Movement Behaviors and the Role of Self-Reported Symptoms and Well-Being: A Dynamic Structural Equation Modeling Approach Among Head and Neck Cancer Patients.
Sabina Ulbricht, Antje Ullrich, Christina Grosse-Thie, Daniel Strueder, Markus Blaurock, Chia-Jung Busch, Christian Junghanss, Sabine Felser
Abstract
Open AccessAIM: This study aimed to identify individual variability in day-to-day physical activity (PA) and sedentary behavior (SB) as well as effects of pain, fatigue, and well-being among patients with head and neck cancer (pwHNC) on both movement behaviors. METHODS: A total of 52 pwHNC (age ≥ 18 years, participation rate 73.2%) consented to wear an accelerometer during waking hours for seven consecutive days and to answer questionnaires. Applying a predefined accelerometer protocol resulted in a final analysis sample of 41 pwHNC. A dynamic structural equation modeling approach (DSEM) was employed that combines time series modeling and multilevel modeling. At the within-person level, autoregressive and bidirectional relationships between PA and SB were analyzed. At the between-person level, moderator effects of pain, fatigue, and well-being on these relationships were analyzed. RESULTS: A proportion of 87.8% among the pwHNC were men, mean age 65.8 years, and 72.2% had finished curative care. The average accelerometer wear time was 710.2 ± 153.1 min/d. Of this time, an average of 461.3 ± 151.0 min/d was spent in SB and 248.9 ± 121.6 min/d in PA. At the within-person level, PA and SB data showed no significant autoregressive relationships; that is, previous PA (or SB) did not predict PA (or SB) on the next day, respectively. No bidirectional relationships were found between PA and SB; that is, previous PA did not predict the next-day SB and vice versa. At the between-person level, neither fatigue, pain nor well-being significantly changed the slope and path parameters of PA and SB. CONCLUSION: This study serves as an example of how DSEM can be used to examine how different movement behaviors interact with each other. Presumably, there is a need for more long-term assessment for the detection of consistency in the temporal information of movement behaviors in pwHNC. TRIAL REGISTRATION: German Registry of Clinical Trials (DRKS00028062).