Using Mendelian randomization to investigate causal relationships in evolutionary theories of development and behaviour

Abstract

There are multiple instances where standard methods in evolutionary approaches to studying human behaviour cannot easily test causality and/or examine the effects of psychological mechanisms on reproductive success. For example, research into life history theory, where manipulation of exposures is not typically possible, has previously been limited to standard analytical approaches that remain vulnerable to potential confounding bias. Additionally, although there has been a longstanding cliff-edge hypothesis for the maintenance of schizophrenia, investigation has been limited due to the constraints of family studies and the inability to manipulate the exposure or test long terms outcomes such as fitness.

Mendelian randomization combines genetic and phenotypic information to investigate psychological and key evolutionary traits with fitness outcomes using a causal framework that does not rely on manipulating the exposure. I applied Mendelian randomization and other related methods to these two areas of evolutionary human behaviour research. According to life history this theory, earlier age at menarche and age at first sexual intercourse can be viewed as directing effort towards reproductive goals as part of a fast life history strategy and therefore show causal effects on reproductive and behavioural outcomes. The schizophrenia paradox refers to the evolutionary conundrum for how schizophrenia, a heritable disorder, is maintained in the population despite being associated with lower reproductive success for those diagnosed.

I find some evidence that earlier age at menarche is causally related to traits that characterize a fast life history strategy, such as earlier age at first and last birth and lower educational attainment. Additionally, it appears that increased genetic liability for schizophrenia does not confer a fitness advantage and therefore the disorder is likely maintained through other explanations than cliff-edge effects. This thesis is novel in its application of epidemiological methods to test evolutionary theories of human behaviour and demonstrates the potential for evolutionary epidemiology.

Date of Award	28 Nov 2019
Original language	English
Awarding Institution	University of Bristol
Supervisor	Ian S Penton-Voak (Supervisor), Marcus R Munafo (Supervisor) & Abigail Fraser (Supervisor)