Population-based health surveys have provided important contributions to our understanding of health and disease. Findings are important for research communities, and policymakers for health rely decisions based on these findings. However, it is impossible to study a whole population. Therefore, a selection of people is chosen to represent the specific population of interest and to allow for generalization of results. But how does the selected group compare to the population which it intends to represent?
Participation rates in population- based research have declined in the past few decades. Participation is based on volunteerism and extensive information on participant characteristics (demographic and health-related) is required. On the other hand, information on non-participants is often lacking. A systematic review conducted in 2009, of articles published in a journal publishing population-based research, showed that only 10% of the studies provided information of non-participants.
Why is it important to gather such information? Not providing information regarding non-participants can compromise the validity of the results. Validity determines to what extend the research measures what it truly intends to measure. Also, it is a great threat to the generalizability of research results if participants differ from non-participants. Generalization refers to what extend the findings can be applied to the wider population.
Read also: Why “Survival of the fittest”?
The Nord-Trøndelag health study is one of the largest population-based studies ever conducted. Three waves, each performed approximately ten years apart, have been performed and a significant reduction in participation rate is reported between all three waves. Data from non-participants has been collected on all three waves. In this case, comparisons between participants and non-participants can be made and guide to what extent the data also represent those choosing not to take part.
Read also: Exercise as medicine in acute lung injury?
Future research should include data from invited people who do not want to take part in population-based studies. This enables comparisons on demographics and important health-related information participants and non-participants. The information is essential for appropriate interpretation of findings and to get a more comprehensive picture for further research.
Trude Carlsen, PhD Candidate at CERG