In infancy, it’s the extremes of arousal that are ‘sticky’

In adult psychology we are quite used to the idea of self-reinforcing mood states: how processes such as rumination or attention biases can act both as a consequence, and cause, of elevated arousal - leading to small initial fluctuations in arousal becoming amplified over time. Weirdly, though - and despite the usefulness of this idea for parents - this idea of self-reinforcing mood states hasn't made its way yet into child psychology. It's long been an aim of mine to take some of the computational processes used, for example, in modelling epilepsy (such as the image above, which is from this paper) to look at this.

I've just managed to finish off a paper which takes a step in that direction. We recorded day-long spontaneous fluctuations in autonomic activity (heart rate, heart rate variability and actigraphy) in 12-month-old infants. We just measured, quite simply, the relationship between a child's overall level of arousal and how stable their arousal state was. Naïvely (because we didn't have much to base a prediction on) we predicted that, if spontaneous arousal fluctuations are random, with fluctuations above and below the mean corrected for via allostasis, then extreme high or low arousal states would be more short-lived than intermediate states. In fact, we found the opposite - that extreme high or low arousal states were more long-lived than intermediate states. 

The vast majority of research in this area focuses on processes of reactivity and regulation - how we return to baseline following an externally induced increase in arousal. We think it's interesting, then, that if you look at how infants' arousal actually fluctuates, you get such a different pattern. Rather than increases and decreases in arousal getting corrected for and pulled back towards the mean, it's more like things getting pulled away from the mean, and sucked towards high or low extremes. 

But there are a couple of different reasons why the data might look like this. First, it might be that extreme high or low arousal states just have intrinscially different hysteresis. Sleep, for example, is an intrinsically more stable arousal state than others. Even though we excluded sleeping sections in this study - we only looked at sections where the infants were at home, and awake - it could be that high and low arousal states are just intrinsically more stable than intermediate states. 

An alternative possibility is that both extreme low and high arousal states may lead to changes in how we interact with the external environment and with people - changes that may in turn lead to extreme arousal states becoming progressively amplified over time. So, for example, an infant in a low state of arousal might engage in fewer interpersonal interactions, and seek out less stimulation from the environment; and a child in a high state of arousal might engage in more interpersonal interactions, including oppositional behaviours; and seek out more stimulation from the environment. 

If that's true, then these 'metastatic' processes - in which small increases and decreases in arousal become amplified over time - might, we think be a useful subject for further study. I'm certainly working at the moment, trying to find new methods to look at this some more!