Design
This study is based on secondary analysis of a cross-sectional survey of Northern Irish adults conducted in 2009/10 and known as the Northern Ireland Sport and Physical Activity Survey (SAPAS). This was the first nationwide survey specifically on sport and physical activity, commissioned by Sport Northern Ireland and carried out by an independent market research agency (IPSOS Mori). 4653 adults (aged 16+) completed face-to-face interviews conducted in their homes using computer assisted personal interviewing. This study involved the mining of an existing data set collected by Ipsos MORI who adhere to a code of conduct that encompasses ethical and legislative principles in the UK. Permission to use the data was granted by Sport Northern Ireland. The authors did not collect the data and therefore did not require ethical approval.
Sample
The sampling procedures ensured proportionality with the Northern Ireland population based on estimates of the number of residents aged 16 or older provided by the Census Office for Northern Ireland (1.4 million). The sample was stratified by local council area (26 strata) and random samples of households, selected from the Royal Mail’s Postal Address File, were drawn within each stratum. The sampling fraction was the same for each council area except for Belfast and Derry the two major cities, where oversampling was used to facilitate detailed analysis. Within selected households, one adult was randomly selected using the last birthday rule. Participants completed a detailed interview (average time - 30 minutes) with a trained interviewer. 4,663 interviews were completed, representing a response rate of 54.6%. The dataset was weighted to control for the oversampling in the Belfast and Derry strata, differences in household size, and to ensure that the age/gender closely matched that of the adult population of Northern Ireland.
Survey description
The survey instrument was designed in partnership with Sport Northern Ireland and was cognitively tested and piloted by the market research agency. The survey was based on the Active People Survey (Sport England) conducted annually by Sport England since 2005-06. The 30 minute interview collected data on participation in sport and physical activity, perceived health and happiness, fruit and vegetable intake, alcohol consumption and smoking habits as well as sociodemographic information. The list of physical activities was taken from the Department of Culture Media and Sport (DCMS) and Sport England (Ipsos MORI).
Piloting and cognitive testing consisted of interviews with 30 respondents covering both genders and a wide range of ages, employment statuses and levels of sport participation. A mix of ‘think aloud’ (whereby respondents are instructed to verbalise their thought processes as they answer the survey questions) and ‘verbal probing’ techniques were employed, which were adapted to suit individual respondents (with both concurrent and retrospective probing). As a result of this, some survey items were amended to maximise recognition, recall and decision-making by respondents.
Outcome measures
Domestic physical activity
Participants were asked to recall how long (in minutes) they spent walking and cycling, being physically active at work and in or around the home, and participating in sport during the previous 7 days. For each activity participants were asked to classify the intensity of physical activity depending on whether or not participation raised breathing or heart rate and to report only bouts of 10 minutes or more.
Participants were asked about any physical activity in the home (domestic physical activity) that raised breathing rate over the last 7 days. Domestic physical activity was classified into 4 categories; housework, DIY, gardening or other activity. Total time spent in domestic physical activity in bouts of 10 minutes or more was reported in minutes. For each activity, participants were asked to report whether the effort they put into the activity was “usually enough to make them out of breath or sweat”. On the basis of this assessment, activity was classified as low intensity physical activity or moderate to vigorous intensity physical activity (MVPA).
Body mass index
Self-reported height and weight were used to calculate Body Mass Index (BMI). The inverse of Body Mass Index (iBMI) was calculated by the formula height (cm) 2/weight (kg). We chose iBMI as our response variable as this ratio has been shown to be more linearly related to, and hence able to explain more of the variance in, percentage body fat. It also has the advantage of being more symmetric, better approximated by the normal distribution and hence more suitable than BMI in detecting differences in fatness/leanness when adopted in statistical/epidemiological studies [13].
Statistical analyses
Because time spent doing moderate to vigorous domestic physical activity and the percentage of time reported doing domestic MVPA activity as a ratio of total MVPA were not normally distributed, gender and age group differences in both variables were explored using non-parametric Kruskal-Wallis tests of significance.
We also explored the effect of various sources of physical activity (at work, at home, sporting, cycling, and walking activities) on iBMI using ANCOVA, having controlled for the confounding effects of age, gender, socio-economic and smoking status. The categorical variables (gender, age group, socio-economic and smoking status) were entered as fixed factors and the continuous variables (time spent at various types of exercise) were incorporated as continuous linear covariates. Time spent at various types of exercise were entered initially as time spent at any level of exercise intensity and subsequently as time spent at a moderate or higher level of exercise intensity only.