Study design
This study used cross-sectional data from the 2019 follow-up survey of a cohort of patients with diabetes (CoDiab-VD) and the 2017 Swiss Health Survey (SHS). CoDiab-VD is a community-based cohort of individuals with diabetes originally recruited in pharmacies in 2011-2012 and 2017. The objectives of CoDiab-VD were to provide an overview of that population and of its diabetes management in the canton of Vaud, Switzerland (22). The SHS is a national population-based survey carried out every five years since 1992 with the objectives of monitoring the health of the general population and assessing the adequacy of prevention and health promotion programs (23).
Population and setting
Data from three groups of survey participants were considered. Two groups included people with diabetes, and the third included individuals from the general population. The groups are described below.
CoDiab-VD: All respondents of the 2019 follow-up survey (n=649) were included in the analysis. At recruitment, the inclusion criteria were being a noninstitutionalized adult (≥18 years) diagnosed with any type of diabetes for at least 12 months, living in the canton of Vaud, and giving consent to participate. Individuals with gestational diabetes only, insufficient knowledge of the French language, and obvious cognitive impairments were excluded.
SHS: Two groups of respondents were considered from this survey: the whole sample of noninstitutionalized individuals aged 18 years or older (n=21,430), and a subsample including only those reporting diabetes (n=951). The latter were those having answered yes to the question: “Has a medical doctor ever told you that you have diabetes?” or having answered that they took diabetes medication or insulin during the last 7 days. In the general SHS population were also included people living with diabetes, as we wanted to include the general population as a whole, including all people suffering from chronic diseases.
Measurements
Self-reported health, the main outcome (dependent) variable, was assessed in the CoDiab-VD survey using the first question of the SF-12 questionnaire “In general, would you say your health is?”, with five response options (excellent, very good, good, fair, poor) (24). A similar question was used in the SHS: “How is your overall health?”, also with five response options (very good, good, average, poor, very poor).
The main exposure (independent) variables were measures of physical activity and sedentary behaviors used in both the SHS and the CoDiab-VD survey.
The following four questions were used to measure physical activity and allowed the construction of a five-category score of physical activity levels (25):
- “In your leisure time, do you engage at least once a week in a physical activity that makes you sweat such as running, cycling, playing sports, etc.?”
- “On average, how many days per week?”
- “If you can think of physical activities in which you are at least a little out of breath, such as brisk walking, hiking, dancing, gardening, or various sports: how many days per week do you engage in such physical activities?”
- “What is the average duration of these physical activities during these days?”
Categories of physical activity per week, as coded by the Federal Statistical Office, were: inactive: < 30 minutes of moderate physical activity or < 1 time of intense activity; partially active: 30-149 minutes of moderate physical activity or 1 time of intense activity; irregularly active: ≥ 150 minutes of moderate physical activity, or 2 times intense physical activity; regularly active: ≥ 150 minutes of moderate physical activity (at least 5 days per week, at least 30 minutes); trained: at least 3 times intense physical activity (25).
The two different sedentary behavior variables related to the following two questions:
- “In total, how many hours do you sit during a typical weekday (Monday to Friday only)?” (Sitting hours: continuous; categorized for the analyses as 0-3 h, 4 h, 5 h, 6-7 h, 8-16 h).
- “How often do you interrupt your sitting position, i.e., how often do you stand up?” (Sitting interruption: every 5-10 min, every 15 min, every 30 min, every hour, less than every hour).
Variables considered in the analyses, either as descriptive variables or potential confounders, had to be available in both the SHS and CoDiab-VD questionnaires, and included age (continuous), sex (male/female), living status (alone: yes/no), educational status (primary: completion of mandatory school or less; secondary: vocational training or high school; tertiary: university or technical college), monthly household income (≤3399 CHF; 3400-5799 CHF; 5800-9599 CHF; ≥9600 CHF; unknown), current smoker (yes/no), BMI (underweight and normal: <25.0 kg/m2; overweight: 25.0-29.9 kg/m2; obese: ≥30.0 kg/m2), and the number of comorbidities (0; 1; 2; ≥3).
Statistical analysis
Analyses were conducted independently in the three populations of interest. In the two samples stemming from the SHS, weighted analyses were conducted using the weights provided by the Federal Statistical Office; the latter weights improved the representativeness of the samples.
First, descriptive uni- and bivariate analyses were conducted. Then, using ordinal logistic regression models, the associations between the exposure variables and the self-reported health variables were explored. Unadjusted models were used, followed by adjusted models including potential confounding variables (age, sex, living alone, educational status, monthly household income, current smoker, BMI, and the number of comorbidities). The significance of the interactions between physical activity and sedentary behavior variables was tested in the adjusted models. A p-value ≤0.05 was considered statistically significant. The analyses were performed using Stata 16.1 (26) and R statistical software (27).