METHODS

Participants

Data from women who participated in the baseline assessment for the British Women's Heart and Health Study between May 1999 and July 2000 were used. These comprised women aged 60 to 79, selected from the age-sex register of one group general practice in each of 15 towns in England, Wales, and Scotland. The criteria for selecting the town, the general practice and the participants were the same as those used for the British Regional Heart Study.²² Each woman who attended completed the following: a detailed questionnaire providing information on sociodemographic, lifestyle and health factors; a research nurse-led interview providing more detailed cardiovascular disease information and a drugs history; a physical examination including anthropometric measurements, lung function tests, and ECG. In addition general practitioner notes were reviewed and details of major diseases extracted. Full ethics committee approval was obtained for the study.

Activity levels

Participants were asked to indicate their usual duration of activity in hours per week for several types of activity: walking, cycling, physical exercise (such as fitness classes, aerobics, swimming, jogging, tennis), light and heavy housework, light and heavy gardening and do it yourself (see appendix). In addition they were asked to indicate whether their usual walking pace was slow, steady, brisk, or fast. The questions concerning physical activity were similar to those used in the British Regional Heart Study with the addition in this study of questions on participation in housework, which had not been included in the British Regional Heart Study.^8,23 Each type of activity was defined as moderate or vigorous based on the US Surgeon General's report on physical activity¹² with brisk or fast walking, cycling, heavy gardening, heavy housework, and do it yourself being categorised as moderate and physical exercise categorised as vigorous. Hours spent on light housework and gardening were not included in the measure.

• Download figure
• Open in new tab
• Download powerpoint

APPENDIX: PHYSICAL ACTIVITY QUESTIONS

Recommended levels of activity are 30 minutes of moderate activity on at least five days a week or 20 minutes of vigorous activity three times a week. Women were considered physically active at recommended levels if they engaged in at least 2.5 hours of moderate activity per week or one hour of vigorous activity per week. The small number of women who engaged in both vigorous activity and moderate activity per week were considered active if the total duration exceeded 2.0 hours. In addition, women reported duration of activity throughout the summer and winter. Difference between reported levels of activity in winter and summer were compared with season of completion of baseline assessment to assess possible recall bias.

The effect of different activity types on cardiorespiratory fitness and obesity were assessed by examining the associations between type of activity and resting pulse rate measured as the mean of two readings obtained with the woman seated and relaxed using a Dinamap 1846SX vital signs monitor, and body mass index (BMI, weight in kg/height in metres²). Standing height was measured without shoes using a Harpenden Stadiometer, which recorded to the nearest millimetre. Weight was measured in light clothing to the nearest 0.1 kg using Soenhle portable scales.

Factors associated with activity levels

Participation in specific physical activities according to disease status, smoking, socioeconomic position, and area of residence was examined. Social class was defined by the longest held occupation and classified according to the new government classification.²⁴ Two measures of social class were separately assessed, one based on the women's longest occupation and the other based on her husband's. Region of residence was defined by whether the town of residence was north or south of a line joining Bristol and the Wash, which divides the more prosperous south east from the more deprived north.²²

Statistical analysis

Linear regression was used to assess the association between each activity type and resting pulse after adjustment for the presence of coronary heart disease and age. Multivariate logistic regression was used to assess whether participation in each activity type was associated with reduced odds of being overweight (BMI greater than 25 kg/m²) after simultaneous adjustment for participation in any other activity type, poor health status, presence of coronary heart disease, respiratory disease, arthritis, a recent fall, age, and socioeconomic position. Multivariate logistic regression was used to assess the odds of participating in different activities associated with disease status, socioeconomic position and area of residence, each entered as categorical variables as indicated in table 3. These variables were included simultaneously in the model. In multivariate analysis only participants with complete data for all variables were included. The clustered design of the study was assessed in multivariate analyses but did not change results, so all findings are presented unadjusted for town. All analyses were conducted using Stata version 6.²⁵

RESULTS

Of the 4063 invited 2341 (58%) both completed questionnaires and attended the interview and examination. The age and prevalence of general practitioner recorded serious illness (stroke, coronary heart disease, cancer, diabetes) did not differ significantly between those who participated and those who did not. There was no significant difference between activity levels of participants who completed their questionnaire during the spring or summer months (April to September) compared with those who completed it during the rest of the year.

Association between activity types and cardiorespiratory fitness and obesity.

There were significant trends of decreasing resting pulse and BMI with increasing duration of physical exercise and brisk walking, but no such trends with heavy housework (table 1). After adjustment for age and presence of coronary heart disease physical exercise was associated with a lower heart rate: β (95% confidence intervals) =−0.8 (−1.2 to −0.5) beats per minute per hour of exercise, p<0.01. Brisk walking was weakly associated with a lower heart rate, after adjustment for age and heart disease: β =−0.2 (−0.3 to −0.1) beats per minute per hour of walking, p=0.03 and heavy housework was not associated β =−0.002 (−0.1 to 0.1) p=0.9. After adjustment for other forms of activity and potential confounding factors (table 1) the odds of being overweight were noticeably lower in women who participated in at least 2.5 hours of brisk walking per week and were slightly reduced in those who participated in at least one hour of physical exercise per week but there was no association between heavy housework and being overweight (table 1).

Levels of activity

Table 2 shows the prevalence of being active and of engaging in different types of activity all year round. Activity levels during the summer were only slightly higher than during the winter. Over two thirds of participants were active at recommended levels when domestic activities were included in the assessment. When these activities were excluded only 21% were defined as regularly active. Heavy housework was by far the commonest activity undertaken, followed by physical exercise and brisk walking. Very few participants regularly cycled, did heavy gardening or do it yourself.

Factors associated with participation in different activities

Table 3 summarises the odds of participating in different types of activities for chronic diseases and social factors. Age, self reported poor health status, coronary heart disease, and respiratory disease were associated with reduced odds of participating in all types of physical activity. In addition to these factors, participation in physical exercise was less likely in current smokers, those from socioeconomic group III, and those living in the north. Participation in brisk walking was less likely in those with arthritis, among women who had fallen in the previous 12 months, current smokers, socioeconomic group III, and living in the north. Participation in heavy housework was less likely in women reporting depression, arthritis, and falls, but socioeconomic group and living in the north were not associated with heavy housework.

DISCUSSION

Achievement of recommended levels of physical activity depends on the criteria used; inclusion of domestic activities, in particular heavy housework, increases the proportion of elderly women who are active by over threefold. If domestic activities are excluded, then only 21% of this sample of British women aged 60–79 were regularly active. Heavy housework, gardening and “do it yourself” have similar levels of energy expenditure as other moderately intense activities such as brisk walking and it has been argued that people who spend similar amounts of time on either domestic activities or other forms of moderate activity should achieve equal health benefits.⁹ A strong and graded response between duration of customary physical activity, comprising predominantly domestic activity, and survival over 10 years has been reported among both men and women over the age of 65 years,²⁶ suggesting that domestic activities have important health benefits. However, studies examining the health benefits of specific domestic activities have not been published.

Discrete episodes of aerobic activity improve cardiorespiratory fitness (that is, maximal oxygen uptake, VO²_max) and have a smaller effect on total energy expenditure, whereas more regular episodes of moderate activity primarily increase energy expenditure and therefore prevent obesity if calorie intake is not increased.^9,17 Regular participation in brisk walking and heavy housework would not be expected to have a major effect on cardiorespiratory fitness but as both expend similar levels of energy they should be equally beneficial in terms of preventing obesity.

Resting heart rate—a marker of cardiorespiratory fitness^16,27,28—and BMI were associated with both physical exercise and brisk walking as expected.^9,16,17 These findings provide a validation of the brisk walking and physical exercise components of the physical activity assessment used. No relation was seen between heavy housework and resting heart rate or being overweight suggesting that the energy expenditure involved may be less than that involved in brisk walking or physical exercise. It is possible that either recall of time spent in heavy housework is over-estimated or that women who participate in heavy housework to this level simply consume more calories. Further validation of the predictive value of duration of heavy housework will be provided by prospective follow up of the cohort for survival and incident cardiovascular disease.

Walking and leisure time exercise have been found in this and other studies²⁹ to be associated with higher socioeconomic position and reduced odds of smoking and other adverse life style risk factors. This clustering of behavioural and socioeconomic factors explains some, but not all, of the health benefits of these activities. By contrast, participation in heavy housework is more commonly carried out in lower socioeconomic groups and is not associated with avoidance of smoking. Any potential health benefits from heavy housework may therefore be lost through adverse social circumstances and health behaviours. Wider environmental determinants of participation in physical activity have been reported,^30,31 and it is likely that these may explain both the socioeconomic relations and north-south geographical differences seen. Possible mechanisms include range and quality of local amenities, perceived safety of the environment, and cultural acceptability.

Furthermore, there may be psychological and social pathways through which physical activity improves health. Exercise and activities such as walking are often social experiences, improve psychological wellbeing and social networks, and for some, are an important break from family and work demands.³² Although being “house-proud” may give some psychological benefit from participation in housework this is unlikely to have the same magnitude of effect as other forms of activity.

Poor health, in particular being diagnosed with coronary heart disease, respiratory disease, arthritis, and having a fall in the previous 12 months were associated with being less active. This cross sectional study cannot determine whether poor health results in decreased activity or poorer health is the result of a lifetime of low activity levels. However, a study of older women drawn from a retired occupational cohort in England found that poor health—particular painful joints and lack of energy—was frequently mentioned as a barrier to activity.³³ Work from the US also suggests that health is a determinant of activity levels.³⁴ As patients with coronary heart disease may benefit in the long term from regular activity,³⁵ but increases in activity in the short-term may lead to an increase in symptoms,³⁶ it is essential that policy on physical activity in older people accommodates both those in good and less good health.

Study limitations

The response rate, although moderate, was similar to other recent studies in this age group³⁷ and responders were not significantly different from non-responders in terms of age or general practitioner recorded serious illness suggesting that major selection bias did not occur. Data on the participants usual weekly duration for each activity were collected but it was not possible to examine the intensity with which activities were performed. Duration, rather than intensity, of activity does seem to be of relevance in determining health benefits,²⁶ and recent evidence suggests that the accumulation of shorter episodes of activity (as little as 15 minutes) is beneficial for cardiovascular health.³⁸ Validation of physical activity levels using seven day diary methods, while feasible, may not accurately reflect habitual levels of activity that are of greater relevance to reducing cardiovascular disease risk. Where such validation has been attempted moderate agreements have been found.^39,40

The choice of instrument to measure physical activity was determined by the use of a similar protocol to male participants in the British Regional Heart Study, which shares the same design. Different instruments show moderately strong correlations with each other, so choice of instrument is not a critical issue.⁴¹

Recall bias may have affected the accuracy of reported levels of activity but there is evidence that people are able to recall habitual activities reasonably well, even over long time periods.^42,43 This form of questioning was used in men in the British Regional Heart Study and shows good predictive validity for all cause and coronary heart disease mortality,⁸ suggesting that the approach used provides valid information. In addition there was no seasonal effect on reporting of levels of activity.

In conclusion over two thirds of elderly women in this sample are reaching new recommended levels of physical activity. Heavy housework makes up an important component of activity for this group but may be insufficient in itself to generate health benefits and prospective studies of its health potential are needed. Among older women, social and health factors are important determinants of physical activity levels.