H Karlsson, H Dal, RM Gardner, EF Torrey and C Dalman,
Journal of psychiatric research, May 27 2019
The objective of the present study was to examine if the monthly variation in births of individuals diagnosed with schizophrenia currently differs from that of unaffected individuals in Sweden. In an extensive linkage of Swedish national and regional population registers we here investigate the birth pattern of the population born 1940-97 (5,995,499 individuals) which included 30,684 individuals diagnosed with schizophrenia in the National Patient Register by December 31, 2016. Among 2,409,862 individuals born since 1973 we investigated potential confounding by co-variates associated with pregnancy and birth. We also compared the monthly birth pattern of 22,570 affected individuals to that of their 41,528 unaffected full siblings. We observe a significant birth excess of individuals with schizophrenia in December, HR 1.07 95%CI (1.01-1.13). Patients born in December received a registered diagnosis of schizophrenia at a slightly younger age than those born during other months. A number of co-variates were associated not only with schizophrenia but also varied across birth months. Inclusion of these in the models however had virtually no influence on the risk for schizophrenia associated with December birth. In comparisons between full siblings, the association between December birth and later diagnosis of schizophrenia remained, albeit slightly attenuated, HR 1.06 (0.99-1.12). Risk for schizophrenia associated with birth in December in Sweden during the study period does not appear to be fully explained by our investigated co-variates or factors shared between family members and may thus represent monthly/seasonal variation in environmental factors involved in the etiology of schizophrenia.