Stress, Genes, and Generalizability Across Gender: Effects of Maoa and Stress Sensitivity on Crime and Delinquency*

Abstract

In the current study, we extend the gene-by-environment interaction (cGxE) literature by examining how a widely studied polymorphism, the MAOA upstream variable number tandem repeat (MAOA-uVNTR) interacts with distal and proximal stressors to explain variation in crime and delinquency. Prior research findings have revealed that MAOA-uVNTR interacts with single indicators of environmental adversity to explain criminal behavior in general-population and incarcerated samples. Nevertheless, the genetically moderated stress sensitivity hypothesis suggests that increased risk for criminal behavior associated with variation in the MAOA-uVNTR can be best understood in the context of both distal stress during childhood and proximal stress in adulthood. Therefore, we employed Tobit regression analyses to examine a genedistal-proximal environment (CGxExE) interaction across gender in a sample of university students (n = 267) and with data from the National Longitudinal Study of Adolescent to Adult Health (Add Health; n = 1,294). The results across both sets of analyses demonstrate that variation in the MAOA-uVNTR interacts with distal and proximal stress to lead to increased risk for criminal behavior in males. Although proximal life stress is associated with an increase in crime and delinquency, this effect is more pronounced among MAOA-L allele carriers that have experienced distal stress.

The Effect of MAOA and Stress Sensitivity on Crime and Delinquency: A Replication Study

“…The resulting correlations revealed nonsignificant correlations between MAOA-uVNTR genotype and distal stress (r = .05, p = .313), as well as MAOA-uVNTR genotype and proximal stress (r = −.06, p = .177). Similar to the findings reported by Wells et al (2017), these findings fail to indicate the presence of an rGE between MAOA-uVNTR genotype and stress measures examined in the current study (see Supplementary Table 1). The next step was to estimate a series of Tobit regression models.…”

“…This state of affairs further underscores the need for adequately powered samples. Taken together, it is reasonable to assume that the size of the sample analyzed in the current study, in addition to the samples analyzed by Wells et al (2017), are all underpowered and thus the cG × E × E effect should be interpreted with caution. This conclusion is all the more compelling in view of the fact that the only statistically significant cG × E effect observed in the present study was in the opposite direction to what is typically reported in the literature.…”

“…Specific details of the current sample and measurements are outlined in the following paragraphs. A direct comparison of methodological features of our study and of Wells et al (2017) is provided in Appendix.…”

“…Items are measured on a 5-point scale (from 1, "Never True," to 5, "Very Often True"). To replicate Wells et al (2017) as closely as possible, only the physical abuse (5 items) and verbal abuse (2 items within the emotional abuse subscale) subscale items were used. The variable was dichotomized to 0 ("Never or rarely experienced abuse"; 63.5%) or 1 ("Experienced abuse at least sometimes"; 36.5%).…”

“…Items are measured on a 5-point scale (from 1, "Never," to 5, "Very Often"). To replicate Wells et al (2017) as closely as possible, items were dichotomized to 0 ("Never," "Almost Never," or "Sometimes") or 1 ("Fairly Often" or "Very Often") before calculating a total sum score, where higher scores indicate higher levels of perceived stress.…”