Abstract
Keywords
The world's population is ageing rapidly. In 2001 the number of Australians aged 65 years or over was 2.3 million (12.4% of the total population) – this number is expected to increase to 6 million during the next 50 years (24.2% of the total population) [1]. This dramatic demographic change is a desirable and welcome phenomenon, but the social, financial and health consequences of an ageing society cannot be ignored. Old age is associated with increased vulnerability and susceptibility to disease and disability. Movement disorders, osteoporosis, arthritis, cancer, sensory deficits, cardiovascular diseases, depression and dementia are all highly prevalent in later life. Dementia and depression, for example, are the most frequent mental health disorders of older people and the two leading causes of years of life lost due to disability in Australia [2]. It is clear therefore that the introduction of effective measures capable of preventing depression and dementia in later life would be associated with significant personal, social and financial benefits to Australia and other countries around the world.
For a preventative program to be successful, it needs to be capable of modifying risk factors known to contribute to the pathogenic process that ultimately leads to the manifestation of illness. For example, physical activity has been associated with a decreased risk of depression [3] and cognitive impairment among older adults [4]. It is therefore possible that interventions designed to increase the levels of physical activity may also contribute to decrease the risk of depression and dementia in later life.
We designed the present study with the aim of investigating the association between potentially modifiable lifestyle factors and cognitive abilities/depressive symptoms in community-dwelling women aged 70 years and over.
Methods
Participants
Two-hundred and seventy-eight women aged 70 years or over living in the metropolitan area of Perth, Western Australia, volunteered to participate in the study. They were recruited through advertisements in local newspapers, radio, local community groups and clubs to take part in a study investigating the effects of oestrogen upon mood and memory in elderly women. Participants were excluded from the study if they had a Cambridge Cognitive Examination for the Elderly (CAMCOG) score lower than 80, difficulty with written and spoken English, severe sensory impairment, prior history of strokes, or had used hormone replacement therapy up to 6 months prior to assessment. This study was approved by the Human Research Ethics Committee of the University of Western Australia.
Assessment procedures and instruments
Demographic and basic clinical information was collected from all participants using a semistructured clinical interview, the Cambridge Examination for Mental Disorders in the Elderly (CAMDEX-R) [5]. In addition, mood, quality of life and cognitive functions were assessed with the following rating scales and neuropsychological tests:
Beck Depression Inventory (BDI) [6]. The BDI is a widely used self-rating scale that was designed to evaluate the severity of depression in clinical and research settings. It includes 21 questions with possible ratings ranging 0–3. The scale is particularly useful in the assessment of negative thoughts associated with depression. The BDI has high internal consistency (0.86 or greater) and is sensitive to change in the severity of depression (high scores are associated with increasing severity of depression).
Beck Anxiety Inventory (BAI) [7]. This self-rating scale includes 21 items describing common symptoms of anxiety that can be rated according to their intensity from 0–3. Internal consistency (α = 0.92) and test-retest reliability ratings (r = 0.75) are high, and so are different measures of validity.
Cambridge Cognitive Examination for Mental Disorders of the Elderly (CAMCOG) [5]. The CAMCOG is an instrument of general cognitive assessment divided into several subsections measuring various aspects of cognitive functioning: orientation, language, memory, attention and concentration, praxis, perception, calculation and executive functions. The total score can range from 0–105, and is highly correlated with the Mini-Mental State Examination (MMSE) total score (which can also be computed from the CAMCOG). Reported test-retest reliability scores are greater than 0.8.
Short Form 36 Health Survey (SF-36) [8]. The SF-36 consists of 36 items designed to assess eight different areas of health-related quality of life (physical functioning, physical role, body pain, general health, vitality, social functioning, emotional role and mental health) as well as physical and mental health summary measures. Published internal consistency and test-retest reliability statistics are in excess of 0.70 and validity is also considered to be appropriate.
Assessment of exposures
Physical activity
Current levels of self reported physical activity were assessed using the following question: ‘Do you engage in regular physical activity? If yes, on average how many hours per week?’ This incorporated all forms of physical activity, ranging from light exercise (e.g. housework, walking the dog) to vigorous exercise such as aerobics. According to recommended guidelines, participants were categorized as physically active or physically inactive using the criteria of 30 minutes of exercise on most days of the week. Thus, only subjects who reported three or more cumulative hours of physical activity per week were classified as physically active.
Smoking
Cigarette smoking was assessed with the following questions: (i) ‘Have you ever smoked?’ (Age at which subjects started and/or stopped smoking and the average number of cigarettes smoked per day were also noted); and (ii) ‘Have you ever been a heavy smoker, say 20 cigarettes or more a day for a year or more?’ Subjects who reported having ever smoked more than 20 cigarettes per day were classified for the purposes of this study as ‘heavy smokers’.
Alcohol consumption
The assessment of past alcohol consumption was determined using the following question: “Have you ever had an alcoholic drink and if so what was the most you ever drank on a regular basis?' The average number of alcoholic units consumed over a typical two-week period was recorded. Harmful use of alcohol was defined as a daily consumption of three or more units of alcohol over a period equaling or greater than two weeks. A unit was defined as 10 fluid ounces of beer/lager, one glass of wine or one standard measure of spirits or liqueurs. This definition was based on the National Health and Medical Research Council guidelines for risk consumption of alcohol for women [9].
Body mass index
Weight was measured to the nearest kilogram using a set of scales. Height was measured to the nearest 0.5 cm using a wall-mounted stadiometer. Body mass index (BMI) was calculated as the weight in kilograms divided by height in centimetres squared (kg/cm2). Participants with a BMI of 25–29.9 were considered overweight. A BMI of 30 or above was considered indicative of obesity.
Data analysis
Data was analysed using the SPSS statistical package (V. 11). Descriptive statistics were used to determine frequencies and means. Mann–Whitney U-tests were used for between-group comparison of ordinal data and Student's t-test for the between group comparison of parametric variables. Analysis of contingency tables using Pearson's χ2 was used to investigate the association between categorical variables (such as physical activity and marital status). Spearman's correlation coefficient rho was used to determine the association between two ordinal variables, such as scores on the BDI and BAI. Logistic regression was then calculated to determine the association between independent variables of mood and memory. Ninety-five percent confidence intervals were also calculated. Alpha-value was set at 0.05 and all tests were two tailed.
Results
Two-hundred and seventy-eight women volunteered to take part in this study. Eight subjects had to be excluded because they met one or more exclusion criteria (two had CAMCOG < 80 and six reported prior history of stroke). As a result, data on 270 women were available for analysis. Their mean age was 74.6 years (SD = 4.3, range = 70–92 years) and 40% were married at the time of assessment. One hundred and forty-four (53.3%) subjects reported having had more than minimum statutory education.
One hundred and sixty-nine women (62.6%) reported being physically active for at least three hours per week. Forty-six (17.0%) described having been heavy smokers in the past (i.e. at least 20 cigarettes per day), although only ten women (3.7%) were still smoking at the time of assessment. As such, there was inadequate power to further investigate current smoking. Ninety-two women (34.1%) also informed us that they consumed alcohol and eight (3.0%) of them were consuming alcohol at a potentially risky level. Information on height and weight was available for 208/270 women: 25.5% of them were obese.
Tables 1 and 2 display relevant clinical and cognitive information for women according to their level of physical activity, heavy smoking, alcohol consumption and body weight. Women who reported being physically active were found to be significantly less likely to be depressed or anxious and to have higher health-related quality of life. Women who had ever been heavy smokers were less likely to be married, were more depressed and had lower health-related quality of life than women who had not smoked heavily. There was no significant difference between subjects who had ever smoked and never smoked in relation to their age (t = 0.28, p = 0.779), BDI (t = −1.30, p = 0.193) and SF-36 scores (t = 0.96, p = 0.337) (nb: this information is not included in the table). Women who used alcohol moderately were found to have higher cognitive scores in comparison to women who did not use alcohol at all. Obese women had lower health-related quality of life as compared to women of normal weight. There was no association between B12 or folate deficiency and health-related quality of life, mood or memory.
Clinical characteristics of older women according to lifestyle factors (activity, smoking and alcohol use)
Clinical characteristics of older women according to lifestyle factors (obesity and B12 or folate defficiency)
Table 3 displays clinical characteristics for older women according to the presence of clinically significant depression, anxiety and cognitive abilities. As would be expected, younger women and women who have higher levels of education, displayed higher cognitive scores. Women with higher health-related quality of life ratings were shown to have better memory abilities and women who used moderate levels of alcohol also scored within the upper 50% of CAMCOG results. There were significant associations between depression and being physically inactive, between depression and a history of heavy smoking, and between depression and having a lower health-related quality of life. Women who displayed clinically significant levels of anxiety had lower health-related quality of life ratings and were less physically active than nonanxious women.
Clinical characteristics of older women according to the presence of clinically significant depression, anxiety and cognitive abilities
BDI and BAI scores were inversely correlated with SF-36 scores (Spearman's rho = −0.66 and −0.60, respectively, p < 0.001 for both analyses). BDI and BAI scores were directly correlated (Spearman's rho = 0.52, p < 0.001).
Variables that were associated with depression, anxiety and CAMCOG fifth percentile scores were included in three separate logistic regression analyses (using 〈-value of up to 10%). The results showed that physical activity halved the risk of depression (OR = 0.5, 95% CI = 0.3–0.8), whereas ever having regularly smoked more than 20 cigarettes per day increased the risk of depression (OR = 2.8, 95% CI = 1.4–5.5). Subjects who were physically active were also less likely to experience clinically significant anxiety symptoms (OR = 0.5, 95% CI = 0.3–0.8). Finally, higher education attainment (OR = 2.0, 95% CI = 1.2–3.3) and non-risky alcohol use (OR = 2.0, 95% CI = 1.1–3.5) were associated with higher CAMCOG scores, whereas increasing age was associated with lower CAMCOG fifth percentile (OR = 0.9, 95% CI = 0.8–0.9).
Discussion
Physical activity
The results of this study show that even in later life, a greater level of physical activity is associated with better mood, reduced anxiety and better quality of life. Women who were physically active were found to be 50% less likely to be depressed than physically inactive women. These findings are consistent with the results of previously reported studies showing significant associations between low levels of physical activity and depression in older women [10–12]. Interestingly, Kritz-Silverstein et al. [10] compared the findings from a cross-sectional study with a prospective study and found that while exercise was associated with lower depression scores in the cross-sectional study, introduction of an exercise regime for women diagnosed with depression did not appear to reduce depression scores. Conversely, Blumenthal et al. [13] reported in a 16-week randomised controlled trial of physical activity for people with major depression that activity was as efficacious as antidepressant medication. Another study by Singh et al. [14] suggested that previously conflicting evidence may have been partly due to different methodologies in ascertaining levels of ‘physical exercise’. For example, our study has included housework and leisure pursuits (ranging from golf to walking the dog to tennis) in addition to strenuous activities (e.g. aerobic exercise) whereas others have only included strenuous methods of physical exercise [10]. As outlined by Simonsick [15], future research should aim at obtaining greater consistency of measurement.
The results of our study further indicate that sedentary lifestyle is associated with increased anxiety. As far as we are aware, this is the first time physical activity, or lack of it, has been clearly associated with clinically significant anxiety. However, this is not surprising, as anxiety is closely correlated with depression in later life (post hoc Pearson r = 0.53, p < 0.001), particularly among women [16].
We also observed the existence of an association between greater levels of physical activity and improved health-related quality of life. These findings are compatible with those of several previous studies [11], [17], [18]. In addition, there is evidence from a randomised trial that exercise training improves health-related quality of life of patients with intermittent claudication [19].
The mechanisms underlying the association between physical activity and depression/anxiety may be attributed to several factors. First, the increased health benefits of being physically active decreases the risk of cardiorespiratory diseases, diabetes, obesity, osteoporosis and cerebrovascular events, which have been shown to be associated with an increased risk of depression and lower health-related quality of life [20–22]. Second, physical activity has been shown to decrease the physiological response to stress [23]. Third, other biological mechanisms such as an increase in serotonin and endorphins and alterations to central norepinephrine activity, may explain why exercise is associated with better mood and quality of life [23–25]. Finally, psychological mechanisms are also positive outcomes of physical activity, which promotes increased feelings of self-efficacy and mastery and, in turn, better quality of life [26], [27].
Findings from the current study indicate that there is no association between physical activity and cognition – these results are consistent with the findings of a previously published prospective Australian study [28]. However, several other studies have reported that women who are physically active have better scores on tests of cognition [28–31]. Yaffe et al. [29] and Pignatti [31] have both shown that women with higher levels of baseline physical activity, and with an absence of dementia, are less likely to experience cognitive decline at follow-up. In addition, Barnes et al. [30] showed that physical fitness, as measured by a treadmill test, was positively associated with the preservation of cognitive function. These latter findings seem robust in that an extensive neuropsychological battery was used to measure cognitive function, and physical fitness was measured in a standardized way and did not rely upon self-report measures, such as ours.
Body mass index
In our study, women who were overweight on BMI had significantly lower scores on the SF-36 than women of normal weight. The mean BMI (27.43, SD = 5.23) for the sample was within the overweight range, although not substantially different from population norms. Cameron et al. [32] reported that 36.4% of Australian women aged 75 years or over were overweight and 15.6% were obese. Our study found that 39.3% were overweight and 25.5% obese. Previous research had described a relationship between obesity and depression and between obesity and poor health-related quality of life [18]. The consequences of being overweight are likely to affect all aspects of a person's life, including psychological, physical and social factors. For example, overweight individuals run an increased risk of cardiovascular disease and diabetes, they are likely to experience more bodily pain and arthritis, dissatisfaction with their appearance and perhaps be limited in their capacity to do the things they want to do, in comparison with nonobese individuals. Although consistent with the current study findings, BMI may not be a useful indicator of health risk in people who are very short because of deficiencies of the calculation of BMI, or in older people, given that muscle mass decreases with age. Unfortunately, waist circumference, which does not have the same limitation, was not measured in this study.
Alcohol
We found that moderate alcohol consumption was associated with better cognitive performance, but had no obvious association with mood or quality of life. These findings are in line with those published by the Bordeaux Study in France [33], the Rotterdam Study [34] and by Mukhmal et al. [35]: they all indicate that moderate alcohol intake is associated with a decreased risk of dementia. However, a protective effect of light to moderate alcohol use on cognition has not been a universal finding [28],[36–39]. Light to moderate alcohol consumption has previously been shown to decrease the risk of ischaemic stroke, heart disease and mortality [40–42]. Moderate alcohol consumption may also promote the release of acetylcholine in the hippocampus, which in turn is associated with improved memory and learning [43]. The presence of antioxidants, particularly in red wine, has also been postulated as a possible mechanism to explain the decreased risk of dementia among consumers in observational studies. However, it is important to bear in mind that alcohol consumption may be associated with significant clinical complications, including cognitive decline, so that its possible prescription as part of a ‘preventative strategy’ would need to take into account its narrow therapeutic window.
Smoking
In the current study, women who had ever smoked more than 20 cigarettes per day were nearly three times as likely to demonstrate symptoms of depression and were more likely to have poorer quality of life. Smoking has previously been shown to be associated with an increased prevalence of depression in early adulthood [44], higher comorbidity levels of poor mental health [44–46] and poorer health-related quality of life [47], although little data are available for the elderly population. Research shows that depression and anxiety are common among smokers, but the direction of causality is still contentious [48]. It is known that people may use smoking as a form of self-medication; to increase pleasure and as a way of coping with stress and anxiety [49]. Neuropharmalogical effects of nicotine on neurotransmitters could also be associated with improved mood [50].
There were no significant associations between smoking and cognition. We have previously demonstrated that the reported positive association between smoking and cognition found in some case-control studies was contradicted by the negative findings from cohort studies and that case controls studies are influenced by healthy survivorship bias [51].
B12 and folate
Low levels of folate and B12 are associated with high levels of homocysteine, a toxic metabolite that has been linked to an increased risk of cardiovascular and cerebrovascular events [52], with substantial morbidity and mortality. We have previously reported that high plasma homocysteine is associated with increasing levels of depression in later life and that such an association was independent of folate and B12 levels [53]. More recently high plasma homocysteine has been linked to dementia in both case-control studies [54] and cohort studies [55]. The number of people in our sample with vitamin deficiency was small, so that the study was clearly underpowered to detect an association between B12 and folate deficiency and the mental health measure of interest.
Conclusion
Some characteristics of this study should be taken into account when interpreting its results. The cross-sectional nature of the study does not allow for the direction of causality to be determined. One could argue, for example, that people with depression and lower cognitive attainment are more prone to physical inactivity. While such an interpretation is theoretically possible, the results of emerging cohort studies and some randomised trials suggest that depression and cognitive attainment are more likely to be the result rather than the cause of poor physical activity [5], [14].
The study may also have had limited power to investigate some of the associations, particularly between vitamin deficiencies and mood/cognition. In addition, our sample consisted of a select group of female volunteers who responded to an advertisement for a study investigating the effects of hormone on mood and memory. The women were predominantly of Caucasian background and were high functioning, educated and lived in their own homes. The exclusion criteria, namely use of HRT in the previous 6 months, strokes and cognitive impairment may have also limited the representativeness of the sample with regard to the national population of Australian women aged 70 years and above and, as a consequence, may have reduced the range of cognitive scores, depression and quality of life relative to those that are found in the general population. Of note, however, the observed rates of physical activity are comparable to previously published results of representative community surveys [56]. Furthermore, it is unclear whether the associations observed in the present study can be also extended to older men. Finally, the retrospective and self-reported measurement methods for many of the variables (physical activity, alcohol consumption and smoking) may also limit the generalisability of the findings.
In conclusion, the results of our study indicate that older women who are physically active have better mood ratings, whereas those who were ever heavy smokers are more likely to present with clinically significant depressive symptoms. Women who received more than minimum statutory education and use alcohol at nonrisky levels have better cognitive performance. Lifestyle factors should be considered as potential targets for interventions designed to reduce the incidence and prevalence of depression and anxiety in later life. Ideally, the first step toward achieving this goal is the design of randomised trials to determine whether the management of lifestyle factors in older people reduces the incidence of mood disorders and cognitive impairment. The other issue that will need to be addressed is the ‘timing’ of the intervention and its potential preventative effect, that is are interventions in early adulthood more effective than in later life?
Footnotes
Acknowledgements
This study was supported by a project grant from the National Health and Medical Research Council of Australia (Project 211976).