@phdthesis{Ullrich2019,
  author    = {Antje Ullrich},
  title     = {The Role of Sedentary Behavior in Cardio-Preventive Research},
  journal    = {Die Bedeutung von sitzendem Verhalten in der kardio-pr{\"a}ventiven Forschung},
  url       = {https://nbn-resolving.org/urn:nbn:de:gbv:9-opus-39369},
  pages     = {101},
  year      = {2019},
  abstract  = {Background: Sedentary behavior (SB) is a modifiable behavior with increasing prevalence worldwide. There is emerging evidence that time spend in SB and the manner in which SB is accumulated over time is associated with cardiovascular and cardiometabolic health. The requirement for SB data to be accurately measured is minimization, or at least accurate quantification of human-related sources of measurement errors such as accelerometer measurement reactivity (AMR). The present thesis was to examine SB and their associations with cardiovascular and cardiometabolic health, and to focus on challenges related to the assessment of SB. The first aim of the thesis was to identify patterns of SB describing how individuals accumulate their time spend in SB day-by-day over one week, and to examine how these patterns are associated with cardiorespiratory fitness as a marker for cardiovascular health (paper 1). The second aim of the thesis was to examine the multiple types of SB, and how this is associated with a clustered cardiometabolic risk score (CMRS; paper 2). The third aim of the thesis was to examine AMR and the reproducibility in SB and physical activity (PA) in two measurement periods, and to quantify AMR as a confounder for the estimation of the reproducibility of SB and PA data (paper 3). Methods: The three papers were based on data of two different studies. For study 1, 1165 individuals aged 40 to 75 years were recruited in three different settings. Among these, 582 participated in a cardiovascular risk factor screening program including cardiopulmonary exercise testing. For the analyses of paper 1, 170 participants were eligible, agreed to wear an accelerometer, fulfilled the wearing regime, and completed the study period by wearing the accelerometer for seven consecutive days. Patterns in accelerometer data were classified based on time spent in SB per day applying growth mixture modeling. Model‐implied class‐specific peak oxygen uptake (VO2peak) means were compared using adjusted equality test of means (paper 1). The underlying study of paper 2 and 3 were based on data of a pilot study aiming to investigate the feasibility of a brief tailored letter intervention to increase PA and to reduce SB during leisure time. Among the individuals who agreed to be contacted again in study 1, a random sample of those aged between 40 and 65 years was drawn. Of those, 175 attended in a cardiovascular examination program. Assessment included giving blood sample, standardized measurement of blood pressure, waist circumference, body weight, and height at baseline, and after twelve months. Further, they agreed to complete a paper-pencil questionnaire on SB (Last 7-d Sedentary Behavior Questionnaire, SIT-Q-7d) and PA (International Physical Activity Questionnaire, IPAQ), and to receive seven-day accelerometery at baseline, and after 12 months. In addition, self-administered assessments were conducted at months one, three, four, and six after baseline. Only individuals of a random subsample (= intervention group) received up to three letters tailored to their self-reported SB and PA at months one, three, and four. For paper 2, associations between SBs and a clustered cardiometabolic risk score (CMRS) were analyzed using linear as well as quantile regression. To account for missing values at baseline, multiple imputations using chained equations were performed resulting in a total sample of 173 participants. Paper 3 comprised data of 136 individuals who participated at the baseline and twelve months assessments, and fulfilled the wearing regime. AMR was examined using latent growth modeling in each measurement period. Intraclass correlations (ICC) were calculated to examine the reproducibility of SB and PA data using two-level mixed-effects linear regression analyses. Results: Results of paper 1 revealed four patterns of SB: 'High, stable', 'Low, increase', 'Low, decrease', and 'High, decrease'. Persons in the class 'High, stable' had significantly lower VO2peak values (M = 25.0 mL/kg/min, SD = 0.6) compared to persons in the class 'Low, increase' (M = 30.5 mL/kg/min, SD = 3.6; p = 0.001), in the class 'Low, decrease' (M = 30.1 mL/kg/min, SD = 5.0; p = 0.009), and in the class High, decrease' (M = 29.6 mL/kg/min, SD = 5.9; p = 0.032), respectively. No differences among the other classes were found. In paper 2, results revealed that the only factor positively associated with a CMRS in all regression models was watching television. Depending on the regression analysis approach used, other leisure-time SBs showed inconsistent (using a computer), or no associations (reading and socializing) with a CMRS. In paper 3, results revealed that time spent in SB increased (baseline: b = 2.3 min/d; after 12 months: b = 3.8 min/d), and time spent in light PA decreased (b = 2.0 min/day; b = 3.3 min/d). However, moderate-to-vigorous PA remained unchanged. Accelerometer wear time was reduced (b = 4.6 min/d) only at baseline. The ICC coefficients ranged from 0.42 (95\% CI = 0.29 - 0.57) for accelerometer wear time to 0.70 (95\% CI = 0.61 - 0.78) for moderate-to-vigorous PA. None of the regression models identified a reactivity indicator as a confounder for the reproducibility of SB and PA data. Conclusions: The present thesis highlights SB in the field of cardiovascular and cardiometabolic research that have implications for future research. Individuals sit for different purposes and durations in multiple life domains, and the time spent in SB is accumulated in different patterns over time. Therefore, research should consider the fact that SB is embedded in an individual's daily life routine, hence might have differential effects on cardiovascular and cardiometabolic health. Further, methodological aspects have to be considered when dealing with SB. In order to detect how SB is 'independently' associated to an individual's health, an accurate measurement of SB is fundamental. Therefore, human-related sources of bias such as AMR should be taken into account when either planning studies or when interpreting data drawn from analysis of SB data.},
  language  = {en}
}