Understanding determinants of nutrition, physical activity and quality of life among older adults: the Wellbeing, Eating and Exercise for a Long Life (WELL) study

Nutrition and physical activity are major determinants of health and disease and are associated with risk of premature mortality, coronary heart disease, hypertension, colon cancer, type 2 diabetes, osteoporosis and weight gain [3]. Promoting physical activity and a healthy diet thus has the potential to substantially reduce the burden of disease and improve quality of life. Currently older adults consume too few fruits and vegetables, and have lower than recommended intakes of a range of nutrients important for prevention of chronic disease [4]. It is also estimated that approximately 45% of adults are not sufficiently active to achieve health benefits and older adults are less likely to participate in ‘sufficient’ physical activity than younger adults [5].

There are a number of specific issues relevant to nutrition and physical activity behaviours of older adults. Nutritional needs change during older age with the required intakes of many nutrients increasing alongside a decreased energy requirement [6]. Therefore, the quality of diet with food choices based on nutrient-dense foods becomes increasingly important, particularly for the avoidance of weight gain. In addition, there is an increasing use of medications with potential for interactions with dietary intake and nutritional status [6]. Of particular significance with respect to physical activity, age-associated loss of muscle mass can result in reduced muscle strength in older persons [7] and is a major cause of their increased disability prevalence [8]. Increased physical activity is a potentially important strategy among older adults for maintaining functional status and independence [9]. Later adulthood is a critical period for promotion of nutrition and physical activity, as chronic disease will typically present during this life-stage, there are immediate benefits to improving chronic disease risk profiles and there is an ability to maximize health by avoiding or delaying preventable disability [3].

As well as the biological changes that accompany ageing, it is a period of social and psychological transition. During older adult life, there are a number of transitions that can lead to substantial lifestyle changes which may directly or indirectly impact on health including retirement, relationship breakdown or partner loss and changing household structures (“empty nest”). Populations undergoing transitional life-stages are at increased risk of poor health due to potential changes to lifestyle that impact negatively on nutrition and physical activity behaviours [10, 11]. A life-course approach to prevention is necessary to develop interventions that are relevant to each stage of life, with strategies that are age-appropriate [3]. Existing research in the area of health and well-being of older adults focuses on the predictors or risk factors for chronic disease and use of health services [12] and there is little research focusing on the influences of nutrition and physical activity behaviours. In addition, there are few studies assessing nutrition and physical activity behaviours longitudinally among older adults. Longitudinal research is important for enabling tracking of behaviours and their determinants during this period of potential transition and the development of causal theoretical models of health behaviour.

Socioeconomic differentials in health including those relating to obesity are well recognised [13]. Similarly, nutrition and physical activity behaviours are known to vary according to socioeconomic position. There is little research on the mechanisms underlying socioeconomic variations in nutrition and physical activity behaviours specific to the older adult population group and how socioeconomic differentials in these behaviours are impacted by the life events typical in this life-stage, such as retirement [14, 15].

In addition, rural populations suffer higher rates of socioeconomic disadvantage with lower incomes, and lower levels of educational attainment. Older adults living in rural areas have worse health compared with those in cities with lower life expectancies and higher rates of illness and disease [16]. People living in rural areas face particular challenges which impact upon health, including social isolation, limited access to transport, facilities and services [16]. The rural population is particularly susceptible to the problems associated with an ageing population since rural areas have a higher proportion of older adults compared to urban areas, driven by a combination of inward migration of older adults and outward migration of young people [16].

A variety of models have been applied to the study of health behaviour, such as the theory of planned behaviour, social cognitive theory and the transtheoretical or “stages of change” model [17]. A broader framework is the social ecological model [18] which acknowledges the environment in which the behaviours occur [19] and that there is a need to consider the influence of factors in the social and physical environment, the inter-relationships between environmental and intrapersonal influences, and the ability of the individual to adapt to these influences.

Intrapersonal factors such as self-efficacy, enjoyment, barriers and intentions in relation to nutrition and physical activity and social influences such as social support and sabotage are thought to be important influences on nutrition and physical activity behaviours [20]. However, there is little research concerning these influences among older adults and just as nutrition and physical activity behaviours vary according to subpopulation, it is likely that the determinants also vary. For example, in cross-sectional studies of mid-aged and older adults, nutrition knowledge [21], self-efficacy, family support factors [22] and aspects of the environment have been shown to be associated with eating behaviours [23]. However existing studies focus on broad age ranges (>40 years) and are not specifically focused on older adults in the peri-retirement phase and therefore it is necessary for research on the influences on nutrition and physical activity behaviours to take into account the specific life-stage context [24]. Furthermore, conducting research in the Australian context is important for the development of appropriate strategies and interventions and may be particularly important when trying to understand interactions between intrapersonal, social and environmental influences as important cross-country variations in some determinants have been demonstrated [25].

Participants were selected from the Australian Electoral Commission (AEC) using a stratified random sampling process. In Australia, voting is compulsory for persons aged 18 years and over, and the AEC estimates that the electoral role represents 89.7% of those who were eligible to enrol and vote [26]. Suburbs in Victoria were classified as urban or rural using a classification consistent with the Australian Regional Infrastructure Development Fund Act 1999 [27] and suburbs with populations of less than 1000 or less than 200 55-65 year olds were excluded. All suburbs in urban and rural areas were then classified according to the socioeconomic Index for Areas score (SEIFA) which is assigned by the Australian Bureau of Statistics [28], and divided into tertiles (i.e low, medium and high SEIFA). Fourteen postcodes from each SEIFA tertile (i.e low, medium and high SEIFA) were randomly selected and an equal number of participants from areas from each tertile of SEIFA score selected. From each suburb, 134 participants (equal numbers of men and women) were selected, resulting in a total sampling pool of 11256. Of the surveys distributed, 380 were returned as undeliverable and 95 were returned from individuals outside of the 55-65 year age bracket. In total, 4,082 completed surveys were returned (38% response rate). Table 1 shows the sociodemographic characteristics of our final sample.

Participants selected from the electoral role were sent a letter inviting them to participate in the study and one week later were sent the survey and a reply-paid envelope for survey return. After three weeks, non-respondents received a reminder letter encouraging them to return their questionnaire. After a further three weeks, the remaining non-respondents received a second reminder letter and a replacement questionnaire and reply-paid envelope. This process of sending two reminders is standard practice [29, 30].

Participants will be re-contacted at T2 and T3 and the same procedures and protocols for postal survey administration will be used. Follow-up after two and four years will allow sufficient time to detect changes in weight, nutrition and physical activity during this life-stage [31]. Recruitment and retention are promoted via media releases in the local survey areas, personalised survey letters, newsletters to participants with details of study results, birthday cards and access to a study website and phone number for information and change of address.

The questionnaire was designed to include measures of the outcome variables (nutrition, physical activity, sedentary behaviour, obesity, quality of life), potential determinants of these outcomes and relevant covariates. The social ecological model was used as a framework for development of the questionnaire and selection of the range of potential determinants of nutrition and physical activity behaviours [24]. Items on the questionnaire examined all three levels of influence on nutrition and physical behaviours (intrapersonal, social and neighbourhood environmental influences). Where possible, established questionnaire items from the literature with known reliability and validity were used. The full range of measures included in the questionnaire are shown in Table 2 and key variables are summarised here.

The Medical Outcomes Study Short-Form General Health Survey (SF-36) is included as a measure of quality of life [32‐34]. Scores for General Health, Physical Health, and Mental Health are computed. The Physical Health Component includes physical functioning, role-physical, bodily pain, and general health. The Mental Health Component includes vitality, social functioning, role-emotional and mental health. The questions were altered to Australian conditions in line with the Australian Longitudinal Study on Women’s Health [33, 35, 36].

Measures of self-reported height and weight were collected. Self-reported height and weight data are strongly correlated with measured height and weight r > 0.9 [37]. Self-reported weight and height information is adequate for use in large epidemiological studies examining weight or body mass index [38‐40] and has been used in several large cohorts in Australia and worldwide to investigate weight change [41, 42].

Diet was measured using a 111-item food frequency questionnaire assessing usual frequency of intake of food and beverages over the last 6 months previously developed for use with Australian adults in the National Nutrition Survey and other national surveys [43‐45]. Additional validated short questions on food habits concerning breakfast consumption, salt use, type of milk consumed, trimming the fat from meat, daily fruit and vegetable consumption and food security were also included [45, 46].

Physical activity in the past week was assessed using the long version of the self-administered International Physical Activity Questionnaire (IPAQ-L). This survey demonstrated excellent one-week test-retest reliability (pooled r = 0.81) and acceptable validity (pooled r = 0.33) when compared to accelerometer-measured physical activity in a 12-country, 14-site study [47]. The IPAQ-L assesses duration, frequency and intensity of leisure, work, commuting and household/yard activities. Data on total sitting time were also collected from the IPAQ-long with respondents asked to report time spent sitting while at work, at home, while doing study, and in leisure-time during the last 7 days [47, 48] Respondents were also asked to report sitting time while doing specific activities (watching tv and during computer activities) [49].

Demographic variables that were considered to be important potential moderators or confounders of the associations between behavioural predictors and outcomes were measured. These included age, country of birth, marital status, measures of socioeconomic position (education, employment, own and household income, postcode as an area level measure of socioeconomic position) [50, 51], retirement status, household composition and living arrangements.

In relation to nutrition and physical activity, the questionnaire included measures of self-efficacy [52], enjoyment, barriers and intentions [53, 54], outcome expectancies[55], perceived behavioural control [54] and nutrition knowledge [56]. It also included measures of perceptions of ageing and retirement [57].

The questionnaire assessed support and sabotage for nutrition and physical activity behaviours (i.e. family and friend support and sabotage) [58], social participation [59], social capital [60] and social cohesion [61] using established measures.

Environmental influences (home and neighbourhood): Participants were asked about their perceptions of their local environment including safety, aesthetics, walking environment [62], and the cost, availability, and convenience of food and food stores. Home availability of fruits, vegetables and high energy foods and beverages and the number of televisions in the house were also assessed.