Chronicity and Timing of Postnatal Depression and Parenting Sensitivity: A Causally Informed Approach

• In Event: Novel Methodological Approaches for Advancing Study of the Intergenerational Transmission of Depression

Abstract

Postnatal depression and co-occurring contextual factors may compromise the cognitive and emotional resources needed for sensitive parenting. Multiple meta-analyses demonstrate associations between maternal depression and less sensitive parenting (e.g., Bernard et al., 2018; Lovejoy et al., 2000), but the wide array of potential confounds makes isolating effects of depression challenging. Further, most studies have not been methodologically equipped to test developmental theories highlighting the importance of chronicity and timing of symptoms in influencing mechanisms of risk (Goodman, 2020). To address these gaps, we employed marginal structural models (MSMs; Robins et al., 2000), an inverse-propensity-of-treatment-weighting (IPTW) approach, to examine whether variations in the chronicity and timing of depression impact parenting sensitivity in early toddlerhood. Based on developmental theories of mother-infant interactions (Beeghly et al., 2016), we hypothesized that chronically elevated depression would have the strongest negative effect on sensitive parenting. Among mothers with more transient depression, we expected depression during the first year postpartum would be more strongly associated with less sensitive parenting than symptoms with onset during the second year (Figure 1a).

We assessed depression with a latent factor derived from the Brief Symptom Inventory-18 (Derogatis, 2000) during home visits at 2, 6, 15, and 24 months postpartum among caregivers (N = 1,292; 99% biological mothers; 40% African American) living in low-income, rural communities (Family Life Project; Vernon-Feagons et al., 2013). At 35 months, parental sensitivity was coded during a mother-toddler interaction (Cox et al., 1999). We first identified time-invariant and time-varying variables that may confound the association between depression and sensitivity (Table 1). We then generated individual MSM weights that rendered depression at each timepoint orthogonal to depression at other timepoints and to included confounds. As such, given the typical IPTW assumptions (e.g., no excluded confounds), maternal depression levels function ‘as if’ repeatedly randomly assigned. We then conducted a weighted linear regression with depression at 2, 6, 15, and 24 months postpartum predicting sensitivity at 35 months. Finally, we statistically compared combined marginal effects of different theoretically relevant profiles of depression (i.e., chronic low, chronic high, early only, and later only) on sensitivity.

Contrary to prior work, our results showed no evidence of ‘single dose’ effects of depression on sensitivity at 35 months (2-months β = -.044, p = .418; 6-months β = -.022, p = .725; 15-months β = -.072, p = .219; 24-months β = -.069, p = .344) postpartum. Rather, differences in sensitivity emerged only after repeated ‘doses’ of heightened (i.e., 75th percentile) maternal depression. Specifically, mothers with chronic depression tended to show less sensitive parenting than mothers with consistently low depression (sensitive parenting was 0.36 SD lower among mothers with chronic depression; χ^2 = 7.764, p = .005), with mothers experiencing heightened depression at two timepoints falling in between.

These results suggest that the chronicity of maternal depression matters for parenting capacities. On average, mothers experiencing transient depression may be able to maintain sensitive parenting, while infants exposed to chronic maternal depression are likely most at risk for the developmental sequelae of less sensitive parenting.

Authors

• Emily Padrutt, University of Minnesota - Twin Cities
  Presenting Author

• Isabella Chask'a Stallworthy, University of Minnesota - Twin Cities

• Meriah L. DeJoseph, University of Minnesota - Twin Cities

• SYLIA WILSON, University of Minnesota - Twin Cities

• Daniel Berry, University of Minnesota - Twin Cities