Do different data analytic approaches generate discrepant findings when measuring mother-infant HPA axis attunement?

Attunement between mothers and infants in their hypothalamic-pituitary-adrenal (HPA) axis responsiveness to acute stressors is thought to benefit the child's emerging physiological and behavioral self-regulation, as well as their socioemotional development. However, there is no universally accepted definition of attunement in the literature, which appears to have resulted in inconsistent statistical analyses for determining its presence or absence, and contributed to discrepant results. We used a series of data analytic approaches, some previously used in the attunement literature and others not, to evaluate the attunement between 182 women and their 1-year-old infants in their HPA axis responsivity to acute stress. Cortisol was measured in saliva samples taken from mothers and infants before and twice after a naturalistic laboratory stressor (infant arm restraint). The results of the data analytic approaches were mixed, with some analyses suggesting attunement while others did not. The strengths and weaknesses of each statistical approach are discussed, and an analysis using a cross-lagged model that considered both time and interactions between mother and infant appeared the most appropriate. Greater consensus in the field about the conceptualization and analysis of physiological attunement would be valuable in order to advance our understanding of this phenomenon.