Combining Life History Calendars and Ecological Momentary Assessment in the Evaluation of Stress in Everyday Life (Preprint)

BACKGROUND Stress is widely acknowledged as a risk factor for various negative health outcomes. Therefore, assessing everyday stress through longitudinal research has gained interest, with a focus on capturing stress and its components using intra-individual approaches and ecological momentary assessment (EMA). OBJECTIVE Building on the proposal of Smyth et al. [1], our study aims to investigate the relationship between different operationalizations of stress components with two relevant constructs: declining mental health and trait anxiety. METHODS Over a 6-month period, we conducted a longitudinal study involving 165 adults (84 females; mean age = 24.91, SD = 4.61 years) with varying anxiety levels. We assessed retrospective stressor evaluations using a life history calendar (LHC) and prospective stress responses via EMA. Our LHC surveyed forty events representing significant changes in the participant´s environment across thirteen domains. EMA anxiety scores were derived from weekly averages of four items measuring anxiety symptoms. After defining three different baselines, we computed indices of pileup, average stress reactivity, and average stressor recovery for each participant, employing various operationalizations. Subsequently, we examined the relationship between these operationalizations and symptom changes using hierarchical multiple regression analyses. RESULTS After data cleaning from 89100 potential notifications (165 retained participants × 180 days × three notifications per day), participants responded to 68554 (77%) of them. Every participant reported experiencing at least one event from the LHC list during the study. Interestingly, different stress reactivity baselines influenced both the total count of stressor episodes for the whole sample (239 with local baseline 1, 176 with local baseline 2, and 228 for cumulative baseline) and the number of stressor episodes per person (local baseline 1, M=1.45, SD=.74; local baseline 2, M=1.07, SD=.25; cumulative baseline, M=1.41, SD=.78). Only a few stress components operationalizations showed the expected associations with increased internalizing symptoms or trait anxiety. When defining a stressor as an event with increased stress reactivity and operationalizing pileup using either local (one week) or cumulative baseline, there was a correlation observed between pileup and stress scores. Additionally, two operationalizations of average stress reactivity correlated with anxiety and stress scores. Moreover, the operationalization using cumulative recovery and the local baseline 2 significantly predicted stress symptoms at follow-up. Finally, only one operationalization, which used the number of weeks with stressors, supported the expected association of trait anxiety with pileup (r=.17, P=.03). CONCLUSIONS Our results underscore the importance of selecting appropriate intra-individual baselines for capturing stress dynamics in everyday life. To our knowledge, this is the first study to combine LHC for retrospective stressor assessment with prospective longitudinal assessment of stress responses using EMA. Future research can benefit from these insights by utilizing the most effective operationalizations identified here and investigating alternative methodologies.