Voigt Lisa, Baumann Sophie, Ullrich Antje, Weymar Franziska, John Ulrich, Ulbricht Sabina
1Institute of Social Medicine and Prevention, University Medicine Greifswald, Greifswald, Walther-Rathenau-Str. 48, D-17475 Greifswald, Germany.
2Partner site Greifswald, German Centre for Cardiovascular Research (DZHK), Fleischmannstr. 42-44, D-17475 Greifswald, Germany.
BMC Sports Sci Med Rehabil. 2018 Jan 23;10:1. doi: 10.1186/s13102-018-0090-8. eCollection 2018.
Measuring physical activity (PA) and sedentary time (ST) by self-report or device as well as assessing related health factors may alter those behaviors. Thus, in intervention trials assessments may bias intervention effects. The aim of our study was to examine whether leisure-time PA, transport-related PA, and overall ST measured via self-report vary after assessments and whether a brief tailored letter intervention has an additional effect.
Among a sample of subjects with no history of myocardial infarction, stroke, or vascular intervention, a number of 175 individuals participated in a study comprising multiple repeated assessments. Of those, 153 were analyzed (mean age 54.5 years, standard deviation = 6.2; 64% women). At baseline, participants attended a cardiovascular examination (standardized measurement of blood pressure and waist circumference, blood sample taking) and wore an accelerometer for seven days. At baseline and after 1, 6, and 12 months, participants completed the International Physical Activity Questionnaire. A random subsample received a tailored counseling letter intervention at month 1, 3, and 4. Changes in PA and ST from baseline to 12-month follow-up were analyzed using random-effects modelling.
From baseline to 1-month assessment, leisure-time PA did not change (Incidence rate ratio = 1.13, = .432), transport-related PA increased (Incidence rate ratio = 1.45, = .023), and overall ST tended to decrease (b = - 1.96, = .060). Further, overall ST decreased from month 6 to month 12 (b = - 0.52, = .037). Time trends of the intervention group did not differ significantly from those of the assessment-only group.
Results suggest an effect of measurements on PA and ST. Data of random-effects modelling results revealed an increase of transport-related PA after baseline to 1-month assessment. Decreases in overall ST may result from repeated assessments. A brief tailored letter intervention seemed to have no additional effect. Thus, measurement effects should be considered when planning intervention studies and interpreting intervention effects.
ClinicalTrials.gov NCT02990039. Registered 7 December 2016. Retrospectively registered.
通过自我报告或设备测量身体活动（PA）和久坐时间（ST）以及评估相关健康因素可能会改变这些行为。因此，在干预试验中，评估可能会使干预效果产生偏差。我们研究的目的是检验通过自我报告测量的休闲时间PA、交通相关PA和总体ST在评估后是否会发生变化，以及一封简短的个性化信件干预是否有额外效果。
在没有心肌梗死、中风或血管介入史的受试者样本中，175人参与了一项包含多次重复评估的研究。其中，153人被纳入分析（平均年龄54.5岁，标准差 = 6.2；64%为女性）。在基线时，参与者接受了心血管检查（标准化测量血压和腰围、采集血样）并佩戴加速度计7天。在基线以及1、6和12个月后，参与者完成国际身体活动问卷。一个随机子样本在第1、3和4个月接受了个性化咨询信件干预。使用随机效应模型分析从基线到12个月随访期间PA和ST的变化。
从基线到1个月评估，休闲时间PA没有变化（发病率比 = 1.13，P = 0.432），交通相关PA增加（发病率比 = 1.45，P = 0.023），总体ST有下降趋势（b = -1.96，P = 0.060）。此外，总体ST从第6个月到第12个月有所下降（b = -0.52，P = 0.037）。干预组的时间趋势与仅接受评估组没有显著差异。
结果表明测量对PA和ST有影响。随机效应模型结果的数据显示，从基线到1个月评估，交通相关PA有所增加。总体ST的下降可能是重复评估导致的。一封简短的个性化信件干预似乎没有额外效果。因此，在规划干预研究和解释干预效果时应考虑测量效应。
ClinicalTrials.gov NCT02990039。2016年12月7日注册。追溯注册。
