasive messages and to be highly persuaded by strong but not weak arguments. In contrast, respondents in happy moods are less likely to elaborate persuasive messages and hence show equal, moderate, persuasion by both strong and weak arguments. To examine how mood-related persuasion effects might be moderated by attributions concerning the source of a person's mood, students were approached on days with either pleasant or unpleasant weather and were asked to participate in a survey relating to higher education. Participants either were or were not cued to the weather as a source of their mood. They listened to either strong or weak arguments supporting the implementation of comprehensive examinations for graduating students. Attitudes toward comprehensive exams were measured. Participants reported more positive moods on pleasant days than on unpleasant days. When the weather was not made salient, weather (i.e., mood) interacted with argument strength: When weather was unpleasant, argument strength had an effect on attitudes, but when weather was pleasant, argument strength had no effect. In contrast, when participants were cued to the weather, there were no weather effects on persuasion; instead, attitudes were affected only by argument strength. Implications, mechanisms, and directions for future research are discussed.
Article
Background. Full-spectrum fluorescent lighting (FSFL) has been credited with causing dramatic beneficial effects on a wide variety of behaviours, mental health outcomes and physical health effects, as compared to other fluorescent lamp types. These effects are hypothesized to occur because of similarity between FSFL emissions and daylight, which is said to have evolutionary superiority over other light sources. Method. This review, covering the period 1941-1999, critically considers the evidence for direct effects of FSFL through skin absorption as well as indirect effects on hormonal and neural processes. Conclusions. Overall, the evidence does not show dramatic effects of fluorescent lamp type on behaviour or health, neither does it support the evolutionary hypothesis.
Article
The CES-D is a well-known index of acute depressive symptoms experienced over a 7-day period that has been used in literally hundreds of studies. This article presents the psychometric derivation of 8- and 4-item screening versions of the CES-D for use in research with community-based samples. These short depressive symptom indices can be used in those instances where a brief assessment is needed for broad screening or research purposes. Using data from a heterogeneous community sample of 411 women, the 8-item CES-D was found to correlate .93 with the full 20-item CES-D while the 4-item CES-D was found to correlate .87 with the full CES-D. In a second sample of 83 women in a residential drug abuse program, the 8- and 4-item measures correlated .54 and .47, respectively, with the BPI Depression scale.
Article
Article
Evidence has accumulated over the past 15 years that a€ect in humans is cyclical. In winter there is a tendency to depression, with remission in summer, and this e€ect is stronger at higher latitudes. In order to determine whether human cognition is similarly rhythmical, this study investigated the cognitive processes of 100 participants living at 698N. Participants were tested in summer and winter on a range of cognitive tasks, including verbal memory, attention and simple reaction time tasks. The seasonally counterbalanced design and the very northerly latitude of this study provide optimal conditions for detecting impaired cognitive performance in winter, and the conclusion is negative: of ®ve tasks with seasonal e€ects, four had disadvantages in summer. Like the menstrual cycle, the circannual cycle appears to in¯uence mood but not cognition. Copyright # 1999 John Wiley & Sons, Ltd. There are three major astronomical cycles that have a period short enough to be important in the study of behaviour. The ®rst is the 24-hour rotation of the Earth on its axis, the second is the monthly rotation of the moon about the Earth and the third is the 12-month orbit of the Earth around the Sun. While cognitive psychologists have studied the daily cycle using time-of-day as an experimental variable (e.g. Smith, 1989) and the monthly cycle in particular in connection with studies of the menstrual cycle (e.g. Richardson, 1991a), the annual cycles have been seldom studied from a cognitive perspective. An investigation of the e€ect of the seasons upon normal cognitive performance is timely because evidence is accumulating of seasonal e€ects on the incidence of mood-related problems. Rosenthal et al. (1984) named this disorder Seasonal A€ective Disorder (SAD). People who su€er from this disorder have symptoms of depression that only occur during one part of the year, but that recur from year to year. The vast majority of SAD patients su€er in winter, with remission in summer, and frequently reported symptoms include sadness, anxiety, irritation, decreased activity and libido, leading both to social problems and diculty at work.