Correspondence: Lorenzo Tonetti, MS, PsyD, Department of Psychology, University of Bologna, Viale Berti Pichat 5, 40127 Bologna, Italy. Email: email@example.com
A previous study has reported a significant month-of-birth effect on mood seasonality in the northern hemisphere. Higher mood seasonality was observed for university students born during spring or summer months (long photoperiod) compared to those born during autumn or winter months (short photoperiod). The aim of this study was to test the hypothesized photoperiod effect by comparing the correlation between month of birth and mood seasonality in two countries located at the opposite poles of the terrestrial globe: Italy (northern hemisphere) and New Zealand (southern hemisphere). On the basis of the photoperiod-at-birth hypothesis, we expected to find higher mood seasonality among individuals born in months with longer photoperiods in both countries.
The Seasonal Pattern Assessment Questionnaire was administered to 1514 young adults (1088 women, 426 men; 1027 Italians, 487 New Zealanders), with ages ranging from 18 to 34 years. The Global Seasonality Score, which is a measure of mood seasonality, was calculated from the Seasonal Pattern Assessment Questionnaire.
A significant overall month-of-birth effect was observed on the Global Seasonality Score, but only for men. Men born in April and August (corresponding in Italy to a long photoperiod, in New Zealand to a short photoperiod) had higher mood seasonality than those born in February, regardless of country.
A significant month-of-birth effect was found on mood seasonality, but results do not support the hypothesis based on the photoperiod effect.
SEASON-OF-BIRTH effect has been extensively explored in relation to schizophrenia. Individuals with schizophrenia tend to be born in winter–spring when compared to the general population, with studies showing a 5–8% winter–spring excess of birth of schizophrenics in both the northern[1, 2] and southern[3, 4] hemispheres. Despite these consistent findings regarding season of birth and schizophrenia, studies focusing on the association between birth season and mood disorders in general have led to an inconsistent pattern of results. Among those disorders, incoherent results have been reported concerning the birth-season effect on seasonal affective disorder (SAD). It has been found that patients suffering from SAD have a higher likelihood of having been born between April and September in the northern hemisphere. Month-of-birth effect has also been shown in a non-clinical sample, with higher mood seasonality for Italian and Spanish university students born in spring or summer months than for those born in autumn or winter months (significant effect only for the male sample). A similar season-of-birth effect has also been observed in adolescents. However, the season-of-birth effect on seasonal changes in well-being was not replicated in a large cohort of Scottish patients who were attending their general practitioners, most of them for symptoms unrelated to depression.
In explaining the season-of-birth effect on SAD, Pjrek et al. gave support to the hypothesis of parental conception habits, since they did not observe any significant differences between birth patterns of SAD patients and those of their healthy siblings. Nevertheless, the photoperiod around birth has been proposed as an alternative explanation for the birth-season effect on mood seasonality, as well as for other aspects of human life. Individuals born in spring (increasing photoperiod) and summer (long photoperiod) set their internal clock with longer days than those born in autumn (decreasing photoperiod) and winter (short photoperiod).
On the basis of seasonal variations, individuals regularly live in optimal and non-optimal daylight. Since the human internal clock more easily adjusts to phase delay than phase advance, there may be a higher adaptation cost for summer-born individuals (long photoperiod) to adjust to very short daylight (winter) than for those born in winter (short photoperiod) to adjust to very long daylight (summer). The higher mood seasonality observed for individuals born in spring and summer months could thus be explained on the basis of this higher adaptation cost. This higher adaptation cost is also observed by the higher mood seasonality detected in evening-type individuals,[12, 13] more frequently born during spring–summer months (increasing and long photoperiod, respectively).[14, 15]
From an experimental point of view, testing the role of the photoperiod-at-birth hypothesis in humans is difficult. A possible way to test this hypothesis is to compare the month-of-birth effect on mood seasonality in individuals born in the northern and southern hemispheres. In line with this hypothesis, higher mood seasonality should be observed for individuals born in spring–summer when compared to those born in autumn–winter in both hemispheres. When referring to the solar year, however, the seasons between the hemispheres are reversed. Given these seasonal differences, if the photoperiod around birth plays the most relevant modulating role on the association between birth month and mood seasonality, a ‘reversed’ pattern of results could be expected. Higher mood seasonality scores should be observed for those born in spring and summer months in the northern (from March to August) and southern (from September to February) hemispheres. The aim of the present study was to test the plausibility of the photoperiod hypothesis by comparing two countries located at the opposite poles of the terrestrial globe: Italy (northern hemisphere) and New Zealand (southern hemisphere).
The Seasonal Pattern Assessment Questionnaire (SPAQ) was used to measure mood seasonality. The SPAQ is the most widely used self-reporting questionnaire for normal and clinical seasonal mood variations.[16-18] Participants indicated on a 5-point Likert scale (0–4) seasonal variations experienced in six areas (sleep length, mood, social activity, weight, energy and appetite). The combined score leads to the so-called Global Seasonality Score (GSS), ranging between 0 and 24, with higher values indicating higher mood seasonality.
All Italian participants were tested and recruited at the University of Bologna (latitude and longitude: 44°30′ N; 11°21′ E) and all New Zealanders were recruited in Wellington City (latitude and longitude: 41°17′ S; 174°47′ E). In both countries, young adults were invited to participate in a cross-national survey about mood seasonality, completing an anonymous questionnaire. The questionnaires were administered in Italy during classes of a psychobiology course by a research assistant, with groups ranging in size from 10 to 30 individuals, and participation was voluntary and unpaid. New Zealander participants answered either an online version or paper-based version of the questionnaire, and Victoria University of Wellington students took part in the study for course credit. The response rate in the Italian sample was 85.7% and 85.2% in the New Zealand sample. Participants provided informed consent prior to their participation in the research project. The ethics committee of both universities approved the protocol and the study complied with the tenets of the Declaration of Helsinki.
In order to test a month-of-birth effect on mood seasonality (GSS), we performed a set of analyses of covariance (ancova) using sex, nationality and month of birth as independent variables, and age as a covariate. If ancova yielded significant results, Tukey's post-hoc test for unequal samples was then performed. The partial eta-squared (ηp2), which is a measure of effect size, was calculated. We also performed a set of multiple linear stepwise regression analyses to examine which of the factors considered were the main predictors of mood seasonality. The significance level was set at P < 0.05.
A total of 1663 mood seasonality questionnaires were administered, but 149 questionnaires were discarded due to missing data. The final sample consisted of 1514 participants (1088 women, 426 men; 1027 Italians and 487 New Zealanders). The mean age of the overall sample was 23.21 ± 3.62 (median = 22; mode = 20), with age ranging from 18 to 34 years. There was a significant age difference between Italian men (23.40 ± 3.55) and women (22.39 ± 2.90) (t1025 = −4.58; P < 0.005; Cohen's d = 0.31), but no age difference between men (24.60 ± 4.20) and women (24.25 ± 4.34) for the New Zealand sample. The New Zealand participants (24.36 ± 4.30; median = 23; mode = 22) were significantly older than the Italian participants (22.66 ± 3.11; median = 22; mode = 20) (t1512 = 8.74; P < 0.001; Cohen's d = 0.45).
The distribution of participants by months of birth in the total sample, as well as separately for sex and nationality is shown in Table 1. The proportion of young adults was quite similar across the months, as well as among sexes and nationality groups.
Table 1. Distribution of participants by month of birth in the total sample and separately for sex and nationality
Month of birth
A two-way ancova was performed to examine the effect of sex (male and female) and nationality (Italy and New Zealand) on the GSS, controlling for age. Sex had a significant effect (F1,1509 = 58.98; P < 0.001; ηp2 = 0.04), with women (10.16 ± 4.03) scoring higher than men (8.22 ± 4.43). A significant effect was also observed for nationality (F1,1509 = 14.53; P < 0.001; ηp2 = 0.01), with Italians (9.71 ± 3.95) having higher scores than New Zealanders (8.67 ± 4.71).
On the basis of the significant differences in GSS between Italy and New Zealand, we converted the GSS into a Z-score to facilitate the comparison of month-of-birth effect between the two countries. The Z-score was computed separately for each country. Given the observed sex differences, we carried out separate two-way ancovas for men and women to examine the effect of month of birth and nationality on the GSS Z-score, again controlling for age. For women, we found no significant effect of nationality, month of birth or interaction between the two factors (Fig. 1). For men, only month of birth was statistically significant (F11,401 = 2.13; P < 0.05; ηp2 = 0.05) (Fig. 2). Post-hoc comparisons (Tukey's test for unequal samples) indicated that participants born during April and August had higher GSS Z-scores than those born in February (P < 0.05 for both comparisons).
Finally, results from multiple linear stepwise regression analysis showed that none of the independent variables (i.e. month of birth, nationality and age) were significant predictors of the GSS Z-score for the female sub-sample. However, for men, month of birth was a significant predictor of mood seasonality score (Beta = 0.10; t422 = 2; P < 0.05).
This paper explored the effect of month of birth on mood seasonality in samples of healthy participants from a northern and a southern hemisphere country (Italy and New Zealand). The paper first examined sex and nationality differences regarding mood seasonality. With reference to sex differences, we found that women reported higher mood variations (as measured by the GSS) than men, confirming that women are usually more susceptible to seasonal sensitivity than men.[12, 19, 20] Regarding nationality differences, Italians had higher mood seasonality than New Zealanders. As Bologna and Wellington are located at similar latitudes, with comparable (even if reversed) photoperiods, it is possible to assume that in this case latitude does not play a primary modulating role on mood seasonality. An alternative explanation is related to the different climates between Italy and New Zealand, as discussed below.
With reference to the photoperiod-at-birth hypothesis, we observed different effects for female and male respondents. On the one side, there was no significant effect for women, which could be explained on the basis that women react less to environmental sunshine-related changes than men. This difference in reactivity could be due to women's genetically programmed circamensual rhythm that can make their circadian system less able to react to environmental changes. On the other side, we detected a significant effect of month of birth in men, although the expected reversed pattern of results between hemispheres was not observed. Both Italian and New Zealander men born in April (spring in Italy and autumn in New Zealand) and August (summer in Italy and winter in New Zealand) had higher GSS Z-scores than those born in February (winter in Italy and summer in New Zealand). These results confirm that men are more sensitive to the month-of-birth effect than women, which is in line with past studies on mood seasonality as well as circadian typology.[22, 23] However, these data disagree with the photoperiod hypothesis because we expected to find the opposite reversed pattern of results between the two countries.
A possible explanation of the discrepancy in month-of-birth effect on mood seasonality, compared to our expectations based on the photoperiod hypothesis, is related to the climatic differences between Italy and New Zealand. Indeed, on the basis of the Köppen–Geiger climate classification, the climate in Wellington is oceanic while in Bologna there is a wet temperate climate with hot summers. Other climatic differences between the two cities are that Wellington is very windy all year round and has a high rainfall (average annual rainfall is 1249 mm), with a peak in June and July (winter in the southern hemisphere). On the contrary, the average annual rainfall in Bologna is much lower (660 mm), with peaks in spring and autumn, and very low incidence during winter and summer. Obviously this is only a reflection, but it can be partially supported by studies showing an effect of climate (e.g. total amount of rain) on features of SAD.
A recently published review has summarized several effects of the early life photoperiod, including behavioral, somatic, and reproductive. Nevertheless, our study indicates that perhaps the photoperiod around birth does not play a primary modulating role on the relation between month of birth and mood seasonality. At this stage of knowledge, we can only assume that other environmental factors, such as weather, humidity and rain, or their combination, could be crucial. There might also be other confounders, such as family history of mental disorder and traumatic events during birth, which were not explored in the present research. Aiming to detect which factors play the main modulating role on the month-of-birth effect on mood seasonality, future studies should compare different nations defined by different environmental parameters and also explore confounder effects.
The authors report no conflicts of interest and declare that they received no financial grants to carry out the study.