Letters

Social and productive activities in elderly people

Activities have been confused

Activity (occupation) is important for survival

Self rated health is important predictor of mortality

Authors' reply

Activities have been confused

EDITORGlass et al have concluded that in elderly people "social and productive activities that involve little or no enhancement of fitness lower the risk of mortality all cause as much as fitness activities do," and that "activity may confer survival benefits through psychosocial pathways."¹ The study design, however, incorporates a misconception that pervades many studies into physical activity.

Neither high levels of cardiorespiratory fitness nor participation in vigorous activities that promote cardiorespiratory fitness are necessary to decrease morbidity and mortality. The main health related variable is the volume, rather than the frequency, intensity, or duration of the activity. Whether the activity is planned (as in sport), or incidental (as in gardening), is obviously immaterial. The body does not care whether the physical activity is undertaken as sport, exercise, hobbies, translocation, or household chores.

In this study, most activities categorised as social and productive can entail significant levels of incidental physical activity. Of the 11 activities mentioned, only day or overnight trips and playing cards, games, or bingo are likely to be completely sedentary. The risk is high that this analysis is seriously confounded by the presence of health enhancing physical activities in all three categories. Glass et al are obviously aware of this as they have omitted "shopping" and "gardening" from one of the analyses. We cannot, however, discount the possibility that all the activities categorised as productive (gardening, preparing meals, shopping, unpaid work, paid work, and other employment) might incorporate significant amounts of physical activity. Significant risk reduction is achieved at a level of 150 kcal/kg/day of expenditure of energy through activity.² This level is embodied in current health related guidelines for physical activity, ^{3 4} which recommend 30 minutes of brisk walking per day. Sufficient energy expenditure can be just as easily accrued through incidental physical activity as by predetermined sports and exercise activity.

The distribution of total physical activity between the three groups in this study remains uncertain, but the findings must be highly questionable. Future health focused studies incorporating physical activity should appreciate that there are many potential pathways to achieving targets of health related physical activity and should not be distracted by notions of sport, fitness, and fitness training.

Chris Riddoch, senior lecturer, exercise and health science. 
Department of Exercise and Health Sciences, University of Bristol, Bristol BS8 1TN chris.riddoch{at}bristol.ac.uk

Activity (occupation) is important for survival

EDITORGlass et al's paper on the relation between activity and survival in older Americans adds to our understanding of how health affects, and is affected by, engagement in purposeful, meaningful activity or what occupational therapists term occupation. Although the occupational therapy profession was founded on a belief in this relation, only relatively recently has it been examined more critically.

Glass et al's paper contributes to the growing body of research based knowledge that can now help us better understand the relation between occupation and health. A similar project was undertaken by Iwarsson et al, who used existing data from a 25 year longitudinal investigation of elderly people in Sweden.² In that study significant differences in survival were found between women who were more active and less active.

In a more rigorous project with a randomised design Clark et al compared groups of elderly people living in Los Angeles.³ One group followed an activity programme led by occupational therapists, one group followed an activity programme led by non-occupational therapists, and one group served as controls. Interestingly, the authors found no significant differences between the control group and the group who followed the non-occupational therapy programme. The occupational therapy intervention, however, had significant benefits "across various health, function, and quality of life domains."

Of note in the current study is the finding that social and productive activities were just as effective in reducing the risk of death as fitness activities.¹ As the authors suggest, this is probably due to a wide range of mechanisms and provides further evidence of the complex nature of occupations. Indeed, Rudman et al found that elderly Canadians considered activity to be any form of mental, social, or physical types of doing.⁴ The findings also suggested that sense of control was an important mediator of the health benefits of engagement in occupation. For these reasons a full understanding of the relation between occupation and health can be achieved only by considering occupations from many perspectives: transcendental, symbolic-evaluative, sociocultural, information processing, biological, and physical.⁵

It is exciting and encouraging that other disciplines are investigating the occupational nature of people, and I look forward to future research.

Matthew Molineux, lecturer. 
Discipline of Occupational Therapy, College of Ripon and York St John, York YO31 7EX m.molineux{at}ucrysj.ac.uk

Self rated health is important predictor of mortality

EDITORGlass et al add important information to the field of gerontology and enlarge our scope for assessing and caring for older people.¹ They note the possibility that activity levels measured at baseline were actually measuring health status and attempted to examine this in subsequent analyses by eliminating deaths in the first five years of follow up.

They do, however, not deal adequately with the probability that "those who can, do, and those who cannot, don't do." Since individuals self select for the observed social and productive activities, any factors pertinent to self selection, such as wellbeing or self health appreciation, may contribute to the outcome (survival). Several longitudinal community cohort studies observed mortality after a self health evaluation.^2-4 Some followed the cohort well beyond five years, ^{2 3} and some adjusted the self appraisal further with a concurrent medical appraisal.² Almost without exception, global self rated health was found to be an important predictor of mortality. The differences of mortality in elderly people between those with highest and lowest health self evaluations proved to be about 20-35%, equal to or greater than the "protective" reductions in mortality both for social and productive activities and for increased physical activity. Does this perhaps mean that a common underlying factor, not readily measured by surveys (for example, subtle perceptions of energy level, absence of substantial preclinical health problems), may prompt one to evaluate self health positively and also to participate in more social and productive activity or greater physical activity?

The findings of Glass et al should prompt broader research into these neglected areas that may be critically important in longevity. It is, however, too soon fully to endorse a premise that improving social and productive activities in otherwise unmotivated, inactive elderly people would extend life. As practitioners, most of us have supported productive activities of our older patients. Without better confirmation of cause and effect, I am not sure we are yet in a position to press all indolent elderly people to become actively involved as a health measure; such an intervention could be harmful rather than salutary.

Gerson T Lesser, assistant professor. 
Department of Geriatrics and Adult Development, Mount Sinai School of Medicine, Jewish Home and Hospital, New York, NY 10025, US glesser{at}jhha.org

Authors' reply

EDITORRiddoch thinks that the beneficial effects of the activities we studied may partly result from the effects of incidental physical activity. An unknown fraction of the effect of our activities results from these known physiological pathways. In the case of social activities, this fraction is probably rather small. There may, however, be an important residual benefit that results from other, lesser known, psychosocial pathways. The data needed to test this hypothesis fully do not exist. All our evidence from this study and others, however, indicates that more or less sedentary social and productive activities are as protective as physical activities.¹

Moreover, we are aware that many scientists will discount these findings as simply the confounding influence of "incidental physical activity." We do not have the data to rule this out. Psychosocial processes may be involved. For example, the evidence in favour of the influence of the interaction of social networks on mortality is quite strong.² New evidence shows that psychosocial factors have a direct impact on physiological systems.³ The reverse may also be true. In many studies of physical activity and health, the benefits of physical activity are confounded by the psychosocial consequences of what Molineux calls "occupation." This may help to understand the recent findings that "lifestyle" interventions are as effective in various populations as exercise interventions. ^{4 5}

We agree with Lesser that self rated health may be an additional factor impinging on the relation between activity and health. Self rated health may be an effect modifier or on the causal pathway. We believe that the latter is more plausible and decided not to include self rated health as a control variable in this single-equation analysis. We have, however, added that variable to the main models presented in the original paper (with the understanding that this may not be the ideal specification). The regression coefficients change only slightly after controlling for self rated health. The effect of social activities drops from 0.069 to 0.061 (P<0.001); fitness activities from 0.078 to 0.066 (P=0.011); and productive activities from 0.112 to 0.103 (P<0.001). In each case the effect is attenuated slightly, although the overall pattern of influence is the same.

We agree with Lesser that it is too soon to endorse a premise that increasing social and productive activities in otherwise unmotivated, inactive elderly people would extend life. This is not a clinical trial, so definitive conclusions about cause and effect are unwarranted. We thank Molineux for pointing out important earlier work in occupational therapy, and we regret not having relied more on that literature, especially Clark's study, in our manuscript.⁶

Thomas A Glass, assistant professor. 
Department of Health and Social Behavior, Harvard School of Public Health, Boston, MA 02115, USA