Implications of rumination for cardiovascular recovery from and adaptation to psychological stress

Background: Rumination has been associated with increased risk of cardiovascular disease (CVD). Potential pathways explaining this link include exaggerated cardiovascular reactions to a single psychological stressor. However, the traditional cardiovascular reactivity (CVR) hypothesis captures only the magnitude of the stress response during the stress exposure and fails to assess the cumulative duration of the stress response that extends beyond the termination of the stressor (cardiovascular recovery) and how responses change with subsequent presentations of the same stressor (cardiovascular adaptation). The variability in the operationalisation of cardiovascular recovery and adaptation across studies is well established in the field but little is done to resolve these issues. Therefore, the aims of this thesis were, 1) to address the measurement issues that exist around the operationalisation of cardiovascular recovery, using a multiverse analysis to identify whether the association between rumination and cardiovascular recovery remained or changed as a function of what computation of recovery was being used, 2) to assess the test-retest reliability of multiple computations of cardiovascular recovery between two separate laboratory visits, 3) to examine the association between rumination and patterns of cardiovascular stress-response adaptation within the same laboratory session and 4) to examine the association between rumination and cardiovascular adaptation to repeated stress between laboratory sessions.

Method: This thesis reports on four empirical studies. Study 1 examined the interaction between trait and state rumination on cardiovascular recovery in a sample of 146 participants who completed a standardized stress testing protocol and had their cardiovascular parameters monitored throughout using a Finometer. Study 2 used previously collected data from the Pittsburgh Cold Study 3 (PCS3) and examined the association between state rumination and cardiovascular recovery from stress. In two laboratory visits (Mean = 48 days apart), participants completed two separate, identical, standardized stress testing protocols and had their blood pressure and heart rate (HR) monitored throughout. Study 3 used the same data as Study 1 to examine cardiovascular adaptation within a single testing session. Study 4 used the same data as Study 2 to examine cardiovascular adaptation across multiple testing sessions.

Results: Study 1 highlights the fragility of significant findings when multiple computations of cardiovascular recovery are considered within the multiverse framework. Therefore, we tentatively note that, high trait rumination was associated with poorer recovery amongst men who also engaged in high levels of state rumination after the stressor. However, we conclude that this interaction effect of rumination on recovery is not robust given that 1) conditional effects were not significant, 2) of the six computations, only one was significant confirming that the status of results strongly depends on the computation of recovery that is chosen, 3) of the five parameters, only one was significant and 4) the sample size for men was relatively small. Study 2 confirms that area under the curve with respect to ground (AUCg) is reliable and potentially the most optimal computation of cardiovascular recovery. It also demonstrates that the rumination-recovery relationship is not in the expected direction with women who engaged in high levels of state rumination exhibiting superior systolic blood pressure (SBP) recovery compared to women who engaged in low levels of state rumination. Study 3 and 4 pinpoint a marked difference in the effects of rumination on cardiovascular adaptation when adaptation is conceptualised as either within-session or between-session. While rumination appears to be maladaptive when encountering the stressor consecutively within the same session, we interpret these findings with caution given the follow-up tests were not significant. However, rumination is considered adaptive when experiencing the stressor across two separate sessions because women who engaged in high levels of state rumination successfully habituated to a repeated exposure to the same task, despite a gap of a number of weeks. On the other hand, women who engaged in low levels of state rumination demonstrated an equally exaggerated diastolic blood pressure (DBP) response to the first and second task, reflecting a failure to habituate to recurrent stress.

Conclusion: Overall, this thesis demonstrates that a broader definition of the reactivity hypothesis to include recovery from and adaptation to stress is warranted. This thesis also provides methodological advancements to the literature on cardiovascular stress responses and rumination. First, addressing the test-retest reliability of multiple computations of cardiovascular recovery between two separate testing sessions can aid in detecting the most stable computations of recovery. Second, the association between rumination and cardiovascular adaptation appears to be somewhat different when adaptation is conceptualised as either within-session or between-session. Third, how rumination is measured and defined is an important methodological consideration for those trying to understand rumination as both a maladaptive and adaptive concept. Finally, rumination potentially functions differentially for men and women in terms of cardiovascular recovery and adaptation, for some cardiovascular parameters.