Correlates of sedentary behaviour among adolescents and adults with hazardous, harmful or dependent drinking in South Africa

Background There is lack of information on the correlates of sedentary behaviour among persons with alcohol use disorders. The study aimed to examine socio-demographic and health correlates among adolescents and adults with hazardous, harmful or probable dependent alcohol use (= problem drinking). Method Data from the cross-sectional South African National Health and Nutrition Examination Survey (SANHANES-1) 2011–12 were analysed. From a total sample of 15 085 persons aged 15 years and older, 2849 adolescents and adults (mean age = 37.1 years, standard deviation [s.d.] = 15.1) were identified as problem drinkers, based on the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C). Multivariable logistic and linear regression were used to determine the associations between socio-demographic characteristics, health variables and high sedentary behaviour (≥8 h/day) and total minutes of sedentary behaviour a day. Results The prevalence of high sedentary behaviour (≥ 8 h/day) was 11.9% overall (11.9% among men and 12.1% among women), and the mean (s.d.) duration of sedentary behaviour was 263 (169) min/day. In bivariate analysis, older age, population group, functional disability, cognitive impairment, having hypertension, having had a stroke and posttraumatic symptoms were correlated with high sedentary behaviour. In adjusted logistic regression analysis, older age and being Indian or Asian were positively, and having been diagnosed with angina was negatively, associated with high sedentary behaviour. In linear regression analysis, older age, not employed and having had a stroke were positively, and being of mixed race and having angina were negatively, associated with total minutes (up to 960 min/day) of sedentary behaviour in a day. Conclusion The study provides socio-demographic and health correlates of sedentary behaviour among problem drinkers. This information can guide possible future interventions in reducing sedentary behaviour among problem drinkers.


Introduction
gender, education, employment status, income, obesity, smoking status, poor physical health and poor mental health (anxiety, depressive symptoms).
Considering the importance of sedentary behaviour interventions, that may differ in terms of correlates from physical activity interventions, 13 there is a special need to target sedentary behaviour in special groups or vulnerable populations such as those who are problem drinkers. 5 Because associations with sedentary behaviour among problem drinkers may differ to those in the general population, 14 it is relevant to study correlates of sedentary behaviour among problem drinkers separately. Therefore, the study aimed to examine sociodemographic and health correlates among adolescents and adults with hazardous, harmful or probable dependent alcohol use (= problem drinking) in a national survey in South Africa.

Sample and procedure
The first South African National Health and Nutrition Examination Survey (SANHANES-1) is a cross-sectional and multistage population-based household health survey conducted in 2012. 18 Participants were interviewed with a questionnaire on socio-demographic and health variables after their informed consent was obtained. The study protocol was approved by the research ethics committee (REC) of the HSRC (REC 6/16/11/11).
The survey response rate of participants was 92.6%.

Measures
Sedentary behaviour was measured with two questions on the time sitting or reclining in the past 7 days during a usual weekday and usual weekend day. 19,20 Responses were recorded in hours and minutes. Sedentary times were truncated at 960 min/day (16 h). Times spent sedentary were summed to calculate the total minutes per day in the last 7 days. 20 High sedentary behaviour was defined as ≥ 8 h/ day. 10,21 Problem drinking was assessed with the three-item Alcohol Use Disorders Identification Test-Consumption (AUDIT-C), with scores of 3 or more in women and 4 or more in men indicating hazardous, harmful or dependent alcohol use (or problem drinking) 22 (Cronbach's alpha 0.89).
Socio-demographic information included sex, age, population group, employment and residential status.
Self-rated health was assessed with the item: 'In general, how would you rate your health today?' 18 Responses were dichotomised into having 'good health' (= 1: very good, or 2: good) or 'poor health' (= 3: moderate, 4: bad or 5: very bad).
Functional disability was assessed with the item: 'Overall in the last 30 days, how much difficulty did you have with work or household activities?' Response options ranged from 1 = none, 2 = mild, 3 = moderate, 4 = severe to 5 = extreme. 18 Functional disability was defined as moderate to extreme difficulties.
Cognitive impairment was assessed with two items: How much difficulty did you have with concentrating or remembering things?
How much difficulty did you have in learning a new task (e.g. learning how to get to a new place, learning a new game, learning a new recipe)? 18 Response options range from 1 = none to 5 = extreme. The two items were summed and cognitive impairment was defined as moderate to severe (scores 6-10).
Current tobacco use was assessed with two questions on daily or less than daily tobacco smoking and use of other tobacco products. 18,23 Fruit consumption: 'How many fruits do you usually eat per day?' Vegetable consumption: 'How many portions of vegetables, excluding potatoes, do you usually eat per day?' Response options were 1 = 4 or more per day, 2 = 1-3 per day, 3 = not every day, but 4 or more a week, 4 = not every day, but less than 4 per week, and 5 = none. 18 Handwashing hygiene behaviour was assessed with the question: 'How often do you wash your hands before eating?' Response options ranged from 1 = always to 4 = never. 18 Bodily pains were measured with the question: 'How much bodily aches or pains did you have?' Response options ranged from 1 = none to 5 = extreme or cannot do. 17 Bodily pains were defined as 'moderate, severe, extreme/can't do'.
Physical ill-health conditions were assessed with the question: 'Has a doctor or nurse or health worker at a clinic or hospital told you that you have had any of the following conditions? High blood pressure, stroke, heart disease, a heart attack or angina (chest pains), high blood cholesterol, high blood sugar or sugar diabetes, and tuberculosis'. 18 Visual difficulties were assessed with two questions asking about the difficulty of near and distant vision. Response options ranged from 1 = none to 5 = extreme or cannot do. 18 Both questions were summed and visual difficulty was defined as 'moderate, severe, extreme/can't do' (scores 6-10).
Hearing difficulties were assessed with two items about the difficulty in 'hearing someone talking on the other side of the room in a normal voice' and 'hearing what is said in a conversation with one other person in a quiet room'. Response options ranged from 1 = none to 5 = extreme or cannot do. 18 Both questions were summed and hearing difficulty was defined as 'moderate, severe, extreme/can't do' (scores 6-10).
Sleep problems were measured with the item: 'How much of a problem did you have with sleeping, such as falling asleep, waking up frequently during the night, or waking up too early in the morning?' Response options ranged from 1 = none to 5 = extreme/cannot do. Sleeping problems were classified in response to this question as 'moderate', 'severe' or 'extreme/cannot do'. 24 Psychological distress in the past month was assessed with the 10-item Kessler 10 scale, 25 and has been validated in South Africa. 26 Response options ranged from 1 = never to 5 = all of the time and the total score is a summation of all the responses, with scores of 20 or more indicating mild, moderate or severe psychological distress. 25 Cronbach's alpha for the Kessler 10 scale was 0.91 in this sample.
Posttraumatic stress disorder (PTSD) was assessed with the 17-item Davidson Trauma Scale (DTS) that assesses all primary DSM-IV symptoms of PTSD related to intrusion, avoidance and hyperarousal symptoms. Participants were considered to have PTSD 'if they score at least one re-experiencing, three avoidance/numbing and two hyperarousal phenomena at a frequency of at least twice in the previous week'. 27 In this analysis, we used any of the three symptom criteria (Cronbach's alpha 0.94).

Data analysis
Data were analysed with STATA software version 13.0 (Stata Corporation, College Station, TX, USA). Descriptive statistics were calculated for the proportions of the study variables. Pearson chi-square statistics was used to test for differences in proportions. We used multivariable logistic and linear regression to determine the associations between sociodemographic characteristics, health variables and high sedentary behaviour (≥ 8 h/day) and total minutes of sedentary behaviour a day. No collinearity was detected. Missing data were excluded from the analysis. All results were adjusted for the multistage sampling design.

Ethical consideration
The study protocol was approved by the research ethics committee (REC) of the HSRC (REC 6/16/11/11).

Sample characteristics
From a total sample of 15 085 persons aged 15 years and older, 2849 adolescents and adults (mean age = 37.1 years, standard deviation [s.d.] = 15.1) had been identified as hazardous, harmful or dependent drinkers (or problem drinkers). In the sample of problem drinkers, the prevalence of high sedentary behaviour (≥ 8 h/day) was 11.9% (11.9% among men and 12.1% among women), and the mean (s.d.) duration of sedentary behaviour was 263 (169) min/day. Compared with individuals without hazardous, harmful or dependent drinking, the prevalence of high sedentary behaviour (13.7%) and the mean duration of sedentary behaviour (267 min/day) did not significantly differ from persons with hazardous, harmful or dependent drinking (11.9%) (p = 0.117) and (263 min/day) (p = 0.223). In the total sample, more men (70.4%) than women (29.6%) were hazardous, harmful or dependent drinkers (p < 0.001), and drinkers were younger than non-drinkers (p < 0.001). Table 1 describes the sample characteristics and prevalence of high sedentary behaviour among problem drinkers.

Associations with sedentary behaviour
In bivariate analyses, older age, population group, functional disability, cognitive impairment, having hypertension, having had a stroke and PTSD symptoms were correlated with high sedentary behaviour (see Table 1).
In adjusted logistic regression analysis, older age and being Indian or Asian were positively, and having been diagnosed with angina was negatively, associated with high sedentary behaviour, while in linear regression analysis older age, not employed and having had a stroke were positively, and being of mixed race and having angina were negatively, associated with total minutes of sedentary behaviour in a day (see Table 2).

Discussion
This is the first study in Africa to investigate correlates of sedentary behaviour among problem drinkers. The investigation found that among problem drinkers (N = 2849, mean age = 37.1 years) the prevalence of high sedentary behaviour was 11.9% and the mean (s.d.) duration of sedentary behaviour was 263 (169) min/day; this did not differ significantly with non-problem drinkers. The sedentary behaviour duration among problem drinkers seems to be higher in this study than in a study among hazardous drinkers in six middle-income countries (with a higher mean age = 45.7 years), including South Africa, with a prevalence of 9% of high sedentary behaviour and a mean duration of sedentary behaviour of 216 min. 7 This study found among problem drinkers that sedentary behaviour increased with age and that there were population group differences: Indian or Asian persons had a higher prevalence of high sedentary behaviour than other population groups, and back Africans had a higher total time of sedentary behaviour than the other, especially mixed race people, population groups. In the general adult population, older age was associated with higher sedentary behaviour. 13,14,15 It is not clear as to why we found such racial or cultural differences, which needs further investigation. In a study among five different ethnic groups in the general population in the Netherlands, no significant ethnic differences were found in objectively measured sedentary behaviour. 28 In our study, not being employed was a strong predictor of increasing minutes per day sedentary behaviour among problem drinkers. A similar correlation was found in a previous study in six middle-income countries, including South Africa. 7 It is possible that having a job increases physical activity and may involve active transportation to go to work. 7 In addition, problem drinking may be more prevalent in the unemployed, can cause unemployment and may reduce chances of getting employed. 7,29 While in the study 7 in six middle-income countries older adults living in urban areas were more likely to engage in sedentary behaviour, this study did not find any urban-rural differences.
In bivariate analyses, functional disability and cognitive impairment were associated with high sedentary behaviour among problem drinkers. In agreement with a previous study, 5 this study also found an association between having had a stroke and spending more time in sedentary behaviour among problem drinkers. Contrary to expectation and previous studies, 30,31 this study found a negative association between angina and sedentary behaviour. It is possible that the individuals with angina have been attending a cardiac rehabilitation programme that includes the promotion of physical activity. A previous review 32 has shown that increasing physical activity can promote healthy ageing (functional ability), possibly also preventing or delaying functional disability and cognitive impairment. In a community survey in Cape Town, South Africa, stroke  survivors were found to engage in a high amount of sedentary time. 33 For people who have had a stroke to engage in physical activity may not be easy, but it would be important to develop possible ways of undertaking physical activity in this population. 7 On the contrary, problem drinking increases the risk of getting a stroke. 5 Our findings seem to show that helping problem drinkers to get employment may be a relevant strategy to decrease sedentary behaviour. 7 An increase in physical activity may also improve the mental and physical health of the problem drinkers. 7 Furthermore, one may need to consider sociodemographic factors such as age and population group and having chronic conditions, such as stroke, in designing interventions in problem drinkers. Strategies to combine interventions to decrease sedentary behaviour with decreasing problem drinking could be effective and should be tested.

Study limitations
The study variables, such as sedentary behaviour, alcohol use and physical illness conditions, were self-reported, and the cross-sectional nature of the study limits our ability to establish causality. Longitudinal studies on sedentary behaviour among problem drinkers are warranted.

Conclusion
This investigation found among problem drinkers that older age, being Indian or Asian, not employed and having had a stroke were positively, and being of mixed race and having been diagnosed with angina were negatively, associated with high sedentary behaviour and/or with total minutes of sedentary behaviour in a day. Findings provide information on possible future interventions that can help to reduce sedentary behaviour among problem drinkers.