Journal of Affective Disorders 132 (2011) 64–70

Contents lists available at ScienceDirect

Journal of Affective Disorders

j ourna l homepage: www.e lsev ie r.com/ locate / j ad

Research report

Association between affective temperaments and season of birth in a generalstudent population

Zoltan Rihmer a,⁎, Peter Erdos b, Mihaly Ormos b, Konstantinos N. Fountoulakis c,Gustavo Vazquez d, Maurizio Pompili e,f,g, Xenia Gonda a,h

a Department of Clinical and Theoretical Mental Health, Kutvolgyi Clinical Center, Semmelweis University; Kutvolgyi ut 4., 1125 Budapest, Hungaryb Department of Finance, Budapest University of Technology and Economics; Muegyetem rakpart 9, 1111 Budapest, Hungaryc 3rd Department of Psychiatry, Aristotle University, University Hospital AHEPA, 1 Kyriakidi Str, 54636 Thessaloniki, Greeced Department of Neuroscience, University of Palermo; Mario Bravo 1259, C.P. 1425. Buenos Aires, Argentinae Department of Neuroscience, Mental Health and Sensory Function - Suicide Prevention Center, Sant'Andrea Hospital, Sapienza University or Rome,Via di Grottarossa, 1035, 00189 Rome, Italyf McLean Hospital, Harvard Medical School, 115 Mill Street , Belmont, MA 02478, USAg Villa Rosa Research Center, Italyh Department of Pharmacology and Pharmacotherapy, Semmelweis University; Nagyvarad ter 4., 1084 Budapest, Hungary

a r t i c l e i n f o

⁎ Corresponding author at: Semmelweis UniversityKutvolgyi Clinical Center, 1125Budapest Kútvölgyi út 4,1 355 8498.

E-mail address: rihmerz@kut.sote.hu (Z. Rihmer).

0165-0327/$ – see front matter © 2011 Elsevier B.V.doi:10.1016/j.jad.2011.01.015

a b s t r a c t

Article history:Received 14 December 2010Received in revised form 25 January 2011Accepted 25 January 2011Available online 21 February 2011

Background: Several studies indicate a significant association between birth season andpersonality and neuropsychiatric disorders. The aim of our present study was to investigate theassociation between affective temperaments and season of birth in a nonclinical sample.Methods: 366 university students completed the standardized Hungarian version of theTemperament Evaluation of Memphis, Pisa, Paris and San Diego-Auto-questionnaire (TEMPS-A).Ordinary Least Squares regression was applied to explain the relationship between TEMPS-Asubscale and birth season of the respondents.Results: We found a significant association between temperament scores and birth season inthe case of the Hyperthymic, Cyclothymic, Irritable and Depressive temperaments, while nosignificant results emerged for the Anxious temperament.Limitations: The relatively small sample size, especially in the case of seasonal and monthlysubsamples limits generalization of our results.Conclusions: Our results support the evidence that there is a strong association between seasonof birth and personality, extending the results to affective temperaments as well. Furthermore,our results are in line with clinical observations concerning the seasonal variation of onset andhospitalization due to affective episodes. This is especially important, since affectivetemperaments are conceived as the subaffective and subclinical manifestations of major andminor affective disorders indicating a risk for the development of these disorders and alsoexerting a possible pathoplastic effect, thus our results also have clinical significance.

© 2011 Elsevier B.V. All rights reserved.

Keywords:Affective temperamentsBirth seasonPersonalitySeasonalityTEMPS-A

1. Introduction

Periodicity, cyclicity, and annual rhythms (including sea-sonality) that are inherent characteristics of living organismshave been reported for diverse physiological and pathological

, Faculty of MedicineHungary. Tel./fax:+36

All rights reserved.

,

conditions and phenomena. Season of birth has been consid-ered to be an important element in determining personalityand illness already in the pre-scientific medical era and thistradition lives on in several contemporary concepts rangingfrom astrological beliefs to science (Fountoulakis et al., 2007).Several studies indicate a significant effect of season of birth onpersonality traits and characteristics (Gupta, 1992; Chotai andWiseman, 2005; Tochigi et al., 2004), handedness (Martin andJones, 1999), suicide (Salib and Cortina-Borja, 2006; Chotai

65Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

et al., 1999; Dome et al., 2010), intelligence (Bibby et al., 1996),neuropsychiatric disorders (Torrey et al., 2000) such asschizophrenia (Davies et al., 2003; Torrey et al., 1997;Castrogiovanni et al., 1998; Mino and Oshima, 2006), majordepression (Mino et al., 2000), bipolar disorder (Torrey et al.,1997;Mino et al., 2000), epilepsy (Procopio andMarriott, 1998;Procopio et al., 2006) and brain tumors (Brenner et al., 2004).Season of birth is of peculiar interest when studying the effect ofenvironment in the development of several psychological andpsychiatric phenomena, since it can be considered an unspecificenvironmental factor which is associated with a complexpattern of varying environmental influences especially in highlatitudes, such as temperature, photoperiod, and consequential-ly behavioral rhythms, nutrition, infections, stress and lifestyle(Chotai et al., 2009; Chotai and Wiseman, 2005). Therefore thepossible mediating factors of birth season are complex andlargely still unknown. Several specific phenomena related toseasons have been suggested as possible underlying factors,which may influence neurodevelopment, and consequentiallythe development of personality and psychiatric disorders aswell. These factors include photoperiod, ambient temperature,weather, nutritional deficiencies, infections, toxins, maternalhormones, as well as pregnancy and birth complications(Kamata et al., 2009; Torrey et al., 1997; Gortmaker et al.,1997;Natale et al., 2007;Natale andAdan, 1999). Seasonof birthcan be considered a complex indicator of environmentalinfluences during season of conception, gestation, delivery andearly postnatal life (Chotai et al., 2002). Because of this complexconstellationof phenomenavaryingwith season, even in spite ofsome strong associations reported between season of birth andseveral factors, we still do not understand the specificmechanisms underlying this association.

Season of birth was also associated with central mono-amine and monoamine metabolite levels in several studies(Chotai and Adolfsson, 2002; Chotai and Asberg, 1999; Chotaiet al., 2006) and there may also be an association betweenseason of birth and dynamics of the turnover of monoaminesand season of birth and interactions betweenmonoaminergicsystems (Brenner et al., 2004). Seasonally varying mono-amine levels and the association between monoamineturnover and birth season are suspected in part to play arole in the background of association of certain personalitytraits with season of birth, since several well-known andvalidated personality traits have been associated withmonoamine functions (Cloninger et al., 1993). In earlierstudies Neuroticism, Impulsivity and Venturesomeness wereassociated with season of birth (Gupta, 1992) and recentlyseveral studies consistently point to an association betweenseason of birth and novelty seeking and reward dependence(Chotai et al., 2002, 2009). However, other personality traitsand temperaments, such as Akiskal's affective temperamentshave not so far been studied with respect to their associationwith birth season. Akiskal developed his framework ofaffective temperaments building on ancient traditions con-ceptualizing personality based on the association of bodyhumors and emotional reactivity (Akiskal and Akiskal, 2007;Pompili et al., 2008a), and extrapolated these theoreticalconsiderations and observations of patients and their healthyrelatives to describe five affective temperaments (Depressive,Cyclothymic, Anxious, Hyperthymic and Irritable), and alsodeveloped a self-report questionnaire to measure affective

temperaments in the general population (Akiskal et al.,2005). In this conceptual model, affective temperaments areconsidered the subaffective, subclinical manifestation ofaffective disorders, also conferring an increased risk towardsthe manifestation of minor and major mood disorders(Akiskal and Akiskal, 2007; Akiskal et al., 2005) and suicidalbehavior (Pompili et al., 2008b; Rihmer et al., 2009; Vazquezet al., 2010).

The aim of our present study was to investigate thepossible association between affective temperaments andseason of birth in a nonclinical sample.

2. Methods

2.1. Study subjects

366 university students (mean age 20.59 years, SD:1.96 years, range: 18–35 years) of Budapest University ofTechnology and Economics (BUTE) completed the standard-ized Hungarian version of the Temperament Evaluation ofMemphis, Pisa, Paris and San Diego-Auto-questionnaire(TEMPS-A) (Rozsa et al., 2008). This scale is different frommost other temperament scales in that it taps subaffectivetrait expressions as they were conceptualized in Greekpsychological medicine and, in more recent times, Germanpsychiatry. While the first version of TEMPS-A contained 84items (Akiskal and Akiskal, 2005), later, clinical and theoret-ical considerations led to the addition of 26 new itemsdescribing the Anxious temperament, resulting in the current110-item-long version of the TEMPS-A (Akiskal and Akiskal,2005; Akiskal et al., 2005).

Among the participants there were 258 males (mean age:20.87 years, SD: 2.08 years, range: 19–35 years) and 108females (mean age: 19.92 years, SD: 1.41 years, range: 18–25 years). All respondents attended the “Principles of Invest-ment Theory” course at the university. Participation in thestudy was voluntary, and a 5% extra credit was offered as acompensation.

The investigation was carried out in accordance with thelatest version of the Declaration of Helsinki. The study hasbeen approved by the Scientific and Research Ethics Commit-tee of Scientific Health Council in charge of experimentationwith human subjects. All subjects were given thoroughexplanation of the procedures of the study and all participantsgave informed consent before participating in the study.

2.2. Statistics

OLS (Ordinary Least Squares) regression was applied toexplain the relationship between TEMPS-A subscale scores andage, sex andbirth season of the respondents.Wedidnot controlfor the distribution of variables. Consistent standard errorswere estimated using White (1980) consistent covariancematrix estimation. We compiled 11 dummy variables for themonth of birth from February to December and three other forthe season of birth from spring to autumn; that is, we usedJanuary and winter as a point of reference. Seasons weredefined according to the astronomical definition as follows:winter is from December 22 to March 20, spring is fromMarch21 to June 20, summer is from June 21 to September 22 andautumn is from September 23 to December 21.

Table 1Descriptive statistics: Mean TEMPS-A scores (SD) by subscales and gender.

TEMPS-A subscales N items Total Males Females t p

Mean St. dev Mean St. dev Mean St. dev

Anxious 26 0.1961 0.1636 0.1701 0.1552 0.2582 0.1670 4.8420 b0.0001Cyclothymic 21 0.3499 0.1712 0.3328 0.1648 0.3907 0.1800 2.9809 0.0031Depressive 21 0.2814 0.1280 0.2669 0.1261 0.3161 0.1265 3.4050 0.0007Hyperthymic 21 0.5863 0.1782 0.6028 0.1757 0.5467 0.1787 2.7700 0.0059Irritable 20 or 21 a 0.3137 0.1671 0.3248 0.1705 0.2870 0.1562 1.9802 0.0484

a Males answer 20 items on the Irritable subscale and females answer 21 items.

66 Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

3. Results

Descriptive statistics of the subscale scores by gender areshown in Table 1. Scores are shown as the total subscale scoredivided by the number of items answered yielding a scorebetween 0 and 1.

We found a significant gender difference in the case of allfive TEMPS-A subscales. Women scored significantly higherin case of the Anxious, Cyclothymic and Depressive subscales(t=4.8420, p=0.0000; t=2.9809, p=0.0031; t=3.4050,p=0.0007, respectively), while men had significantly higherscore on the Hyperthymic and Irritable subscales (t=2.7700,p=0.0059; t=1.9802, p=0.484, respectively) (Table 1). Wecomputed the mean score of each subscales in each monthand for all four seasons separately (Table 2). Birthmonths andbirth seasons show a uniform distribution in the total sampleand also in the gender subsamples, however, the standarddeviations of the gender subsamples are high.

We regressed each temperament subscale on birth monthdummies, the gender dummy and age applying OLS (Table 3).The results indicated the same significant association betweengender and temperament subscale scores as the t-tests above.Furthermore, the score on the Cyclothymic subscale signifi-cantly declines with age (β=−0.0112, p=0.0114), while inthe case of the Depressive temperament it significantlyincreases with age (β=0.0068, p=0.0338) (Table 3).

OLS estimates show that there is a significant relationshipbetween the five temperament scores and birth monthcontrolling for gender and age (Table 3).

Table 2Number of respondents by gender and the means of TEMPS-A subscale scores by s

Birth month/season Male (%) Female (%) Total (%) A

January 25 (9.69%) 7 (6.48%) 32 (8.74%) 0February 11 (4.26%) 11 (10.19%) 22 (6.01%) 0March 20 (7.75%) 7 (6.48%) 27 (7.38%) 0April 14 (5.43%) 9 (8.33%) 23 (6.28%) 0May 25 (9.69%) 11 (10.19%) 36 (9.84%) 0June 27 (10.47%) 9 (8.33%) 36 (9.84%) 0July 18 (6.98%) 15 (13.89%) 33 (9.02%) 0August 30 (11.63%) 4 (3.70%) 34 (9.29%) 0September 24 (9.30%) 7 (6.48%) 31 (8.47%) 0October 23 (8.92%) 10 (9.26%) 33 (9.02%) 0November 17 (6.59%) 13 (12.04) 30 (8.20%) 0December 24 (9.30%) 5 (4.63%) 29 (7.92%) 0Spring 59 (22.87%) 31 (28.70%) 90 (24.59%) 0Summer 74 (28.86%) 26(24.07%) 100 (27.32%) 0Autumn 60 (23.26%) 28 (25.93%) 88 (24.04%) 0Winter 64 (24.81%) 24 (22.22%) 88 (24.04%) 0

In the case of the Hyperthymic temperament, regressionresults indicate a significant association between scores on thistemperament scale and birth month: individuals born duringthe period between April–June and October–November scoresignificantly higher compared to those born in January (April:β=0.1029, p=0.0387; May: β=0.1307, p=0.0014; June:β=0.1124, p=0.0036; October: β=0.1153, p=0.0057; No-vember: β=0.0954, p=0.0327). The September dummy hasthe lowest estimated parameter, indicating the lowest score onthe Hyperthymic temperament scale for those born in(β=0.0474, p=0.2352); although this result was not signif-icant (Table 3).

In the case of the Cyclothymic temperament, only thecoefficient for July was significant, indicating that the scoreson the Cyclothymic temperament scale are highest in thoseborn in July (β=0.1066, p=0.0123). The Cyclothymictemperament showed the lowest score among those whowere born in November (β=−0.0248, p=0.5715); however,this result is not significant (see Table 3).

In the case of the Irritable temperament, OLS regressionresults showed that this temperament has significantlyhigher scores in the period between April and Decemberthan in January: the probability of occurrence is approxi-mately constant between April and June and soars in July,then in the remaining part of the year it has no definitedirection (April: β=0.1047, p=0.0122, 0.1034, p=0.0364;May: β=0.0882, p=0.222; June: β=0.0905, p=0.0176;July: β=0.1586, p=0.0001; August: β=0.1047, p=0.0122,September: β=0.1262, p=0.0026, October: β=0.1079,

eason of birth.

nxious Cyclothymic Depressive Hyperthymic Irritable

.1995 0.3155 0.2887 0.5089 0.2196

.2080 0.3377 0.2792 0.5823 0.2423

.1980 0.3739 0.3157 0.5697 0.2907

.2358 0.3644 0.3188 0.6046 0.3200

.1699 0.3214 0.2394 0.6349 0.3048

.2179 0.3466 0.2712 0.6217 0.3118

.2937 0.4372 0.3434 0.5455 0.3693

.1719 0.3403 0.2773 0.5910 0.3309

.1526 0.3917 0.2873 0.5576 0.3478

.1807 0.3449 0.2367 0.6190 0.3238

.1641 0.3095 0.2730 0.5937 0.3452

.1698 0.3169 0.2660 0.5977 0.3371

.2162 0.3455 0.2804 0.6148 0.3134

.2035 0.3805 0.2976 0.5710 0.3484

.1726 0.3425 0.2500 0.6136 0.3364

.1906 0.3268 0.2955 0.5471 0.2518

Table 3Ordinary Least Squares regression of TEMPS-A scales on gender, age and month of birth.

Variable Anxious Cyclothymic Depressive Hyperthymic Irritable

Coeff. p Coeff. p Coeff. p Coeff. p Coeff. p

C 0.1708 0.0616 0.5848 0.0000 0.1873 0.0084 0.5325 0.0000 0.2823 0.0007Male −0.0867 0.0000 −0.0458 0.0289 −0.0533 0.0003 0.0628 0.0026 0.0459 0.0147Age 0.0046 0.2520 −0.0112 0.0114 0.0068 0.0338 −0.0035 0.4816 −0.0047 0.2063February −0.0152 0.7317 0.0078 0.8415 −0.0236 0.4650 0.0905 0.0609 0.0350 0.3650March −0.0061 0.8780 0.0593 0.1644 0.0232 0.5229 0.0641 0.1815 0.0740 0.0780April 0.0261 0.5562 0.0294 0.5396 0.0280 0.4573 0.1029 0.0387 0.1034 0.0364May −0.0362 0.3644 −0.0004 0.9920 −0.0525 0.0687 0.1307 0.0014 0.0882 0.0222June 0.0187 0.6829 0.0223 0.5683 −0.0147 0.6459 0.1124 0.0036 0.0905 0.0176July 0.0756 0.1329 0.1066 0.0123 0.0448 0.1945 0.0500 0.2452 0.1586 0.0001August −0.0169 0.6911 0.0250 0.5216 −0.0032 0.9310 0.0744 0.0974 0.1047 0.0122September −0.0453 0.2425 0.0705 0.0738 0.0015 0.9623 0.0474 0.2352 0.1262 0.0026October −0.0260 0.5337 0.0252 0.4837 −0.0563 0.0618 0.1153 0.0057 0.1079 0.0014November −0.0503 0.2312 −0.0248 0.5715 −0.0216 0.5128 0.0954 0.0327 0.1316 0.0024December −0.0233 0.5807 −0.0023 0.9546 −0.0166 0.6277 0.0840 0.0756 0.1129 0.0053Adj R² 0.0743 0.0462 0.0590 0.0283 0.0406

67Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

p=0.0014, November: β=0.1316, p=0.0024, December:β=0.1129, p=0.0053) (Table 3).

For the Depressive temperament, OLS results indicatedthat Depressive temperament score is lowest in those born inMay (β=−0.0525, p=0.0687) and October (β=−0.0563,p=0.0618), and although these results were not significant,they showed a strong tendency. Depressive temperamentscore was highest in those born in July (β=0.0448,p=0.1945), but this result was also not significant (Table 3).

In the case of the Anxious temperament as explained bymonth of birth controlled for age and gender none of thecoefficients were significant. Although not significant, theAnxious temperament showed the highest score among thoseborn in July (β=0.0756, p=0.1329), while the Anxioustemperament score was lowest in those born in November(β=−0.050, p=0.2312) (Table 3).

Investigating the association between temperament scoresand season of birth controlling for age and gender, we found nosignificant results for the Anxious temperament. In the case ofthe Cyclothymic temperament, OLS regression indicated thatthis temperament has significantly higher scores in those bornduring summer compared to winter borns (β=0.0505,p=0.0345). For the Depressive temperament, OLS regressionindicated significantly lower scores in those bornduring autumncompared to winter borns (β=−0.0447, p=0.0138).Hyperthymic temperament had significantly higher scores inspring and autumn borns compared to winter borns(β=0.0706, p=0.0068; β=0.0680, p=0.0138). Concerningthe Irritable temperament, the results indicate that this

Table 4Ordinary Least Squares regression of TEMPS-A scales on sex, age and season of birt

Variable Anxious Cyclothymic Depre

Coeff. p Coeff. p Coeff.

C 0.1570 0.0650 0.5784 b0.0001 0.21Male –0.0925 b0.0001 –0.0513 0.0127 –0.05Age 0.0048 0.2309 −0.0102 0.0184 0.00Spring 0.0222 0.3378 0.0083 0.7386 −0.01Summer 0.0170 0.4632 0.0505 0.0345 0.00Autumn −0.0194 0.3813 0.0076 0.7571 −0.04Adj R² 0.0602 0.0381 0.04

temperament has significantly lower scores in those born inwinter compared to subjects born in all other seasons (spring:β=0.0618, p=0.0112; summer: β=0.0398, p=0.0001; au-tumn: β=0.0840, p=0.0003) (Table 4, Fig. 1).

Considering the results significant at pb0.05, for thoseborn in spring the most likely temperament is Hyperthymicand Irritable, for summer borns Cyclothymic and Irritable, inautumn borns the Depressive temperament is the least likely,while Hyperthymic and Irritable are the most likely, while inwinter borns the Irritable temperament is the least likely.

4. Discussion

In our study we found a significant association betweenthe season of birth and four of the five affective tempera-ments as described by Akiskal (Akiskal and Akiskal, 2005,2007). Hyperthymic temperament was most common inspring and autumn borns. The Cyclothymic temperamentproved to be significantly more common in summer bornsthan in winter borns. The Irritable temperament was leastcommon in winter borns compared to the three otherseasons. The Depressive temperament was the least commonin autumn borns. Only in the case of the Anxious tempera-ments was there no evidence for association with birthseason or month.

Some very interesting and clinically relevant patterns canbe identified in our results. Considering the two basictemperaments (Depressive and Hyperthymic), the Depres-sive temperament is most common in summer and winter

h.

ssive Hyperthymic Irritable

p Coeff. p Coeff. p

42 0.0007 0.5544 b0.0001 0.3143 0.000168 0.0001 0.0629 0.0022 0.0419 0.021658 0.0530 −0.0025 0.5931 −0.0044 0.228153 0.4054 0.0706 0.0068 0.0618 0.011259 0.7574 0.0214 0.4209 0.0938 0.000147 0.0138 0.0680 0.0138 0.0840 0.000399 0.0373 0.0496

Hyperthymic Cyclothymic Irritable Depressive Anxious

Spring Summer Autumn Winter0,10

0,15

0,20

0,25

0,30

0,35

0,40

0,45

0,50

0,55

0,60

0,65

Fig. 1. Mean scores of each affective temperament by birth season.

68 Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

borns (and least common in autumn borns), while theHyperthymic is most common in spring and autumn borns(and least common in summer andwinter borns). This meansthat these two basic temperaments show opposite patternswith respect to birth seasonality, which is in line with clinicalobservations (Rihmer et al., 2010). Cyclothymic and Irritabletemperaments on the other hand show a similar pattern ofassociation with birth season. This is also in line with theclassical concept of these two latter affective temperaments;that is, both incorporate the Depressive and the Hyperthymiccomponents. However, in case of the Cyclothymic tempera-ment, depression and hyperthymia occur in an alternativeperiodic pattern, while in the case of the Irritable tempera-ment they are present simultaneously.

It is interesting that the Hyperthymic temperament andthe remaining four temperaments – containing more or lessof a depressive component – differ from each other in severalclinical and biological aspects. In contrast to the Hyperthymictemperament, the Depressive, Cyclothymic, Irritable andAnxious temperaments show a significant association withthe ‘s’ allele of the serotonin transporter gene (Gonda et al.,2006, 2009), with suicidal behavior (Vazquez et al., 2010;Pompili et al., 2008b; Rihmer et al., 2009) and with deepwhite matter abnormalities (Serafini et al., 2011) indicatingthat Hyperthymic temperament seems to be a separatedimension compared to the other affective temperaments.The finding that the Anxious temperament did not show anassociation with birth season is also well in line with clinicalobservations indicating that, unlike affective disorders andsuicide, the course of anxiety disorders shows no seasonalpattern, and with results from prior studies showing that –except for panic disorder – anxiety disorders show no seasonof birth effect (Castrogiovanni et al., 1998, 1999). It is alsowell known that panic disorder is often manifested incomorbidity with bipolar disorders suggesting a possiblecommon point in etiology (Logue et al., 2009; Goodwin andHoven, 2002; Szadoczky et al., 1998) which may in partexplain the effect of birth season in the case of panic disorder.

Another very important finding is that the association ofHyperthymic temperament with birth season (spring–au-tumn peak and summer–winter low) coincides with the birthseason pattern of bipolar I patients (Rihmer, 1980) who arepredominantly characterized by Hyperthymic temperament

(Rihmer et al., 2010), and also with the seasonality of hospitaladmissions due to mania (Rihmer, 1980). Birth seasonalityassociated with Depressive temperament (summer–winterpeak and spring–autumn low), on the other hand, coincideswith the birth seasons associated with bipolar II depression(Rihmer, 1980) which is characterized by Depressive andCyclothymic temperament (Rihmer et al., 2010), and hospitaladmissions of bipolar II patients due to depression follow thesame pattern (Rihmer, 1980). It is an interesting finding thatthe Anxious temperament did not show an association withseason of birth. It is, however, well in line with clinicalobservations, that anxiety disorders, unlike affective disor-ders and suicide, show no seasonal fluctuation, and this wayour result may shed further light on the etiopathology ofaffective disorders as well.

Our results are novel, since no study so far has investi-gated the effect of birth season on affective temperaments.However, in earlier studies a significant association betweenseason of birth and other personality traits has been reported.It has consistently been found in several studies that NoveltySeeking (NS) scores are significantly lower in subjects bornduring winter compared to those born during summer inadult women, although in adolescent women the pattern isthe opposite with higher NS scores for those born duringwinter (Chotai et al., 2001, 2003a, 2009), suggesting thatwomen born during winter exhibit a sharper rise during theiradolescence and then a steeper decrease during adulthood innovelty seeking compared to women born during summer,which suggest an association between season of birth andpersonality development throughout the life span (Chotaiet al., 2009). Reward Dependence was also found to bestrongly associated with season of birth (Chotai et al., 2009);however, the results are somewhat mixed. In a Finnish studymen born during spring had significantly lower RD scorescompared to men born in autumn (Chotai et al., 2009), whilein a Swedish study significant results were found only in caseof pooled genders, and RD peak was found in those born inDecember (Chotai et al., 2001), which the authors explain as aresult of reward dependence being modulated by aninteraction between dopamine and other neurotransmitters(Kim et al., 2006). Similarly, classical Big Five and Eysenckianpersonality dimensions have also been reported to beassociated with birth season. In a study using Neo-Pi-R,Agreeableness scores were significantly lower inwinter bornscompared to other subjects, with a marginally significanthigher score in individuals born in March and April (Tochigiet al., 2004). In another study using the Big Five Observer,however, only scores on the Conscientiousness scale tendedto show a significant association with birth season, withlower scores for summer born individuals compared to othersubjects, but the association was significant in males (Tonettiet al., 2009). In case of Extraversion and Neuroticism, thoseborn during the coldmonthswere reported to score higher onExtraversion and lower on Neuroticism (Forlano and Ehrlich,1941), and in another study summer borns scored higher onNeuroticism (Gupta, 1992). In a third study winter bornyoung adults were reported to score significantly higher onNeuroticism compared to autumn borns, who also scoredhigher than summer borns, while in case of Extraversion,autumn borns scored significantly higher compared tosummer borns (Chichilenko and Babarash, 2001) indicating

69Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

that the results are somewhat contradicting for Neuroticism.These studies therefore support a strong association betweenseason of birth and personality and our results yield furtherevidence for this association using a new approach topersonality and temperament. Earlier moderately strongcorrelations have been described between affective tempera-ments and TCI and NEO-Pi-R dimensions (Rozsa et al., 2008).Considering those personality traits which have previouslybeen shown to be associated with season of birth, Neuroti-cism had a moderately strong positive correlation withDepressive, Cyclothymic, Irritable and Anxious and a moder-ately strong negative correlation with the Hyperthymictemperament. Hyperthymic temperament also showed amoderately strong positive correlation with Extraversion. Astrong negative correlation with Conscientiousness wasobservable in case of the Cyclothymic temperament (Rozsaet al., 2008). Interestingly, the birth season associations ofaffective temperaments partially fit our expectations basedon these correlations in case of the Hyperthymic tempera-ment, where scores are higher in those born during springand autumn (partially corresponding to higher scores duringcold months and autumn as expected due to a positivecorrelation with Extraversion), while Cyclothymic scores arehigher in summer borns as expected, since the Cyclothymictemperament shows a strong negative correlation withConscientiousness, where scores are lower in summerborns. This gives further support to the association betweenaffective temperaments and birth season.

Seasonal factors may play a role during gestation or theperinatal period on monomamine gene expression (Chotaiet al., 2003b, 2009). Accordingly, an association betweenmonoamine metabolite levels in the cerebrospinal fluid andseason of birth has been reported in several studies. Levels ofthe dopamine metabolite homovanillic acid (HVA), and thenoradrenaline metabolite 3-Methoxy-4-hydroxyphenylgly-col (MHPG) were found to be significantly lower in thoseborn in summer compared to those born in winter, whilethe level of serotonin metabolite 5-hydroxyindoleacetic acid(5-HIAA) was lower in those born in spring compared tothose born in autumn, and in some studies in those bornduring winter (Chotai and Asberg, 1999; Chotai and Adolfsson,2002; Chotai et al., 2006). The observed fluctuation in centralmonoamine levels paralleled the results concerning theassociation of birth seasons and personality dimensions.Novelty seeking is thought to be modulated by dopaminergicfunction, with high novelty seeking associated with lowdopaminergic activity (Cloninger et al., 1993), and winterborns exhibit both significantly lower novelty seeking andhigher MHPG levels. On the other hand, results concerning theassociation between birth seasonality and suicidal behaviordescribed in several studies (Salib and Cortina-Borja, 2006;Dome et al., 2010) are in line with the seasonal fluctuation inserotonin turnover (Praschak-Rieder et al., 2008). The neuro-biochemistry of affective temperaments have not been studiedin detail so far; however, these results indicate that differentialmonoamine function corresponding to birth season may playan important role in the development of affective tempera-ments as well.

Our results therefore add to the mounting evidence that –beside the relationship between season of birth and unipolarand bipolar major mood disorders (Torrey et al., 1997; Mino

et al., 2000; Rihmer, 1980) – there is a strong associationbetween season of birth and personality, extending theresults to affective temperaments as well. This is especiallyinteresting, since affective temperaments are conceived asthe subaffective and subclinical manifestations of major andminor affective disorders indicating a risk for the develop-ment of this disease and also exerting a possible patoplasticeffect (Rihmer et al., 2010; Akiskal and Akiskal, 2005, 2007),thus our results also have a clinical significance. Furthermore,as the literature on the relationship between season of birthand season of incidence of different mood episodes in thesame sample of mood disorder patients is almost lacking(Rihmer, 1980), further studies are needed to explore thepossible connection between these phenomena in patientpopulations as well.

Our study has a number of limitations. Firstly, the samplesize limits the generalization of the present findings, andalthough the total number of participants was 366, thenumber of subjects in each seasonal subgroups was muchsmaller, ranging from 88 to 100. Furthermore, the instrumentto assess temperament, although reliable and commonly usedin various setting, was notmatchedwith at least another scaleto verify validity. Third, there may have been factorsdecreasing the validity of the scales, such as social desirability(Banister et al., 1994). It would be of interest to explore theresults of this study in other clinical and nonclinical samples.Another potential limitation is the subtype of no clinicalsample, students of “Investment” is a very specific population,in which affective temperament traits could influence theiracademic interests and/or elections (Figueira et al., 2010;Akiskal and Akiskal, 2007). We also did not report psycho-social variables of subjects and assessment of psychopathol-ogy was not performed.

Role of the funding sourcesThere were no funding sources involved in the work described in this

paper.

Conflict of interestThe authors have no conflict of interest to declare.

References

Akiskal, H.S., Akiskal, K.K., 2005. TEMPS: Temperament Evaluation ofMemphis, Pisa, Paris and San Diego. Journal of Affective Disorder 85(special issue), 1–242.

Akiskal, H.S., Akiskal, K.K., 2007. In search of Aristotle: temperament, humannature, melancholia, creativity and eminence. J. Affect. Disord. 100, 1–6.

Akiskal, H.S., Akiskal, K.K., Haykal, R.F., Manning, J.S., Connor, P.D., 2005.TEMPS-A: progress towards validation of a self-rated clinical version ofthe Temperament Evaluation of the Memphis, Pisa, Paris, and San DiegoAutoquestionnaire. J. Affect. Disord. 85, 3–16.

Banister, P., Burman, E., Parker, I., Taylor, M., Tindall, C., 1994. QualitativeMethods in Psychology: A Research Guide. Open University Press,Buckingham.

Bibby, P.A., Lamb, S.J., Leyden, G., Wood, D., 1996. Season of birth and gendereffects in children attending moderate learning difficulty schools. Br. J.Educ. Psychol. 66 (Pt 2), 159–168.

Brenner, A.V., Linet, M.S., Shapiro, W.R., Selker, R.G., Fine, H.A., Black, P.M.,Inskip, P.D., 2004. Season of birth and risk of brain tumors in adults.Neurology 63, 276–281.

Castrogiovanni, P., Iapichino, S., Pacchierotti, C., Pieraccini, F., 1998. Season ofbirth in psychiatry. A review. Neuropsychobiology 37, 175–181.

Castrogiovanni, P., Iapichino, S., Pacchierotti, C., Pieraccini, F., 1999. Season ofbirth in panic disorder. Neuropsychobiology 40, 177–182.

70 Z. Rihmer et al. / Journal of Affective Disorders 132 (2011) 64–70

Chichilenko, M.V., Babarash, N.A., 2001. Effect of season of birth onpersonality, health and emotional stress in adolescents. Hum. Physiol.27, 507–509.

Chotai, J., Adolfsson, R., 2002. Converging evidence suggests that monoamineneurotransmitter turnover in human adults is associated with theirseason of birth. Eur. Arch. Psychiatry Clin. Neurosci. 252, 130–134.

Chotai, J., Asberg, M., 1999. Variations in CSF monoamine metabolitesaccording to the season of birth. Neuropsychobiology 39, 57–62.

Chotai, J., Wiseman, R., 2005. Born lucky? The relationship between feelinglucky and month of birth. Pers. Individ. Differ. 39, 1451–1460.

Chotai, J., Renberg, E.S., Jacobssen, L., 1999. Season of birth associated withthe age and method of suicide. Arch. Suicide Res. 5.

Chotai, J., Forsgren, T., Nilsson, L.G., Adolfsson, R., 2001. Season of birthvariations in the temperament and character inventory of personality ina general population. Neuropsychobiology 44, 19–26.

Chotai, J., Jonasson, M., Hagglof, B., Adolfsson, R., 2002. The temperamentscale of novelty seeking in adolescents shows an association with seasonof birth opposite to that in adults. Psychiatry Res. 111, 45–54.

Chotai, J., Lundberg, M., Adolfsson, R., 2003a. Variations in personality traitsamong adolescents and adults according to their season of birth in thegeneral population: further evidence. Pers. Individ. Differ. 35, 897–908.

Chotai, J., Serretti, A., Lattuada, E., Lorenzi, C., Lilli, R., 2003b. Gene–environment interaction in psychiatric disorders as indicated by seasonof birth variations in tryptophan hydroxylase (TPH), serotonin trans-porter (5-HTTLPR) and dopamine receptor (DRD4) gene polymor-phisms. Psychiatry Res. 119, 99–111.

Chotai, J., Murphy, D.L., Constantino, J.N., 2006. Cerebrospinal fluidmonoamine metabolite levels in human newborn infants born in winterdiffer from those born in summer. Psychiatry Res. 145, 189–197.

Chotai, J., Joukamaa, M., Taanila, A., Lichtermann, D., Miettunen, J., 2009.Novelty seeking among adult women is lower for the winter bornscompared to the summer borns: replication in a large Finnish birthcohort. Compr. Psychiatry 50, 562–566.

Cloninger, C.R., Svrakic, D.M., Przybeck, T.R., 1993. A psychobiological modelof temperament and character. Arch. Gen. Psychiatry 50, 975–990.

Davies, G., Welham, J., Chant, D., Torrey, E.F., McGrath, J., 2003. A systematicreview and meta-analysis of Northern Hemisphere season of birthstudies in schizophrenia. Schizophr. Bull. 29, 587–593.

Dome, P., Kapitany, B., Ignits, G., Rihmer, Z., 2010. Season of birth issignificantly associated with the risk of completed suicide. Biol.Psychiatry 68, 148–155.

Figueira, M.L., Caeiro, L., Ferro, A., Cordeiro, R., Duarte, P.M., Akiskal, H.S.,Akiskal, K.K., 2010. Temperament in Portuguese university students asmeasured by TEMPS-A: implications for professional choice. J. Affect.Disord. 123, 30–35.

Forlano, G., Ehrlich, V.Z., 1941. Month and season of birth in relation tointelligencce, introversion–extroversion, and inferiority feelings. J. Educ.Psychol. 32, 1–12.

Fountoulakis, K.N., Iacovides, A., Karamouzis, M., Kaprinis, G.S., Ierodiakonou,C., 2007. Season of birth, clinical manifestations and dexamethasonesuppression test in unipolar major depression. Ann. Gen. Psychiatry 6,20.

Gonda, X., Rihmer, Z., Zsombok, T., Bagdy, G., Akiskal, K.K., Akiskal, H.S., 2006.The 5HTTLPRpolymorphismof the serotonin transporter gene is associatedwith affective temperaments asmeasured by TEMPS-A. J. Affect. Disord. 91,125–131.

Gonda, X., Fountoulakis, K.N., Rihmer, Z., Lazary, J., Laszik, A., Akiskal, K.K.,Akiskal, H.S., Bagdy, G., 2009. Towards a genetically validated newaffective temperament scale: a delineation of the temperamentphenotype of 5-HTTLPR using the TEMPS-A. J. Affect. Disord. 112, 19–29.

Goodwin, R.D., Hoven, C.W., 2002. Bipolar-panic comorbidity in the generalpopulation: prevalence and associated morbidity. J. Affect. Disord. 70,27–33.

Gortmaker, S.L., Kagan, J., Caspi, A., Silva, P.A., 1997. Daylength duringpregnancy and shyness in children: results from northern and southernhemispheres. Dev. Psychobiol. 31, 107–114.

Gupta, S., 1992. Season of birth in relation to personality and blood groups.Pers. Individ. Differ. 13, 631–633.

Kamata, M., Suzuki, A., Matsumoto, Y., Togashi, H., Otani, K., 2009. Effect ofmonth of birth on personality traits of healthy Japanese. Eur. Psychiatry24, 86–90.

Kim, S.J., Kim, Y.S., Lee, H.S., Kim, S.Y., Kim, C.H., 2006. An interaction betweenthe serotonin transporter promoter region and dopamine transporterpolymorphisms contributes to harm avoidance and reward dependencetraits in normal healthy subjects. J. Neural Transm. 113, 877–886.

Logue, M.W., Durner, M., Heiman, G.A., Hodge, S.E., Hamilton, S.P., Knowles, J.A.,Fyer, A.J., Weissman, M.M., 2009. A linkage search for joint panic disorder/bipolar genes. Am. J.Med. Genet. BNeuropsychiatr. Genet. 150B, 1139–1146.

Martin, M., Jones, G.V., 1999. Handedness and season of birth: a gender-invariant relation. Cortex 35, 123–128.

Mino, Y., Oshima, I., 2006. Seasonality of birth in patients with schizophreniain Japan. Psychiatry Clin. Neurosci. 60, 249–252.

Mino, Y., Oshima, I., Okagami, K., 2000. Seasonality of birth in patients withmood disorders in Japan. J. Affect. Disord. 59, 41–46.

Natale, V., Adan, A., 1999. Season of birth modulates morningness–eveningness preference in humans. Neurosci. Lett. 274, 139–141.

Natale, V., Adan, A., Chotai, J., 2007. Season of birth modulates moodseasonality in humans. Psychiatry Res. 153, 199–201.

Pompili, M., Girardi, P., Tatarelli, R., Iliceto, P., De Pisa, E., Tondo, L., Akiskal, K.K.,Akiskal, H.S., 2008a. TEMPS-A (Rome): psychometric validation of affectivetemperaments in clinically well subjects in mid- and south Italy. J. Affect.Disord. 107, 63–75.

Pompili, M., Rihmer, Z., Akiskal, H.S., Innamorati, M., Iliceto, P., Akiskal, K.K.,Lester, D., Narciso, V., Ferracuti, S., Tatarelli, R., De Pisa, E., Girardi, P.,2008b. Temperament and personality dimensions in suicidal andnonsuicidal psychiatric inpatients. Psychopathology 41, 313–321.

Praschak-Rieder, N., Willeit, M., Wilson, A.A., Houle, S., Meyer, J.H., 2008.Seasonal variation in human brain serotonin transporter binding. Arch.Gen. Psychiatry 65, 1072–1078.

Procopio, M., Marriott, P.K., 1998. Seasonality of birth in epilepsy: a Danishstudy. Acta Neurol. Scand. 98, 297–301.

Procopio, M., Marriott, P.K., Davies, R.J., 2006. Seasonality of birth in epilepsy:a Southern Hemisphere study. Seizure 15, 17–21.

Rihmer, Z., 1980. Season of birth and season of hospital admission in bipolardepressed female patients. Psychiatry Res. 3, 247–251.

Rihmer, A., Rozsa, S., Rihmer, Z., Gonda, X., Akiskal, K.K., Akiskal, H.S., 2009.Affective temperaments, as measured by TEMPS-A, among nonviolentsuicide attempters. J. Affect. Disord. 116, 18–22.

Rihmer, Z., Akiskal, K.K., Rihmer, A., Akiskal, H.S., 2010. Current research onaffective temperaments. Curr. Opin. Psychiatry 23, 12–18.

Rozsa, S., Rihmer, Z., Gonda,X., Szili, I., Rihmer, A., Ko,N., Nemeth,A., Pestality, P.,Bagdy,G., Alhassoon,O., Akiskal, K.K., Akiskal, H.S., 2008. A studyof affectivetemperaments in Hungary: internal consistency and concurrent validity ofthe TEMPS-A against the TCI and NEO-PI-R. J. Affect. Disord. 106, 45–53.

Salib, E., Cortina-Borja, M., 2006. Effect of month of birth on the risk ofsuicide. Br. J. Psychiatry 188, 416–422.

Serafini, G., Pompili, M., Innamorati, M., Fusar-Poli, P., Akiskal, H.S., Rihmer,Z., Lester, D., Romano, A., de Oliveira, I.R., Strusi, L., Ferracuti, S., Girardi,P., Tatarelli, R., 2011. Affective temperamental profiles are associatedwithwhitematter hyperintensity and suicidal risk in patientswithmooddisorders. J. Affect. Disord. 129 (1–3), 47–55.

Szadoczky, E., Papp, Z., Vitrai, J., Rihmer, Z., Furedi, J., 1998. The prevalence ofmajor depressive and bipolar disorders in Hungary. Results from anational epidemiologic survey. J. Affect. Disord. 50, 153–162.

Tochigi, M., Marumo, K., Hibino, H., Otowa, T., Kato, C., Marui, T., Araki, T.,Otani, T., Umekage, T., Kato, N., Sasaki, T., 2004. Season of birth effect onpersonality in a general population. Neurosci. Lett. 365, 120–123.

Tonetti, L., Fabbri, M., Natale, V., 2009. Season of birth and personality inhealthy young adults. Neurosci. Lett. 452, 185–188.

Torrey, E.F., Miller, J., Rawlings, R., Yolken, R.H., 1997. Seasonality of births inschizophrenia and bipolar disorder: a review of the literature. Schizophr.Res. 28, 1–38.

Torrey, E.F., Miller, J., Rawlings, R., Yolken, R.H., 2000. Seasonal birth patternsof neurological disorders. Neuroepidemiology 19, 177–185.

Vazquez, G.H., Gonda, X., Zaratiegui, R., Lorenzo, L.S., Akiskal, K., Akiskal, H.S.,2010. Hyperthymic temperament may protect against suicidal ideation.J. Affect. Disord. 127 (1–3), 38–42.

White, H., 1980. A Heteroskedasticity-Consistent Covariance Matrix Estima-tor and a Direct Test for Heteroskedasticity. Econometrica 48, 817–838.