Seasonal Variations in Mood and Behavior Among Chinese Medical Students

Although there have been many reports on seasonal variations in mood and behavior in groups in the United States and Europe in the past two decades (1–7), there has been only one previous study from Asia (8). To our knowledge, the present study is the first survey of seasonal changes in mood and behavior in mainland China.

METHOD

A total of 1,358 medical students from Jining Medical College in Jining, China (35.4° north latitude), were surveyed by means of a Chinese translation of the Seasonal Pattern Assessment Questionnaire (1), a screening instrument for assessing seasonality, and a Chinese version of the Beck Depression Inventory (9). Responses were evaluated for winter and summer patterns of mood and behavior change, including seasonal affective disorder and subsyndromal seasonal affective disorder, derived from the case-finding criteria of Kasper et al. (3) and Rosen et al. (6). Summer and winter were defined, respectively, as June 21 to September 20 and December 21 to March 20, as specified in the Seasonal Pattern Assessment Questionnaire. The survey was performed in a classroom setting in early March 1996. All full-time students in the first to third years who were currently enrolled at the school and attending regular classes were approached. Completion and return of the questionnaires was regarded as consent for the confidential use of the data in reports, as is standard practice in China.

Fourteen days after the initial survey, 126 randomly chosen subjects were retested with the same questionnaires. The intraclass correlation coefficient of reliability for a key Seasonal Pattern Assessment Questionnaire subscale, the global seasonality score, was 0.58.

The effect of gender on global seasonality scores was tested by an analysis of covariance, using gender as a grouping factor and age as a covariate; winter and summer patterns were compared by the methods listed in table 1.

RESULTS

Of 1,358 subjects, 814 (59.9%) were men, and 544 (40.1%) were women. The mean age was 20.7 years (SD=1.4, range=17–26). The subjects were highly homogeneous with regard to education (all college students), marital status (all single), and residency history (all living at the same latitude for at least 3 years).

The mean depression and seasonality scores are given in table 1. There was no significant difference in global seasonality score between the men and women (8.4, SD=3.6, versus 8.2, SD=3.7, respectively) (F=0.09, df=2, 1346, p=0.43). No significant correlation between age and global seasonality score emerged (r=0.01, df=1356, p>0.79).

Of the 1,358 subjects, 1,110 (81.7%) regarded behavioral changes associated with the seasons as a problem. Specifically, 810 (59.6%) of the subjects regarded the problem as mild, 235 (17.3%) as moderate, 50 (3.7%) as marked, and 15 (1.1%) as severe or disabling. Table 1 shows the numbers of subjects who reported that changes with the seasons were a moderate, marked, severe, or disabling problem. Only 13 subjects (1.0%) reported no behavioral changes across the seasons, and 248 subjects (18.3%) did not regard seasonal changes as a problem.

Of all the subjects surveyed, significantly more reported summer (30.9%) than winter (20.0%) seasonal mood and behavioral patterns (U=6.34, N=1,358, p<0.001). The estimated frequencies were 2.4% (N=33) for winter seasonal affective disorder, 5.7% (N=78) for winter subsyndromal seasonal affective disorder, 4.4% (N=60) for summer seasonal affective disorder, and 8.0% (N=108) for summer subsyndromal seasonal affective disorder. The rate of summer seasonal affective disorder or subsyndromal seasonal affective disorder was significantly higher than the rate of winter seasonal affective disorder (U=2.82, N=1,358, p<0.01) or subsyndromal seasonal affective disorder (U=2.37, N=1,358, p<0.05), a difference that persisted when the genders were considered separately. No difference was found between men and women with regard to the rates of the four types of seasonal affective disorders. We excluded from analysis the small group of subjects (6.3%, N=86) who met the criteria for both seasonal patterns. A positive correlation (r) of 0.18 (df=1345, p<0.0001) was observed between the global seasonality score and the Beck Depression Inventory total score.

DISCUSSION

In what we believe to be the first study of seasonal variation in mood and behavior in a Chinese study group, such changes were rather common—more so even than in comparable studies in the United States (3, 6). Thus, in the present study, the mean global seasonality score was much higher than that of Kasper et al. (8.3, SD=3.6, versus 5.4, SD=4, respectively). Moreover, 81.7% (N=1,110) of the subjects surveyed claimed that seasonal changes were a problem for them compared with only 27% in the study by Kasper et al. (3). However, the percentage of our subjects who claimed their seasonal problems were of a marked to severe or disabling degree was lower (4.8%) than in the U.S. study (8%). Although a high level of overall depression (Beck Depression Inventory total score=9.7) in this Chinese student group might have biased the responses in favor of reports of higher levels of seasonal changes, this Beck Depression Inventory score is comparable with the mean score of 10.3 (SD=7.3) obtained from another group of Chinese medical students from Hong Kong (10). In addition, we recognize that these findings in a group of medical students cannot be assumed to reflect the seasonal mood and behavior patterns in the population as a whole.

The greater frequency of summer versus winter mood and behavior variations we encountered contrasts with the predominance of winter difficulties seen in most studies of Caucasian populations performed to date (1, 3–7) but is in keeping with the only other published study of an Asian population. In the latter study, Ozaki et al. (8) found that more civil servants in Nagoya in northern Japan reported seasonal mood and behavioral problems during the summer than in the winter.

There is nothing particularly severe about the summers in Nagoya or mainland China compared with some of the climates in which the Caucasian studies were performed. It is possible, however, that the relative lack of air conditioning in China and Japan compared with the West might account for the greater frequency of summer difficulties in these two studies. Although summer activities might account for summer changes, we have no specific information about the vacation patterns of particular students but do know the general tendency of students to remain in their hometowns during vacation. Thus, stress would seem an unlikely explanation for summer variations. It is also possible that genetic differences in people who have evolved in different climates might explain the different susceptibility to summer and winter variations in different ethnic groups (11). Another similarity between the two Asian studies is the absence of a preponderance of women in the seasonally affected group, which has been typical of most studies conducted in the West.

Received Dec. 16, 1996; revisions received May 26, 1998, and March 4 and April 26, 1999; accepted May 17, 1999. From the Clinical Psychobiology Branch, NIMH, Bethesda; and Jining Medical College and the affiliated psychiatric hospital, Jining, China. Address reprint requests to Dr. Han, c/o Norman E. Rosenthal, M.D., Clinical Psychobiology Branch, Bldg. 10/3S-231, NIMH, 10 Center Dr., Bethesda, MD 20892-1390; [email protected] (e-mail). The authors thank the teachers and students of Jining Medical College for their participation and Drs. Chao, Jin, Su, Ren, and Han of the affiliated psychiatric hospital of Jining Medical College for their help with data collection.

TABLE 1

Enlarge table