The author examines how temperature and length of day can affect mood and behavior, both in a general population and a group of inpatients. In both groups, there were two peaks of violent behavior, one in May-June and one in October-November, which correspond with the equinoxes. Is it possible to track violent behavior in various geographical areas depending upon weather and length of day?
In the last century, there has been an increasing interest in the influence of climate on human behavior. Information has mainly been gathered by comparing seasonal variations of behavior with climatic changes. Although pollution, changes in air pressure, rapid changes in temperature, solar activity and humidity all have been discussed, most of the attention has been on the influence of changes in temperature and length of day on human behavior.
Summer and Winter
Since the early 1980s, differences in mood and behavior between summer and winter have been of great interest. Patients with repeated affective episodes that always occur at the same time of the year have been diagnosed with Seasonal Affective Disorder (SAD). The most frequent form of SAD is winter depression, which is associated with increased weight and sleep. The frequency of winter depression increases with increasing latitude, and young women are the most affected. Most authors link winter depression to changes in length of day. Recurrent depressions and manias in summer have also been described; high temperature may be the trigger for these affective episodes.
Spring and Fall
Until Rosenthal et al. (1984) described SAD, the most studied seasonal pattern of behavior was the increased frequency of suicides (Hakko et al., 1998) and hospital admissions for affective disorders (Eastwood and Peacocke, 1976) in spring and, to a lesser extent, in fall.
The start of an affective episode in bipolar patients (Faedda et al., 1993) and use of electroconvulsive therapy (Eastwood and Peacocke, 1976) are also more frequent in spring and fall. Researchers speculate that disturbances in the sleep-wake rhythm may cause these seasonal peaks. The rapidly changing length of day around the equinoxes might disturb the circadian and sleep-wake rhythms and change mood and activity in vulnerable individuals.
Violence and Seasons
Studies of assaults and rape have detailed an increased frequency of recorded violence in summer, linked to high temperature (Anderson et al., 1997; Michael and Zumpe, 1983). In studies of variations of homicide, the results are more conflicting. Most studies find no seasonal variation at all, while a report from Finland described a summer peak in frequency of homicides (Tiihonen et al., 1997).
Studies of seasonal variations of violence among psychiatric patients are based on recordings of inpatients, and some connect these variations to patients with affective disorders (Roitman et al., 1990). D'Mello et al. (1995) described an increase of violence among patients with mania in spring.
Seasonal Variation in Norway
To investigate the influence of changes in length of day on behavior, it is an advantage to study variations in violence at extreme latitudes that do not have very high temperatures during summer.
Norway is situated between 58 degrees north latitude and 72 degrees north latitude, with extreme seasonal variations of light and rather low temperatures during summer. Due to the Gulf Stream, the climate in Norway is milder than at comparable latitudes in North America.
In our two studies, we examined possible seasonal variations of violence in the general population and in acutely admitted psychiatric patients (Morken and Linaker, 2000a, 2000b). Correlations between the monthly numbers of violent episodes in the two groups were also examined and relations to changes in length of day were investigated (Table).
All violent episodes reported to the Norwegian police from 1991 to 1997 (n=82,537) and all violent patient-staff incidents at the psychiatric acute department at our hospital in Trondheim (n=512) were recorded and analyzed.
The monthly numbers of violent episodes were compared to an even distribution of violent episodes, taking into account different number of days in a month and leap years. Both in the general population (
2=343.08, df=11, p<0.0001) and among acutely admitted psychiatric patients (
2=31.40, df=11, p<0.001), the monthly figures of violence differed from what we expected. The monthly number of violent episodes in the general population and among acutely admitted patients correlated (n=12, r=0.63, p
The monthly variation of violence among acutely admitted psychiatric patients varied throughout the year in the same manner as in the general population, but with greater amplitude. If such a correlation between violence in the general population and among psychiatric patients could be replicated in other locations, information about seasonal variation in risk of violence among patients could be extrapolated from studies of variations of violence in the general population in the same location.
Influence of Light
There are at least two possibilities for the relationship between change in length of day and behavior: 1) The length of day (with maximum impact at midsummer) is the significant factor; and 2) The change in length of day (with maximum impact at the equinoxes) is important. In both cases, there may be a few weeks delay for the light to sufficiently influence biological systems.
The seasonal variation of violence in the hospital and in the general population were both correlated with monthly length of day and changes in length of day (defined as the mean length of day in one month subtracted from the mean length of day in the preceding month). Due to the possibility of biological systems reacting slowly, we also introduced delays.
In areas dominated by hot summers and with little changes in length of day during the year, high temperature seems to increase the risk of violence. In northern areas, an increased frequency of violence in May and June and October through December seems to be more prevalent (Morken and Linaker, 2000a).
To my knowledge, only psychiatric studies of seasonal variation of violence among inpatients have been published. Comparing seasonal variation of violence in different diagnostic groups of patients shows that patients with affective disorders are at greater risk than other patients. D'Mello et al. (1995) showed that patients with mania admitted to the hospital in spring had a high risk for violence.
It is reasonable to assume that the seasonal variation of violence among acutely admitted inpatients reflects the seasonal variation with outpatients. Patients with a history of recurrent violent episodes and an affective disorder should be examined for seasonal variation of violence.
Clearly, geographic location may influence affective behavior. In southern parts of the northern hemisphere, higher temperatures are associated with greater risk of violent behavior. In northern climates, the time after the equinoxes, especially in spring, are high-risk periods. In all studies, winter is a low-risk period. Among psychiatric patients, those with affective disorders are at greater risk for seasonal variations in violence than other patients.
Anderson CA, Bushman BJ, Groom RW (1997), Hot years and serious and deadly assault: empirical tests of the heat hypothesis. J Pers Soc Psychol 73(6):1213-1223.
D'Mello DA, McNeil JA, Msibi B (1995), Seasons and bipolar disorder. Ann Clin Psychiatry 7(1):11-18.
Eastwood MR, Peacocke J (1976), Seasonal patterns of suicide, depression and electroconvulsive therapy. Br J Psychiatry 129:472-475.
Faedda GL, Tondo L, Teicher MH et al. (1993), Seasonal mood disorders. Patterns of seasonal recurrence in mania and depression. Arch Gen Psychiatry 50(1):17-23.
Hakko H, Rasanen P, Tiihonen J (1998), Seasonal variation in suicide occurrence in Finland. Acta Psychiatr Scand 98(2):92-97 [see comment].
Michael RP, Zumpe D (1983), Sexual violence in the United States and the role of season. Am J Psychiatry 140(7):883-886.
Morken G, Linaker OM (2000a), Seasonal variation of violence in psychiatric patients and the general population. European Journal of Psychiatry 14(3):153-159.
Morken G, Linaker OM (2000b), Seasonal variation of violence in Norway. Am J Psychiatry 157(10):1674-1678.
Roitman G, Orev E, Schreiber G (1990), Annual rhythms of violence in hospitalized affective patients: correlation with changes in the duration of the daily photoperiod. Acta Psychiatr Scand 82(1):73-76.
Rosenthal NE, Sack DA, Gillin JC et al. (1984), Seasonal affective disorder. A description of the syndrome and preliminary findings with light therapy. Arch Gen Psychiatry 41(1):72-80.
Tiihonen J, Rasanen P, Hakko H (1997), Seasonal variation in the occurrence of homicide in Finland. Am J Psychiatry 154(12):1711-1714.