I am going to quote from the discussion, and show the diagram, then comment.

In accordance with our hypothesis and previous cross-sectional studies, we observed greater depressive symptom severity among individuals with the s/s genotype who reported a history of severe child abuse compared with those without a history. We also provide novel evidence which suggests that the s/s genotype confers a persistent moderating effect on depressive symptom severity without alteration to the rate of change over time. There was no significant difference in depressive symptoms between those with or without a history of child abuse in the s/l or l/l genotype groups. These findings support the notion that long-lasting consequences of adverse childhood experiences may be dependent on an individual’s genomic context. More specifically, these findings suggest that the s/s genotype may be associated with susceptibility to more severe depressive symptom trajectories in adulthood following exposure to severe abuse in childhood.

Among the s/s group, those with a history of severe, child abuse had a mean PHQ-9 score greater than 10 (i.e. moderate depression) at each time point over the 5-year period, whereas those without a history of child abuse consistently scored under 10 (i.e. mild depression), a differential effect not observed among s/l and l/l genotype carriers. However, among individuals without a history of severe child abuse, we detected a trend in which s/s genotype carriers had more favourable depressive symptom trajectories compared with s/l and l/l genotype carriers. This pattern of results suggests that the 5HTTLPR polymorphism may be a marker of ‘phenotypic plasticity’ rather than ‘vulnerability’ in that s/s carriers appear to be the most susceptible to the negative effects of severe child abuse (i.e. increased depressive symptom severity) but also marginally more likely to benefit from the absence of a severe child abuse experience, albeit only a trend difference (P=0.068) was seen between s/s carriers and their l/l or l/s carrying counterparts in the absences of severe child abuse. Nonetheless, this so-called ‘differential susceptibility’ has been observed previously. Most notably, in the landmark paper by Caspi et al in that s/s genotype carriers with stressful life events reported the most depressive symptoms but conversely those without such events reported the fewest depressive symptoms.

Longitudinal measurements of depressive symptoms over 5 years by 5HTTLPR and history of severe child abuse. Points are observed mean PHQ-9 scores with standard error bars. Lines are predicted values with 95% confidence intervals shaded based on linear mixed model analysis. Dashed line/circles, no history of severe child abuse; solid line/squares, history of severe child abuse; PHQ-9, Primary Care Evaluation of Mental Disorders Patient Health Questionnaire-9; pU, covariate-unadjusted P-value; pA, covariate-adjusted P-value. P-values based on estimated marginal means.

The papers by Caspi on depression and the short serotonin genome (and the MAOI variation with antisocial behaviour, and similar work with genome and psychosis) do not explain why some people get depressed and others do not. What they demonstrate is that the genome is not your fate, nor is your life experience. It is the interaction, and this is not simple. These variations do not explain all the reasons why one may suffer and another may not.

We simply don’t know enough.
We may never know enough.
But this I do know: the rate of depression and distress is increasing, and the correlation seems to be with diversity. Replacing traditions and the traditional population of a place causes harm. Avoiding this does not.