Affect variability and predictability: Using recurrence quantification analysis to better understand how the dynamics of affect relate to health.

Changes in affect over time have been associated with health outcomes. However, previously utilized measurement methods focus on variability of affect (e.g., standard deviation, root mean squared successive difference) and ignore the more complex temporal patterns of affect over time. These patterns may be an important feature in understanding how the dynamics of affect relate to health. Recurrence quantification analysis (RQA) may help alleviate this problem by assessing temporal characteristics unassessed by past methods. RQA metrics, such as determinism and recurrence, can provide a measure of the predictability of affect over time, indexing how often patterns within affective experiences repeat. In Study 1, we first contrasted RQA metrics with commonly used measures of variability to demonstrate that RQA can further differentiate among patterns of affect. In Study 2, we analyzed the associations between these new metrics and health, namely, depressive and somatic symptoms. We found that RQA metrics predicted health above and beyond mean levels and variability of affect over time. The most desirable health outcomes were observed in people who had high mean positive affect, low mean negative affect, low affect variability, and high affect predictability. These studies are the first to demonstrate the utility of RQA for determining how temporal patterns in affective experiences are important for health outcomes. (PsycInfo Database Record (c) 2020 APA, all rights reserved).