Psychology
2012. Vol.3, Special Issue, 848-855
Published Online September 2012 in SciRes (http://www.SciRP.org/journal/psych) http://dx.doi.org/10.4236/psych.2012.329128
Copyright © 2012 SciRes.
848
Seasonal Mood and Behavioral Changes for Japanese Residents
in the United Kingdom
Yumiko Kurata1, Yumiko Nomura2
1Graduate School of Human Sciences, Waseda University, Saitama, Japan
2Faculty of Human Sciences, Waseda University, Saitama, Japan
Email: kurata@asagi.waseda.jp
Received July 10th, 2012; revised August 10th, 2012; accepted September 5th, 2012
The aim of this study was to investigate seasonal changes in mood and behavior for Japanese residents in
UK A questionnaire survey was conducted with Japanese residents in the UK (n = 100) who participated
both a combination winter and summer research. First, a longitudinal study comparing two surveys—one
in summer and another in winter—was carried out to determine how the level of seasonal changes influ-
enced depression among Japanese living in the UK. Then, we examined seasonal changes in mood and
behavior over a 12-month period based on the degree of seasonal dependence. Paired t-tests on Global
Seasonality Score (GSS score) and the Center for Epidemiologic Studies Depression Scale (CES-D)
scores by winter and summer demonstrated that each score had a significant seasonal difference; individ-
ual scores were higher in winter than in summer. We examined the difference between high seasonality
group, medium seasonal group, and non-seasonal group, regarding to the winter CES-D and summer
CES-D scores. The ANOVA revealed a significant difference on the winter score (Winter: F(2,97) = 4.62,
p < .01, Summer: F(2,97) = 3.24, p < .05). Although we did not find any interaction between seasonal
change and season, the main effect was significant for season. The results showed fluctuations in which
mood, social activity, and sleep all declined during the winter and then improved during the summer. It
indicated that depressive symptoms among Japanese living in the UK fluctuate due to seasonality; over a
period of 12 months, their mood and behavior declined during winter and improved during summer. As
described, Japanese living in the UK experience environmental changes due to seasonality. This suggests
that the environmental factor called seasonal change can partly explain why Japanese living in the UK
suffer from mental and physical disorders. Mental health measures specific to the local environment are
necessary to support individuals to adapt to and live under an environment different from home country.
Keywords: Seasonality; Seasonal Change; Depression
Introduction
Seasonal affective disorder (SAD) is a syndrome character-
ized by recurrent depression that occurs annually at the same
time each year (Rosenthal, Sack, Gilin, Lewy, Goodwin, Dav-
enport, Mueller, Newsome, & Wehr, 1984). Typically, episodes
begin in the fall or winter and remit in the spring (Rosen, Tar-
gum, Terman, Bryant, Hofman, Kasper, Hamovit, Decherty,
Welch, & Rosenthal, 1990). Rates of winter SAD and sub-syn-
dromal winter SAD (S-SAD) were significantly higher at the
more northern latitudes (Rosen, Targum, Terman, Bryant, Hofman,
Kasper, Hamovit, Decherty, Welch, & Rosenthal, 1990; Magnus-
son & Stefansson, 1993; Magnusson & Axelsson, 1993).
An analysis of seasonal changes in mood, emotion, sleep,
and activity levels of people in four regions in Japan with vary-
ing latitudes showed that the fluctuations were large in the
high-latitude region (Shirakawa, Okawa, Uchiyama, Oguri, Kos-
aka, Mishima, Inoue, & Kamei, 1993). Seasonal changes in be-
havior occur in Japan, but at a lower frequency and with a dif-
ferent profile than in the US or Europe (Ozaki, Ono, Ito, & Ro-
senthal, 1995). Seasonal depression peaked in three groups as
indigenous white, non-indigenous Asian and Asian-British,
with the incidence of winter depression being highest in the
Asian group (Suhail & Cochrane, 1997). Then, since Asian
group showed significant seasonal variation in depression, with
a greater number of depressive episodes in winter, gender and
ethnicity may be major risk factors in winter low mood (Suhail
& Cochrane, 1998). Japanese participants staying in Stockholm
showed more mental and somatic depressive symptoms in the
winter than in the summer. And the Beck Depression Inventory
(BDI) score was higher in participants who had lived in Stock-
holm for more than 10 years, compared with newcomers who
lived in the same city for less than 2 years (Murase, Murase,
Kitabatake, Yamauchi, & Mathé, 1995).
On the basis of these factors, effects on mood and emotions
could be predicted when Japanese reside over a long period of
time in high-latitude regions with seasonal changes.
Along with the increase in the number of Japanese people
living overseas on a long-term basis, the prevalence of mental
disorders, mental illnesses, and suicides in such people has
increased (Ministry of Foreign Affair of Japan, 2010). In some
overseas regions, they may end up living in an environment
different from the one home country. Mood and behavior may
fluctuate among people in high-latitude regions, where seasonal
effects are greater. Exploring the environmental factors affect-
ing mood disorders and mood changes in Japanese people liv-
ing in a high-latitude region on a long-term basis might provide
valuable insight into the pathogenesis, treatment, and preven-
tion of mental illness.
Y. KURATA, Y. NOMURA
Therefore, we conducted a longitudinal study to compare the
seasonal changes in summer and winter, aiming to understand the
its impact on mood and behavior among Japanese living in the UK.
In Study 1, a longitudinal study comparing two surveys—one
in summer and another in winter—was carried out to determine
how the level of seasonal change influenced depression among
Japanese living in the UK. In Study 2, we examined seasonal
fluctuations in mood and behavior over a 12-month period
based on the degree of seasonal dependence.
Material and Methods
Study 1
Sample and Procedure
We distributed the questionnaire to Japanese men and wo-
men over the age of 18 who resided in major areas in the UK
(total of 100 individuals: 22 men, average age = 40.55, SD =
9.64; 78 women, average age = 41.28, SD = 7.78) and asked
them to respond in both summer and winter. Recruitment was
performed through Japanese companies and Japanese cram
schools, and survey forms were sent and collected either by
post or by mail. Research consent was obtained after providing
participants with written information regarding the research
purpose, privacy protection policy, and the voluntary nature of
participation. The participants were instructed to fill out the
questionnaire anonymously.
Scales
The Seasonal Pattern Assessment Questionnaire (SPAQ)
(Rosenthal & Wehr, 1987) has several parts covering topics
such as the patterns of seasonal dependence, degree of seasonal
dependence (Global seasonality score: GSS score), degree of
difficulty with seasonal fluctuation, and the degree of that difficulty,
and also includes questions about hours of sleep by season.
This study used the Japanese version of the SPAQ (Shira-
kawa et al., 1993), a standard scale used in Japan for the epi-
demiological study of SAD.
In this study, the GSS score from the Japanese SPAQ was
used as an indicator that measures seasonal fluctuations in
mood and behavior. GSS score is the central component of the
seasonal pattern Assessment Questionnaire (SPAQ). The GSS
retrospectively measures seasonal variations on six different
items which are sleep, mood, weight, sleep length, social activ-
ity and appetite. The GSS are scaled 0 for no seasonality, 1for
mild, 2 for moderate, 3 for marked and 4 for extremely marked
seasonal change. The total score is ranging between 0 to 24.
The cut-off level of the GSS are set to 8 - 10 for sub-SAD and
11 for SAD (Rosenthal et al., 1987).
The Center for Epidemiologic Studies Depression Scale
(CES-D) Japanese version (Shima, Shikano, Kitamura, & Asai,
1985) is a 20-item instrument with a 4-point rating system and
is a translation of the same scale developed by the National
Institute of Mental Health (NIMH). These 20 items are intended
screen for depression in the general population. CES-D items
are rated with regard to frequency: “Rarely or never (less than
once a day)”, “Some or a little of the time (1 - 2 days)”, “Occa-
sionally or a moderate amount of time (3 - 4 days)”, and “Most
or all of the time (5 - 7 days)”.
Statics
First, in order to examine the seasonal fluctuations in indi-
vidual scores, paired t-tests were performed on the GSS and
CES-D scores to compare winter and summer scores. In addi-
tion, t-tests were performed to determine gender differences in
the GSS and CES-D scores. Next, comparisons using an analy-
sis of variance (ANOVA) were performed to check whether the
length of stay affected the GSS and CES-D scores. Length of
stay was compared by dividing participants into four groups:
less than three years, less than five years, less than ten years,
and ten years or more. Subsequently, we checked for differ-
ences in depressive symptoms during summer and winter based
on seasonal dependence (GSS scores). This was done by creat-
ing groups based on the winter GSS score—high group (11
points or higher), medium group (8 to 10 points), and low
group (7 points or lower)—and conducting a one-way ANOVA
to test the effect on the winter and summer CES-D scores. Later,
various comparisons were made using Tukey’s method.
Finally, a two-way ANOVA was conducted to test the effect
of two factors—seasonal changes (GSS score) and season (sum-
mer and winter)—on the CES-D scores in summer and winter.
On the basis of their GSS scores, participants were categorized
into a low group (lower than 8 points) or high group (8 points
or higher).
Study 2
Sample and Procedure
A questionnaire survey was administered to 100 Japanese
residents in the UK who participated in both the winter and
summer research. Study 2 used the GSS scores and the table of
seasonal mood and behavior patterns. The Seasonal Pattern
Questionnaire covers topics such as seasonal patterns in mood
and behavior.
The Seasonal Pattern Questionnaire contained the following
ten items: 1) Feel best, 2) Gain most weight, 3) Socialize most,
4) Sleep most, 5) Eat most, 6) Lose most weight, 7) Socialize
least, 8) Feel worst, 9) Eat least, 10) Sleep least. These items
may refer to a single month, a cluster of months, or no particu-
lar month that stands out as extreme on a regular basis.
Statics
Responses 1 through 5 were given the score of 1, and re-
sponses 6 through 10 were given the score of –1. To compare
those scores, the high, medium, and low groups were created on
the basis of the seasonal assessment score (winter GSS score),
and t-tests were performed.
PASW for Windows version 18.0 was used for statistical
analysis. The demographic characteristics of the participants in
Studies 1 and 2 are shown in Table 1.
Results and Discussion
Study 1
Paired t-tests on GSS scores and CES-D scores by winter and
summer demonstrated that each score had a significant seasonal
difference; individual scores were higher in winter than in
summer (Table 2).
There was no significant gender difference in CES-D and
GSS scores in summer or winter. Further, a relationship be-
tween seasonal dependence, depression, and the period of resi-
dence could not be identified in either the summer period or the
winter period.
Copyright © 2012 SciRes. 849
Y. KURATA, Y. NOMURA
Copyright © 2012 SciRes.
850
Table 1.
Demographic features of participants.
Number n %
Total 100 100
Male 22 22.00
Female 78 78.00
Age n %
Under 29 7 7.00
30’s 31 31.00
40’s 49 49.00
50’s 13 13.00
Mean age years SD
Total 41.12 8.18
Male 40.55 9.64
Female 41.28 7.78
Duration of stay n %
<3 years 27 27.00
<5 years 22 22.00
<10 years 27 27.00
Over 10 years 24 24.00
Mean months of residence months SD
Total 88.79 86.17
Male 82.23 52.40
Female 90.64 93.71
Marital status n %
Married 17 17.00
Unmarried 76 78.00
Others 7 7.00
Children n %
With children 61 61.00
Without children 39 39.00
Occupation n %
Student 4 4.00
Housewife 43 43.00
Employee (full time) 38 38.00
Employee (part time) 7 7.00
Others 8 8.00
Next, we examined the difference between the high seasonal-
ity group, the medium seasonal group, and the non-seasonal
group, regarding to the CES-D scores in winter and summer.
The ANOVA revealed a significant difference on the winter
score. Multiple comparisons using Tukey’s method showed a
significant difference between the high group and low group
during both winter and summer (Winter: F(2,97) = 4.62, p < .01,
Summer: F(2,97) = 3.24, p < .05) (Figures 1 and 2).
In addition, we conducted a two-way ANOVA to test the ef-
fect of two factors—seasonal dependence (GSS score) and sea-
son (summer and winter)—on the CES-D score in summer and
winter. Although we did not find any interaction between sea-
sonal change and season, the main effect was significant at
the .1% level for season (Figure 3).
Study 2
An analysis to determine whether seasonal fluctuations in
mood and behavior differ over a year indicated significant dif-
ferences in mood during October, November, and July; in
weight during January and December; in social activity during
April; in sleep during January and February; and in appetite
during January and December.
In terms of mood change, results indicated that the high-GSS
group tends to feel worse compared to the low-GSS group dur-
ing October and November, when the summer has ended and
changes such as shortened daylight hours and deteriorations in
weather are observed. They also indicated that the high-GSS
group tends to feel better than the low-GSS group during July.
As for changes in weight, the results indicated that it tends to
Table 2.
Paired t-test on the GSS score and CES-D.
Mean
Summer Winter t-Value
GSS 6.63 5.73 2.84 ***
CESD 22.5 10.67 16.04 **
Note: ***p < .001, ***p < .01; GSS: Global seasonality score; CES-D: The Center
for Epidemiologic Studies Depression Scale Japanese version.
Figure 1.
Mean CES-D for High Seasonal group, Medium Seasonal group and Non-Seasonal group by GSS
score (winter).
Y. KURATA, Y. NOMURA
Figure 2.
Mean CES-D for High Seasonal group, Medium Seasonal group and Non-Seasonal group by GSS
score (summer).
Figure 3.
Graphic representation of the CES-D score in winter and summer by two-factor factorial ANOVA.
increase for the high-GSS group compared to the low-GSS
group during January and December. The changes in social
activity indicated that they tend to be better for the high-GSS
group than the low-GSS group during April. Changes in sleep
indicated that the high-GSS group tends to sleep longer than the
low-GSS group during January and February. The results indi-
cated that appetite tends to increase during January and De-
cember for the high-GSS group. The result of Study 2 is shown
in Figures 4 (a)-(e) and Table 3.
Discussion
This study employed a longitudinal design in which two ques-
tionnaire surveys—one in summer, another in winter—were
conducted to examine depression. Seasonal change within indi-
viduals and aggravation of depression in Japanese residents in
the winter was confirmed. In agreement with the findings of
other studies, the total CES-D score was significantly higher in
winter than in summer. The study also indicated that individu-
als who are more susceptible to seasonal change are likely to
develop more severe depressive symptoms in winter than in
summer. In addition, depressive symptoms among Japanese
living in the UK were more influenced by the season than the
susceptibility to seasonal change.
These results suggested that seasonal change has an impor-
tant effect on mood among Japanese living in the UK. Further,
this study confirmed that depression in Japanese long-term re-
sidents of the UK is affected by the season regardless of the
length of their residence in the UK.
The results of the present study are not in agreement with
previous findings of a study in which participants who had
lived in Stockholm for more than 10 years had higher BDI
Copyright © 2012 SciRes. 851
Y. KURATA, Y. NOMURA
(a)
(b)
(c)
Copyright © 2012 SciRes.
852
Y. KURATA, Y. NOMURA
(d)
(e)
Figure 4.
Monthly difference of Mood, Weight, Social Activity, Sleep Length and Appetite in participants
with high (Global Seasonality Score 11) or low (Global Seasonality Score < 11) seasonality
group. GSS means Global Seasonality Score. (a) Monthly variation of “Mood” between GSS
group; (b) Monthly variation of “Weight” between GSS group; (c) Monthly variation of “Social
Activity” between GSS group; (d) Monthly variation of “Sleep Length” between GSS group; (e)
Monthly variation of “Appetite” between GSS group.
scores than those who had lived in the same city less than 2
years. The difference based on the length of stay may not have
arisen in the present study because we used a different scale to
measure depressive symptoms and because Stockholm is lo-
cated at higher latitude than the UK. Since it has been shown
that depressive symptoms among Japanese living in the UK are
exacerbated during the winter, we believe mental health meas-
ures that take seasonal changes into account are necessary re-
gardless of the length of stay.
In this study, no gender differences in depression were found.
However, previous research has indicated that women are more
vulnerable than men to winter depression (Suhail et al., 1997;
Kasper, Wehr, Bartko, & Rosenthal, 1989; Michalak, Wilkin-
son, Dowrick, & Wilkinson, 2001).
The previous findings accord with data showing a lower
prevalence of and a lower female-male ratio in seasonal affec-
tive disorder in Japan than in Western countries (Shiakamoto,
Kamo, Nakadaira, Tamura, & Takahashi, 1993). Gender may
be major risk factor in winter low mood (Suhail et al., 1998). In
this study, longitudinal research was conducted using two ques-
tionnaire surveys, but the number of participants from both
sexes was unequal; hence, gender differences may have been
Copyright © 2012 SciRes. 853
Y. KURATA, Y. NOMURA
Table 3.
Related independent t-test on the seasonal pattern score.
January February March April May June
t-value t-value t-value t-value t-value t-value
Mood 8.50 1.69 .57 .98 .88 1.48
Weight 2.6** .97 .75 1.31 .96 .66
Social activity 1.40 .56 1.80 2.05* .43 .67
Sleep 2.95** 2.32* .59 .21 .15 1.91
Appetite 2.43* .19 .77 1.88 .51 .49
July August September October November December
t-value t-value t-value t-value t-value t-value
Mood 3.46*** 2.21 1.23 1.93* 2.47* 1.39
Weight .85 1.58 .77 1.23 1.74 2.61**
Social activity 1.86 .58 .80 .33 .76 .93
Sleep 1.71 1.78 1.38 .62 1.02 1.25
Appetite 1.19 .82 1.08 1.56 1.56 3.03**
Note: *p < .05, **p < .01, ***p < .001.
obscured. It is not conclusive whether the results of this study
are specific to the Japanese or due to the gender bias in the
sample.
In Study 2, we examined how mood and behavior fluctuate
over a period of 12 months among Japanese living in the UK.
The results showed fluctuations in which mood, social activity,
and sleep all declined during the winter and then improved
during the summer. Based on a study of the general population
in Japan, Shirakawa et al. (1993) reported that the high-seasonal-
ity group experienced larger fluctuations in mood, social life,
sleep, and body weight compared to the low-seasonality group.
Our study also demonstrated that the high group had larger
seasonal fluctuations than the low group did. In particular, the
high group experienced a considerable mood decline during the
winter, and their mood improvement during the spring and sum-
mer was more evident than that of the low group. Similar to
studies within Japan, our study showed seasonal fluctuations in
mood and behavior. However, while people in both Japan and
the UK experienced a decline in mood and behavior in winter,
people experienced improvement in summer for the UK and au-
tumn for Japan (Shirakawa et al., 1993; Kamo, Kamo, Nakadaira,
& Sakamoto, 1993). In addition, while a seasonal mood decline
was observed twice in Japan, from the rainy season to summer
and again during winter, mood declined only during the winter
in the UK. Seasonal mood fluctuations differ in the UK and
Japan. This is probably because the UK is located at higher
latitude than Japan, which would lead to different seasonal
changes such as shortened daylight hours and poor weather.
Conclusion
We found statistically significant changes in mood according
to season in Japanese residents of the UK in both the summer
and winter studies. In particular, individuals who are more sus-
ceptible to seasonal change experience considerably worse
depressive symptoms. As described, Japanese living in the UK
experience environmental changes due to seasonality. This
suggests that the environmental factor called seasonal change
can partly explain why Japanese living in the UK suffer from
mental and physical disorders. Mental health measures specific
to the local environment are necessary to support individuals to
adapt to and live under an environment different from home
country.
However, a limitation of this study is the skewed ratio: the
number of female respondents was higher. Collecting more data
by inviting participants without any gender bias will be a key to
future research. Hereafter, an increase in the number of exami-
nees should be considered.
Furthermore, although Study 2 demonstrated seasonal mood
and behavior fluctuations among Japanese living in the UK, we
have not directly compared the findings with those of previous
studies. Future studies should investigate differences in sea-
sonal fluctuations in the UK and Japan or those experienced by
Japanese living in other countries in order to further clarify the
effect of seasonal environment changes.
REFERENCES
Haggag, A., Eklund, B., Linaker, O., & Gotestam K. G. (1990). Sea-
sonal mood variation: An epidemiological study in northern Norway.
Acta Psychiatrica Scandinavica, 81, 141-145.
doi:10.1111/j.1600-0447.1990.tb06467.x
Kamo, T., Kamo, K., Nakadaira, S., & Sakamoto, K. (1993). A Inves-
tigation of seasonality in mood and behavior in normal subjects using
the seasonal pattern assessment questionnaire. Japanese Journal of
Psychiatry, 35, 837-840.
Kasper, S., Wehr, T. A., Bartko, J. J., & Rosenthal, N. E. (1989). Epi-
demiological findings of seasonal changes in mood and behavior: A
telephone survey of Montgomery Country, Maryland. Archives of
General Psychiatry, 46, 823-833.
doi:10.1001/archpsyc.1989.01810090065010
Magnusson, A., & Stefansson, J. G. (1993). Prevalence of seasonal af-
fective disorder in Iceland. Archives of General Psychiatry, 50, 941-
946. doi:10.1001/archpsyc.1993.01820240025002
Magnusson, A., & Axelsson, J. (1993). The prevalence of seasonal
affective disorder is low among descendants of Icelandic emigrants
in Canada. Archives of General Psychiatry, 50, 947-951.
doi:10.1001/archpsyc.1993.01820240031004
Michalak, E. E., Wilkinson, C., Dowrick, C., & Wilkinson, G. (2001).
Seasonal affective disorder: Prevalence, detection and current treat-
ment in North Wales. The British Journal of Psychiatry, 179, 31-34.
doi:10.1192/bjp.179.1.31
Ministry of Foreign Affairs of Japan (2010). Annual report of statistics
on Japanese nationals overseas. Tokyo: Ministry of Foreign Affairs
of Japan.
Ministry of Foreign Affairs of Japan (2010). Kaigai houjin yougo tou-
kei. Tokyo: Ministry of Foreign Affairs of Japan.
Copyright © 2012 SciRes.
854
Y. KURATA, Y. NOMURA
Murase, S., Murase, S., Kitabatake, M., Yamauchi, T., & Mathé, A. A.
(1995). Seasonal mood variation among Japanese residents of Stock-
holm. Acta Psychiatrica Scandinavica, 92, 51-55.
doi:10.1111/j.1600-0447.1995.tb09542.x
Ozaki, N., Ono, Y., Ito, A., & Rosenthal, N. E. (1995). Prevalence of
seasonal difficulties in mood and behavior among Japanese civil ser-
vants. Journal of the American Psychiatric Association, 152, 1225-
1227
Rosen, L. N., Targum, S. D., Terman, M., Bryant, M. J., Hofman, H.,
Kasper, S. F., Hamovit, J. R., Decherty, J. P., Welch, B, & Rosenthal,
N. E. (1990). Prevalence of seasonal affective disorder at four lati-
tudes. Psychiatry Research, 31, 131-144.
doi:10.1016/0165-1781(90)90116-M
Rosenthal, N. E., Sack, D. A., Gillin, J. C., Lewy, A. J., Goodwin, F. K.,
Davenport, Y., Mueller, P. S., Newsome, D. A., & Wehr, T. A. (1984).
Seasonal affective disorder. Acta Psychiatrica Scandinavica, 41, 72-
80. doi:10.1001/archpsyc.1984.01790120076010
Rosenthal, N. E., & Wehr, T. A. (1987). Seasonal affective disorders.
Psychiatric Annals, 17, 670-674.
Sakamoto, K., Kamo, T., Nakadaira, S., Tamura, A., & Takahashi, K.
(1993). A nationwide survey of seasonal affective disorder at 53 out-
patient university clinics in Japan. Acta Psychiatrica Scandinavica,
87, 258-265. doi:10.1111/j.1600-0447.1993.tb03368.x
Shima, S., Shikano, T, Kitamura, T., & Asai, M. (1985). Seishinigaku.
New Self-Rating Scales for Depre ss i o n, 27, 717-723.
Shirakawa, S., Okawa, M., Uchiyama, M., Oguri, M., Kosaka, M., Mi-
shima, K., Inoue, H., & Kamei, K. (1993). Seasonality in the mood
and behaviors of a normal population sample in Japan. Journal of
Mental Health, 39, 81-93.
Suhail, K. & Cochrane, R. (1997). Seasonal changes in affective state
in samples of Asian and white women. Social Psychiatry and Psy-
chiatric Epidemiology, 32, 149-157
Suhail, K. & Cochrane, R. (1998). Seasonakl variation in hospital ad-
missions for affective disorders by gender and ethnicity. Social Psy-
chiatry and Psychiatric Epidemiology, 33, 211-217.
doi:10.1007/s001270050045
Copyright © 2012 SciRes. 855