Early life stress (ELS), primarily encompassing childhood neglect and abuse, is associated with many adverse psychiatric and physical health outcomes in later life. What remains unclear, however, is precisely how these links are mediated. Answering this question is challenging, partly because there are many other exposures that may accompany childhood maltreatment or neglect, but also because there are many physical, social, and other life events that occur between childhood and adulthood which could interact with the effects of early life stress to together result in adulthood pathology.

Here, I conducted a large, controlled experiment in rats that sought to isolate key behavioural, immunological, and neurobiological effects into adulthood of early life stress itself. To do this, I used the repeated maternal separation (RMS) model of chronic early life stress, and I focused particularly on those effects of possible relevance to anxiety, depression, and inflammation-related physical disease.

In Chapter 3, I describe the long-term effects of RMS on commonly used measures of anxiety- and depression-like behaviour, as well as on comparatively sophisticated tasks capable of providing detailed insights into reward and punishment sensitivity, as well as attentional control. The probabilistic reversal learning task revealed long-lasting effects of RMS on the degree to which negative outcomes shaped animals’ future decisions, as well as evidence suggesting that RMS animals were comparatively inefficient at directing their attention, even where they were equally accurate. Further, RMS animals exhibited a long-lasting sensitization to later-life stress on several behavioural metrics. These effects all persisted into late adulthood despite RMS having no effects on conventional measures of anxiety- or depression-like behaviour, even in early adulthood.

In Chapter 4, I present findings from my experiment and from a systematic review examining the short-term and long-term effects of RMS on cytokine levels in blood and non-blood tissue, as well as on microglial activation and density. I show that RMS causes short-term increases in pro-inflammatory signalling, but only causes long-term increases in pro-inflammatory signalling if animals are subjected to a later-life stress. Thus, I demonstrate that RMS causes a long-lasting sensitisation of the neuroimmune pathway that links stressor perception ultimately to pro-inflammatory cytokine release. However, these effects were largely limited to non-blood tissue such as brain tissue: in plasma, serum, or whole blood, studies generally found no effect of RMS on cytokine levels in the short- or long-term, even following later-life stress.

In Chapter 5, I present analyses of regional brain volumes determined from 9.4 Tesla structural magnetic resonance imaging scans at three timepoints following RMS. I show that RMS had no effect on the volume of any of six regions examined at post-natal day (PND) 20 or 62, but resulted in a larger amygdala during the scan at PND 285, which occurred after 9-13 days of adult stress. Given that the PND 62 and PND 285 scans both occurred in adulthood, this suggests that RMS may have interacted with later-life stress to increase amygdala volume.

In the General Discussion, I describe how these findings are concordant and together provide valuable insight into how early life stress can alter physiology and behaviour in such a way that may directly increase risk for mental and physical pathology.