Association between autozygosity and major depression: stratification due to religious assortment

Abstract

The effects of inbreeding on the health of offspring can be studied by measuring genome-wide autozygosity as the proportion of the genome in runs of homozygosity (F roh) and relate F roh to outcomes such as psychiatric phenotypes. To successfully conduct these studies, the main patterns of variation for genome-wide autozygosity between and within populations should be well understood and accounted for. Within population variation was investigated in the Dutch population by comparing autozygosity between religious and non-religious groups. The Netherlands have a history of societal segregation and assortment based on religious affiliation, which may have increased parental relatedness within religious groups. Religion has been associated with several psychiatric phenotypes, such as major depressive disorder (MDD). We investigated whether there is an association between autozygosity and MDD, and the extent to which this association can be explained by religious affiliation. All F roh analyses included adjustment for ancestry-informative principal components (PCs) and geographic factors. Religious affiliation was significantly associated with autozygosity, showing that F roh has the ability to capture within population differences that are not captured by ancestry-informative PCs or geographic factors. The non-religious group had significantly lower F roh values and significantly more MDD cases, leading to a nominally significant negative association between autozygosity and depression. After accounting for religious affiliation, MDD was not associated with F roh, indicating that the relation between MDD and inbreeding was due to stratification. This study shows how past religious assortment and recent secularization can have genetic consequences in a relatively small country. This warrants accounting for the historical social context and its effects on genetic variation in association studies on psychiatric and other related traits.

Figure 1

Map of the Netherlands, its major rivers, and its 26 largest municipalities (population size >100k as of April 2012) (CBS 2012).

Figure 2. Geographic distribution of religious groups in the Netherlands

A shows the geographical distribution of Catholic, Protestant and not religious groups in the Netherlands in 1849 (Dimitri 2007; Knippenberg 1992). B and C show the distribution of the current genotyped dataset (including related individuals: N=6,464) and the full dataset with a measurement for religious affiliation respectively (N=25,450). Each postal code was given the color of its most prevalent religious group. D shows the geographic distribution of the North-South PC, where the mean PC value per postal code (current living address) was computed, divided into 10 percentiles, and plotted.

Figure 3. Density plot of the F_roh distribution for religious and non-religious individuals

The bars have been scaled to have a total density of 1 for both groups in order to allow a comparison between the groups.