Variability of relationship evaluations & physical health outcomes : testing the moderating role of implicit theories of relationships

Variability (i.e., greater fluctuations) of relationship evaluations over time undermines current and future relationship and individual well-being. To date, greater variability of relationship evaluations has not been linked to individual physical health, yet overall relationship quality (e.g., initial or mean levels) is a strong predictor of several markers of physical health. The aims of the current study are three-fold--first, to explore new methods of capturing intra-individual variability; second, to replicate and extend previous findings on the damaging effects of variability by examining variability's association with relationship fate and physical health outcomes; and third, to investigate the potential moderating role of individuals' general relationship beliefs on these associations. I drew from a 9-month longitudinal study of 202 individuals who were in new dating relationships at the start of study participation (M[subscript length] = 3.28 months). The study included baseline self-reports of relationship satisfaction and Implicit Theories of Relationships (ITRs) as well as bi-weekly self-reports, for a total of up to 20 assessments per individual, of relationship status (together or not) and quality (semantic differential) and physical health (shortened SF-36). A series of analyses were conducted to obtain de-trended levels of intra-individual variability and a previously understudied measure of intra-individual change in relationship evaluations over time -- temporal dependency (i.e., the extent to which one day's relationship evaluation is correlated with the previous day's evaluation). These two variables were subsequently entered into a Cox proportional hazards model to explain the effect of each on the likelihood of relationship dissolution. Both variability and temporal dependency of relationship evaluations (but not their interaction) predicted increased likelihood of experiencing relationship dissolution. Contrary to predictions, however, greater variability and temporal dependency (and their interaction) were not associated with changes in self-reported physical health over time; all three terms, however, predicted higher levels of average self-reported physical health across assessments. Finally, the hypothesized moderating effects of individuals' implicit theories of relationships were not supported. These findings emphasize the predictive power of individuals' patterns of relationship evaluations over time for relationship fate yet suggest additional research is needed to understand the association between variable relationship evaluations and health outcomes.