Seasonal Variation in the Occurrence of Homicide in Finland
Abstract
OBJECTIVE: Although seasonal variation in impulsive aggression related to circannual rhythms of central serotonin neurotransmission is a topic of current interest, there is little firm knowledge on seasonality in the occurrence of homicide. Longitudinal studies on the seasonal rhythms of platelet imipramine binding and l-tryptophan levels have placed the circannual peaks around January and February and the nadirs around May and August. The aim of this study was to test the hypothesis that the number of homicides is the lowest during winter and the highest during spring and summer. A secondary hypothesis was that the seasonal variations in homicides and violent suicides are correlated. METHOD: The largest database on the monthly occurrence of homicide thus far (N=4,553) was used in this study, in which the monthly occurrence of all murders and manslaughters in Finland during the years 1957–1995 was analyzed. RESULTS: During winter the homicide rate was 6% below the expected rate. Correspondingly, during summer there was a 6% elevation above the expected homicide rate, but no significant peak was observed in spring. There was a significant association between the monthly occurrence of homicides and violent suicides but not between homicides and nonviolent suicides. CONCLUSIONS: The results suggest that a seasonal variation in the occurrence of homicide exists. On the basis of current literature, it could be hypothesized that this seasonal variation and the correlation between the monthly occurrence of homicides and violent suicides are associated with the observed circannual rhythms of serotonin transmission. (Am J Psychiatry 1997; 154:1711–1714)
Low concentrations of the major serotonin metabolite 5-hydroxyindoleacetic acid (5-HIAA) in the CSF are known to be related to impulsivity and aggression (1–4), as are low levels of platelet paroxetine binding (5). A substantial seasonal variation in the frequency of suicide has been found, with peaks occurring in the spring and summer and lows in the autumn and winter (6–11). Seasonal effects associated with aggressive and impulsive behavior have been studied by investigating weekly or monthly variation in the occurrence of suicides and by examining the annual variation in factors reflecting serotonin (5-HT) function in the brain (12). Some studies on platelet 5-HT uptake (13, 14) and [3H]imipramine binding (14–17) have detected the lowest levels in the spring and the highest levels in the late summer, whereas some other platelet 5-HT uptake studies (18) and imipramine binding studies (19, 20) have shown the lowest levels in the spring or summer and the highest levels in the winter. It has been reported that among normal control subjects, 5-HIAA concentrations were lower in the late winter and early spring than the values obtained in the summer and fall (21). There are no published data that would reveal the quantitative correlation between CSF 5-HIAA and platelet 5-HT uptake values, but several studies have confirmed that the 5-HT transporter protein in 5-HT neurons and the one in platelets are identical (22, 23).
Recent state-of-the-art longitudinal studies on plasma L-tryptophan levels and platelet imipramine binding indicate that tryptophan levels are highest in January and February and lowest in May (12, 24) and that the peaks of imipramine binding density are in February and the troughs are in August (25, 26). Maes et al. (12) noted a significant temporal relationship between seasonal variations in L-tryptophan availability and the occurrence of violent suicide in Belgium: there was a significant negative correlation between plasma L-tryptophan levels and the weekly total number of suicides.
There are grounds to suspect that seasonality of dysfunctional central 5-HT neurotransmission is related to both internally and externally directed violent behavior (21, 27). Studies on seasonal variation in the occurrence of suicide and 5-HT function in humans indicate that there might exist seasonal variations in the occurrence of homicide, but no significant seasonal variation has been reported thus far. On the basis of the platelet L-tryptophan and imipramine binding studies (12, 24–26) it can be assumed that if a seasonal variation in the occurrence of homicide does exist, the peaks of occurrence would be in May and August and the lows in January and February.
Although the seasonality of suicides and the seasonal variations in biochemical factors related to aggression and impulsivity are topics of current interest, very little research has been carried out to evaluate circannual rhythms in the severity of violent acts or in the incidence of homicide. To our knowledge, no seasonality has been reported for impulsive or aggressive behavior other than homicidal or suicidal behavior, except for a study by Michael and Zumpe (28) which showed that the rape rate reaches its peak during the summer. So far, one of the largest studies of seasonality in homicide has been that of Maes and co-workers (10), in which all homicides in Belgium during the years 1979–1987 (N=1,462) were studied by means of spectral analysis. No significant circannual rhythms were found in the time series of homicides, nor has any significant seasonal variation in the occurrence of homicide been reported thus far in any other studies with smaller samples (28–32), although Goodman et al. (33), with a large sample (N=2,702), observed a slightly higher frequency of homicides during July and August.
If the magnitude of seasonal variation is related to the magnitude of the variation in climatic conditions, Finland is an ideal country in which to study the issue because of its extreme northern location. In addition, for several decades homicide statistics have been registered each month, providing a very large database. The aim of this study was to investigate seasonality in homicides occurring in Finland during the years 1957–1995 and to test the hypothesis that the number of homicides would be the lowest around January and February (winter) and the highest around May (spring) and August (summer). The relationship between the monthly frequency of homicides and violent versus nonviolent suicides was also investigated.
METHOD
The monthly numbers of homicides in Finland in the period from 1957 to 1995 were obtained from the Official Statistics of Finland from the series “Criminality Known to the Police” (Statistics Centre, 1957–1995). The numbers of all murders and manslaughters known to the police were sampled and analyzed in season intervals. The winter season was defined as December, January, and February; the spring season as March, April, and May; summer as June, July, and August; and autumn as September, October, and November. The monthly numbers of suicides were available for the years 1980–1995 and were also obtained from the Official Statistics of Finland, Statistics Centre. Suicides were classified in two categories: violent (hanging, drowning, shooting, wrist slashing, jumping from a high place) and nonviolent (poisoning, gas, other methods).
The chi-square test for multinomials was used as an overall measure of deviation (34). The null hypothesis was that homicides occur in a time interval (season) with a probability proportional to the length of that time interval. If there is no difference in homicide rates between time intervals, the chi-square value is small. The expected frequency of homicides in each time interval was calculated taking into account the unequal numbers of days in time intervals and the effect of leap years during the study period.
To locate more precisely the time intervals in which there was a possible departure from the null hypothesis, we calculated for each time interval the equation
where Π is the observed proportion of homicides in the sample, Π0 is the expected proportion of homicides when the null hypothesis is true, and n is the total number of homicides in the study period. If the null hypothesis is true, the value of the ratio Π/Π0 is approximately 1.00. The null hypothesis was rejected if the 95% confidence interval of the ratio did not include the value 1.00. The relationship between homicides and suicides was studied with use of the Spearman correlation coefficient.
RESULTS
During the study period in the years 1957–1995, 4,553 homicides were recorded. The null hypothesis of no difference in homicide rates between seasons was rejected (χ2=8.29, df=3, p=0.04). To identify the seasons in which the departure of the observed frequency of homicides from the expected frequency was statistically significant, the ratio Π/Π0 with the 95% confidence interval was calculated for each season (table 1). In the winter the observed frequency of homicides was about 6% lower than the expected frequency, and in the summer about 6% higher. In the spring and autumn, the observed frequency of homicides did not differ statistically from the expected frequency. Table 2 demonstrates the deviations (%) between the observed and the expected homicide frequencies by month. The monthly mean number of homicides varied between 8.38 (SD=2.94) and 10.85 (SD=4.26).
During the years 1980–1995, the total number of suicides in Finland was 21,279, and 70.3% (N=14,965) of them were violent. When the monthly ratios of suicides and homicides (observed divided by expected) were analyzed, the homicide ratios were correlated with the violent suicide ratios (r=0.62, N=12, p=0.03) but not with the nonviolent suicide ratios (r=0.15, N=12, p=0.63).
DISCUSSION
The main finding of this study was that there is seasonal variation in the occurrence of homicides in Finland, with the peaks in summer (especially in July and August) and the troughs in winter (especially in January and February), and thus our hypothesis was accepted. The result was confirmed by analyzing the data with two statistical methods (34). To our knowledge, this is the first study to demonstrate a statistically significant seasonal variation in the occurrence of homicide.
The previous published studies with relatively small samples (10, 11, 28–31) failed to detect any seasonal variation in homicides, but the study by Goodman et al. (33) with a large sample showed a minor seasonal trend for homicides with a peak in summer, and especially in August, which is consistent with our results. The reasons for the partial discrepancy among different studies may be that true seasonal variations in aggressive behavior are so small that a very large database and a long follow-up time are needed for the phenomenon to be revealed. In addition, there are reports that latitude and climatic factors (i.e., day length, daily temperature, daylight, and humidity) may influence mood (35–37) and impulsive behavior (9). Since Finland is located in the extreme north (with the geographical coordinates between 60° and 70° N) and the climatic conditions vary greatly between seasons, the effect of season on violent offenses should be seen more clearly there than in countries nearer the equator.
A large proportion of Finnish homicide offenders have been reported to be impulsive antisocial alcoholics (38–40), and therefore it is possible to hypothesize that the frequency of homicides is lowest when serotonergic turnover has its seasonal peak and highest when serotonergic turnover is at its lowest. The low observed number of homicides in January and February is in line with longitudinal studies on plasma L-tryptophan levels (12, 24) and on platelet imipramine binding density (25, 26) which indicate that the seasonal peak in serotonergic turnover is around January and February. Our results showed a significant association between the monthly occurrence of homicides and violent suicides but not between homicides and nonviolent suicides. This is in accord with studies suggesting that both externally and internally directed violent behaviors are associated with dysfunctional 5-HT neurotransmission (21, 27).
The seasonal peak of homicide occurrence was seen in summer, which is consistent with studies on platelet imipramine binding showing the lowest values in August (25, 26). However, no prominent peak was seen in spring, as would have been anticipated on the basis of studies on seasonal variation of L-tryptophan levels (12, 24). It is well established that there is a seasonal variation in several other biochemical variables, not just in 5-HT-related markers (24, 41). Therefore, it is possible that seasonal variations in dopamine and testosterone may contribute to the seasonal variation in the occurrence of homicide, because both of these factors are associated with aggressive behavior (4, 42, 43).
 |
 |
Received March 24, 1997; revision received July 3, 1997; accepted July 31, 1997. From the Department of Forensic Psychiatry, University of Kuopio, and the Department of Psychiatry, University of Oulu, Oulu, Finland. Address reprint requests to Dr. Tiihonen, Department of Forensic Psychiatry, University of Kuopio, Niuvanniemi Hospital, FIN-70240 Kuopio, Finland; [email protected] (e-mail).
1. Brown GL, Goodwin FK, Ballenger JC, Goyer PF, Major LF: Aggression in humans correlates with cerebrospinal fluid amine metabolites. Psychiatry Res 1979; 1:131–139Crossref, Medline, Google Scholar
2. Linnoila M, Virkkunen M, Scheinin M, Nuutila A, Rimon R, Goodwin R: Low cerebrospinal fluid 5-hydroxyindoleacetic acid concentration differentiates impulsive from nonimpulsive violent behavior. Life Sci 1983; 33:2609–2614Google Scholar
3. & Aring;sberg M, Nordström P, Träskman-Bendz L: Cerebrospinal fluid studies in suicide, in Psychobiology of Suicidal Behavior. Edited by Mann JJ, Stanley M. New York, Annals of the New York Academy of Science, 1986, pp 243–255Google Scholar
4. Virkkunen M, Rawlings R, Tokola R, Poland RE, Guidotti A, Nemeroff C, Bissette G, Kalogeras K, Karonen S-L, Linnoila M: CSF biochemistries, glucose metabolism, and diurnal activity rhythms in alcoholic, violent offenders, fire setters, and healthy volunteers. Arch Gen Psychiatry 1994; 51:20–27Crossref, Medline, Google Scholar
5. Coccaro EF, Kavoussi RJ, Sheline YI, Lish JD, Csernansky JG: Impulsive aggression in personality disorder: correlates with tritiated paroxetine binding in the platelet. Arch Gen Psychiatry 1996; 53:531–536Crossref, Medline, Google Scholar
6. Näyhä S: Autumn incidence of suicides re-examined: data from Finland by sex, age and occupation. Br J Psychiatry 1982; 141:512–517Crossref, Medline, Google Scholar
7. Näyhä S: The bio-seasonal incidence of some suicides. Acta Psychiatr Scand 1983; 67:32–42Crossref, Medline, Google Scholar
8. Barraclough BM, White SJ: Monthly variation of suicidal, accidental and undetermined poisoning deaths. Br J Psychiatry 1978; 132:279–282Crossref, Google Scholar
9. Souetre E, Salvati E, Belugou JL, Douillet P, Braccini T, Darcourt G: Seasonality of suicides: environmental, sociological and biological covariations. J Affect Disord 1987; 13:215–225Crossref, Medline, Google Scholar
10. Maes M, Cosyns P, Meltzer HY, De Meyer F, Peeters D: Seasonality in violent suicide but not in nonviolent suicide or homicide. Am J Psychiatry 1993; 150:1380–1385Google Scholar
11. Maes M, Meltzer HY, Suy E, De Meyer F: Seasonality in severity of depression: relationships to suicide and homicide occurrence. Acta Psychiatr Scand 1993; 88:156–161Crossref, Medline, Google Scholar
12. Maes M, Scharpe S, Verkerk R, D'Hondt P, Peeters D, Cosyns P, Thompson P, De Meyer F, Wauters A, Neels H: Seasonal variation in plasma L-tryptophan availability in healthy volunteers: relationships to violent suicide occurrence. Arch Gen Psychiatry 1995; 52:937–946Crossref, Medline, Google Scholar
13. Arora RC, Kregel L, Meltzer HY: Seasonal variation of serotonin uptake in normal controls and depressed patients. Biol Psychiatry 1984; 19:795–804Medline, Google Scholar
14. Egrise D, Rubinstein M, Schoutens A, Cantraine F, Mendlewicz J: Seasonal variation of platelet serotonin uptake and 3H-imipramine binding in normal and depressed subjects. Biol Psychiatry 1986; 21:283–292Crossref, Medline, Google Scholar
15. Egrise D, Desmedt D, Schoutens A, Mendlewicz J: Circannual variations in the density of tritiated imipramine binding sites on blood platelets in man. Neuropsychobiology 1983; 10:101–102Crossref, Medline, Google Scholar
16. Goziotis A, Tang SW: A longitudinal study of intact platelet 3H-imipramine binding in 12 normal human subjects. Psychiatry Res 1988; 26:157–162Crossref, Medline, Google Scholar
17. Pine DS, Trautman PD, Shaffer D, Cohen L, Davies M, Stanley M, Parsons B: Seasonal rhythm of platelet [3H]imipramine binding in adolescents who attempted suicide. Am J Psychiatry 1995; 152:923–925Link, Google Scholar
18. Wirz-Justice A, Richter R: Seasonality in biochemical determinations: a source of variance and a clue to the temporal incidence of affective illness. Psychiatry Res 1979; 1:53–60Crossref, Medline, Google Scholar
19. Whitaker PM, Warsch JJ, Stancer HC, Persad E, Vint CK: Seasonal variation in platelet 3H-imipramine binding: comparable values in control and depressed populations. Psychiatry Res 1984; 11:127–131Crossref, Medline, Google Scholar
20. Kanof PD, Coccaro EF, Johns CA, Siever LJ, Davis KL: Platelet 3H-imipramine binding in psychiatric disorders. Biol Psychiatry 1987; 22:270–286Crossref, Google Scholar
21. Brewerton TD, Berrettini WH, Nurnberger JI: Analysis of seasonal fluctuations of CSF monoamine metabolites and neuropeptides in normal controls: findings with 5-HIAA and HVA. Psychiatry Res 1988; 23:257–265Crossref, Medline, Google Scholar
22. Ramamoorthy S, Bauman AL, Moore KR, Han H, Yang-Feng T, Chang AS, Ganapathy V, Blakely RD: Antidepressant- and cocaine-sensitive human serotonin transporter: molecular cloning, expression, and chromosomal localization. Proc Natl Acad Sci USA 1993; 90:2542–2546Google Scholar
23. Lesch K-P, Wolozin BL, Murphy DL, Riederer P: Primary structure of the human platelet serotonin uptake site: identity with the brain serotonin transporter. J Neurochem 1993; 60:2319–2322Google Scholar
24. Maes M, Scharpe S, D'Hondt P, Peeters D, Wauters A, Neels H, Verkerk R: Biochemical, metabolic and immune correlates of seasonal variation in violent suicide: a chronoepidemiologic study. Eur Psychiatry 1996; 11:21–33Crossref, Medline, Google Scholar
25. DeMet EM, Chicz-DeMet A, Fleischmann J: Seasonal rhythm of platelet 3H-imipramine binding in normal controls. Biol Psychiatry 1989; 26:489–495Crossref, Medline, Google Scholar
26. DeMet EM, Reist C, Bell KM, Gerner RH, Chicz-DeMet A, Warren S, Wu J: Decreased seasonal mesor of platelet 3H-imipramine binding in depression. Biol Psychiatry 1991; 29:427–440Crossref, Medline, Google Scholar
27. Linnoila M, Virkkunen M: Aggression, suicidality, and serotonin. J Clin Psychiatry 1992; 53:46–51Medline, Google Scholar
28. Michael RP, Zumpe D: Sexual violence in the United States and the role of season. Am J Psychiatry 1983; 140:883–886Link, Google Scholar
29. Lester D: Temporal variation in suicide and homicide. Am J Epidemiol 1979; 109:517–520Crossref, Medline, Google Scholar
30. Abel EL, Strasburger EL, Zeidenberg P: Seasonal, monthly, and day-of-week trends in homicide as affected by alcohol and race. Alcohol Clin Exp Res 1985; 9:281–283Crossref, Medline, Google Scholar
31. Abel EL, Welte JW: Temporal variation in violent death in Erie County, New York, 1973–1983. Am J Forensic Med Pathol 1987; 8:107–111Crossref, Medline, Google Scholar
32. Abel EL: Childhood homicide in Erie County, New York. Pediatrics 1986; 77:709–713Medline, Google Scholar
33. Goodman RA, Herndon JL, Istre GR, Jordan FB, Kelaghan J: Fatal injuries in Oklahoma: descriptive epidemiology using medical examiner data. South Med J 1989; 82:1128–1134Google Scholar
34. Wonnacott TH, Wonnacott RJ: Introductory Statistics, 5th ed. New York, John Wiley & Sons, 1990, pp 550–555Google Scholar
35. Potkin SG, Zetin M, Stamenkovic V, Kripke DF, Bunney WE: Seasonal affective disorder: prevalence varies with latitude and climate. Clin Neuropharmacol 1986; 9:181–183Medline, Google Scholar
36. Rosen LN, Targum SD, Terman M: Prevalence of seasonal affective disorder at four latitudes. Psychiatry Res 1990; 31:131–144Crossref, Medline, Google Scholar
37. Okawa M: Seasonal variation of mood and behaviour in a healthy middle-aged population in Japan. Acta Psychiatr Scand 1996; 94:211–216Crossref, Medline, Google Scholar
38. Tiihonen J, Eronen M, Hakola P: Criminality associated with mental disorder and intellectual deficiency. Arch Gen Psychiatry 1993; 50:917–918Crossref, Medline, Google Scholar
39. Tiihonen J, Hakola P: Psychiatric disorders and homicide recidivism. Am J Psychiatry 1994; 151:436–438Link, Google Scholar
40. Eronen M, Hakola P, Tiihonen J: Mental disorders and homicidal behavior in Finland. Arch Gen Psychiatry 1996; 53:497–501Crossref, Medline, Google Scholar
41. Dabbs JM: Age and seasonal variation in serum testosterone concentration among men. Chronobiol Int 1990; 7:245–249Crossref, Medline, Google Scholar
42. Lycke E, Modigh K, Roos BE: Aggression in mice associated with changes in the monoamine metabolism of the brain. Experientia 1969; 25:951–953Crossref, Medline, Google Scholar
43. Bandecchi A: An ethological model of long-term modification of dimensions of the temperament (abstract). Neuropsychopharmacology 1994; 10:725SGoogle Scholar
