Psychological Health Benefits of Companion Animals Following a Social Loss

Dawn C. Carr, PhD; Miles G. Taylor, PhD; Nancy R. Gee, PhD; Natalie Sachs-Ericsson, PhD


Gerontologist. 2020;60(3):428-438. 

In This Article

Selection Factors

Complicating mixed findings in the literature, the factors that lead older individuals to own a pet (or not own a pet) may be related to how individuals respond to stressful life events. For instance, an individual who has psychological health problems may not be able to adequately care for a CA, and these deficits may also lead to greater vulnerability in stressful situations. These individuals may, in turn, erroneously appear to suffer more psychological deficits after a social loss compared with CA owners. In contrast, individuals might seek out a CA as a means to ameliorate depressive symptoms following a social loss. Such preexisting psychological problems may give the false impression that individuals who obtain a CA after experiencing a social loss fare more poorly.

Various statistical techniques may be used to adjust for these selection factors. One approach involves identification and estimation of the "treatment-effects" of pets on health, accounting for selection factors that predict pet ownership (and nonownership). It would be extremely difficult to conduct a randomized control trial (RCT) to examine the role of pets on the health of middle-aged and older adults; however, models designed to account for selection into CA ownership make it possible to use observational data to mimic an RCT (Rubin, 1974). Propensity score methods, for the present study, use observational data to parse out effects related to pet ownership (Hahn, Todd, & Van der Klaauw, 2001; McCaffrey, Ridgeway, & Morral, 2004; Saunders, Parast, Babey, & Miles, 2017).

An important preliminary study from the authors recently identified a number of underlying factors related to pet ownership, and specifically key sources of heterogeneity in pet owners and nonowners (Carr, Taylor, et al., 2018). Based on data drawn from the human–animal interaction (HAI) experimental module collected in the 2012 wave of the Health and Retirement Study (HRS), we used latent class analysis to identify distinct clusters of CA owners and nonowners. The typologies revealed important variation in factors likely to select individuals into pet ownership/nonownership, suggesting that neglecting to account for these selection factors is likely to lead to biased findings and potentially inaccurate conclusions.