Aim: This study describes the study design, study participants, participation rate and underreporting in the Danish National Surveys of Diet and Physical Activity (DANSDA) from 1995 to 2011-2013.
Methods: DANSDA are government-funded surveys of food and nutrient intake, physical activity and lifestyle, undertaken to support nutritional policy, risk assessment and public health research. The surveys are cross-sectional based on primarily simple random samples (ages 1-80 years in 1995, 4-75 years in 2000-2013) drawn from the Danish Civil Registration System. Approximately 4800 individuals in 1995, 8200 in 2000-2002, 8400 in 2003-2008 and 7300 in 2011-2013 were invited to participate. Participants completed a seven-day food diary, a physical activity questionnaire (2000-2008), a step diary (2011-2013) and a face-to-face interview. Self-reported anthropometrics (1995-2013) were supplemented with device-based measures (2011-2013). Pedometers were included in 2011-2013.
Results: The number of participants included per survey round was 3100-4400. Participant rates decreased from 66% (1995) to 54% (2011-2013). Non-participation was primarily refusal. Ages 18-30 years, 61-75 years (2000-2013), 61-80 years (1995) and low educated and individuals living alone were underrepresented. Underreporting of energy intake among adults ranged from 14% (1995) to 26% (2008).
Conclusions: The methods in DANSDA have been developed to include device-based measures on physical activity and anthropometrics. This has improved the applicability of the results of the survey. The participation rate has fallen, which has affected sample representativity, and underreporting has increased. Future DANSDA surveys should explore and consider new initiatives to counteract non-response and underreporting, with the aim of enhancing data representativeness and applicability.
Keywords: Nutrition; cross-sectional study; data collection; participation rate; population survey; public health; study design; underreporting.