With increased life expectancy, the incidence of

Transcription

1 Eriksson Sörman, D., Sundström, A., Rönnlund, M., Adolfsson, R., & Nilsson, L.G. (2014). Leisure activity in old age and risk of dementia: a 15-year prospective study. Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 69(4), , doi: /geronb/gbt056. Advance Access publication June 13, 2013 Leisure Activity in Old Age and Risk of Dementia: A 15-Year Prospective Study Daniel Eriksson Sörman, 1 Anna Sundström, 1,2 Michael Rönnlund, 1 Rolf Adolfsson, 3 and Lars-Göran Nilsson 4 1 Department of Psychology, 2 Centre for Population Studies/Ageing and Living Conditions, and 3 Department of Clinical Sciences, Division of Psychiatry, Umeå University, S Umeå, Sweden. 4 Department of Psychology, Stockholm University and Stockholm Brain Institute, S Stockholm, Sweden. Objectives. The aim of this study was to investigate whether leisure activity is associated with incident dementia in an older sample. Method. We examined a sample of 1,475 elderly ( 65 years) who were dementia free at baseline over a follow-up period of up to 15 years. In addition to analyses involving the total time period, separate analyses of three time periods were performed, 1 5, 6 10, and years, following baseline measurement of leisure activity. Results. After controlling for a variety of potential confounders, analyses of data for the total time period revealed that higher levels of Total activity and Social activity, but not Mental activity, were associated with decreased risk of dementia. However, analyses of the separate time periods showed that this association was only significant in the first time period, 1 5 years after baseline. Discussion. The results from this study provide little support for the hypothesis that frequent engagement in leisure activities among elderly serve to protect against dementia diseases across a longer time frame. The finding of a relationship for the first time period, 1 5 years after baseline, could indicate short-term protective effects but could also reflect reverse causality. Key Words: Cognitive aging Dementia Leisure activities Lifestyle Longitudinal. With increased life expectancy, the incidence of dementia diseases, among which Alzheimer s disease (AD) is the most common (Bird, 2008; Corder et al., 1993), is increasing rapidly. According to estimates by the World Alzheimer Report (Alzheimer s Disease International, 2010), 35.6 million people currently suffer from dementia, and this number will reach 65.7 million by 2030 and escalate to million by the year Thus, it is important to establish factors that slow, or even prevent, this disabling process in the elderly population. Although the cause and prevention of dementia is still not fully understood, several factors appear to increase or decrease the risk of developing the disease. One factor that has been suggested to reduce the risk of AD or dementia is frequent engagement in various leisure activities. Previous studies have indicated that socially stimulating activities (e.g., Fabrigoule et al., 1995; He, Zhang, & Zhang, 2000; Saczynski et al., 2006; Wang, Karp, Winblad, & Fratiglioni, 2002; Wilson et al., 2007) and mentally stimulating activities (e.g., Karp et al., 2006; Scarmeas, Levy, Tang, Manly, & Stern, 2001; Verghese et al., 2003; Wilson et al., 2002a, 2002b) are associated with decreased risk of dementia. The cognitive reserve hypothesis (Scarmeas & Stern, 2003; Stern, 2002) provides a potential account of these findings. The concept of cognitive reserve stems from observations that there is no direct relationship between the degree of brain pathology and the clinical manifestation of that damage. Two related models have been proposed: the passive model and the active model. The passive model, which is also referred as the brain reserve, presumes that individuals differ in their reserve capacity due to individual differences in the structural neural substrate (e.g., overall brain size or synapse counts). Individuals with a larger brain weight are assumed to have a greater brain reserve (due to increased number of synapses or neurons in the cortex), which allows them to compensate for age-associated decline and pathological changes associated with dementia diseases. Neuropathology occurs when a critical threshold of reserve has been reached. The active model of the cognitive reserve focuses on individuals ability to compensate for brain degradation. It suggests that the same degree of brain pathology will have different effect depending on the individual s amount of reserve. According to this model, environmental factors such as more frequent engagement in leisure activities can leave an individual with a larger cognitive reserve, resulting in a more efficient cognitive network. This in turn enables them to compensate for age-related cognitive decline and pathological changes such as dementia by making better use of alternate networks or recruitment of alternative brain networks (Scarmeas & Stern, 2003; Stern, 2002). The Author Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please journals.permissions@oup.com. Received December 21, 2012; Accepted May 14, 2013 Decision Editor: Bob G. Knight, PhD 493

2 494 ERIKSSON SÖRMAN ET AL. In line with the reserve hypothesis, brain imaging and postmortem examinations show that certain individuals exhibit normal cognitive performance even in the presence of quite extensive pathological changes in structure and function (Esiri & Chance, 2012). Further, it has been shown that even in the presence of higher levels of fibrillar brain amyloid associated with dementia, individuals with higher educational attainment function better cognitively than less educated individuals (Kemppainen et al., 2008; Roe et al., 2008). In regard to physical activity, cognitive and neural beneficial effects have similarly been suggested to prevent the onset of dementia (Churchill et al., 2002), and several studies have also found an association between physical activities and dementia or AD status (Larson et al., 2006; Laurin, Verreault, Lindsay, MacPherson, & Rockwood, 2001; Yoshitake et al., 1995). Whereas a recently published review of studies investigating the relationship between engagement in various leisure activities and dementia by Wang, Xu and Pei (2012) confirmed a relationship, in line with the aforementioned studies (see also Fratiglioni, Paillard-Borg, & Winblad, 2004; Stern & Munn, 2010), it should be noted that there are several studies showing a lack of relationship between certain types of leisure activities and risk of dementia (Akbaraly et al., 2009; Wilson et al., 2002a). With the mixed findings in mind, the National Institutes of Health in United States (Williams, Plassman, Burke, Holsinger, & Benjamin, 2010) stated that the evidence are insufficient to draw any firm conclusions with regard to the association between activity engagement and incident dementia and suggested that more well-designed longitudinal studies with long follow-ups are warranted. The differences among prior studies can possibly be explained by differences in methods and study populations. In particular, the association between leisure activities and dementia can be biased by a number of confounding factors. Because participation in leisure activities may be related to a variety of other factors, for example, health-related behaviors and well-being, including depression, it is important to conduct research that controls for these factors. Further, a well-established risk factor for dementia is the presence of apolipoprotein E (APOE) epsilon4 (ɛ4) allele (Corder et al., 1993). Carriers of one setting of ɛ4 have an increased risk of developing AD by three to four times and homozygous carriers have 10 to 12 times increased risk, compared with noncarriers (Farrer et al., 1997). Despite its importance, however, most studies do not control for the influence of the APOE genotype when considering the relationship between leisure activities and dementia (Stern & Munn, 2010). Of further concern, in prior studies that used a relatively short follow-up between measurement of activity level and dementia onset, the significant relationship demonstrated between activity level and dementia possibly signal a causal influence reversed to that suggested before such that lowered activity represent early, preclinical symptoms of the disease. With these considerations in mind, the objective of this present study was to investigate whether participation in leisure activities is related to incident dementia within a large-scale, population-based sample of older adults ( 65 years at baseline). To this end, the association between composite activity indexes (Total activity, Social activity, and Mental activity) and risk of incident dementia was examined. In order to take into account the influence of length of follow-up on the findings and whether reverse causation could confound the results, we further divided followup time into three time periods; 1 5, 6 10, and years. The influences from a large variety of potential confounding factors (e.g., demographic factors, medical and health factors, and APOE genotype) were considered in the analyses. Method Study Population The data were emanated from the Betula prospective cohort study (Nilsson et al. 1997, 2004), an ongoing study of aging, cognition, and health that started in Umeå, Sweden, in The participants in this study were selected from the population registry of Umeå municipality (a city in Northern Sweden with about 100,000 citizens) using stratified (age, sex) random sampling. Data have been collected for five test occasions: (T1), (T2), (T3), (T4), and (T5). On each of these occasions, the participants were invited to two sessions (1 week apart), the first focusing on health assessment and the second on assessment of cognitive functions. In this present study, baseline data from test occasions T2 T4 were considered, as the same version of the activity questionnaire was administered on these occasions. At T2, 433 individuals in Sample 1 (65, 70, 75, 80, 85 years at baseline) took part in the Betula assessment in addition to 397 individuals from Sample 2 (65, 70, 75, or 80 years) and 466 individuals from Sample 3 (65, 70, 75, 80, 85 years). At T3, a further sample, Sample 4 (65, 70, 75, 80, 85, 90 years) was added (n = 254). Finally, at T4, Sample 5 (n = 262; 65, 70, 75, 80, 85, 90, 95 years) was assessed for the first time. Participants The number of participants aged 65 years (65 years is the normal retirement age in Sweden) or older at baseline eligible for inclusion was 1,812 participants. Participants who had not completed the questionnaire were excluded (n = 58) as were those lost at follow-up examination (n = 41; for example, those who had moved outside the catchment area). In addition, data for a few cases with dementia at time of inclusion had to be discarded (n = 19). (Retrospective analyses performed by the geriatric psychiatrist concluded that some participants were incorrectly included in the study due to ongoing process of dementia.) Finally, we excluded participants who had deceased during the first time period (n = 209) or who had been diagnosed as demented after the

3 LEISURE ACTIVITY IN OLD AGE AND RISK OF DEMENTIA 495 last follow-up examination (n = 10). Thus, data for a total of 1,475 participants were included in this present study. The mean baseline age of this group was (standard deviation [SD] = 6.85) and included a total of 639 men and 836 women from Sample 1 (n = 314), 2 (n = 344), 3 (n = 383), 4 (n = 217), and 5 (n = 217). The design of this study is presented in Figure 1. As can be seen, all samples could be used to examine the association between activity and dementia 1 5 years after baseline (first time period). In a second time period, 6 10 years after baseline, data for samples 1 4 were available for this purpose, and finally, participants in samples 1 3 could be examined at a time period years after baseline. The design could be used to examine the association between baseline activity and dementia on the basis the entire set of data (total time period). In addition, the design allowed for analyses of data for the separate time periods. Participants who were diagnosed as demented were never used as healthy controls in any of the time periods (e.g., time from baseline assessment to follow-up), whereas participants who were alive and not having dementia during a time period could serve as controls in up to three analyses. Measures Leisure activity questionnaire. The questions concerned frequency of participation in 16 leisure-time activities during the last 3 months. The questionnaire was sent home to the participants about 1 week before the health assessment. Samples 1, 2, and 3 responded to the questionnaire for the first time at T2, Sample 4 at T3, and Sample 5 at T4. For each activity, participants were requested to indicate frequency of participation on an ordinal scale including the alternatives: never (coded as 0), occasionally (1), a few times a month (2), sometime per week (3), and every day (4). To obtain a meaningful number of observations, new classification was made into never (coded as 0), occasionally or a few times a month (1), and some time per week or every day (2). The following activities were included: (a) Travel/Trips; (b) Sports/Exercise/Walking in the forest; (c) Hunting/ Fishing; (d) Read books; (e) Read magazines; (f) Read newspaper; (g) Watching TV/Listen to the radio; (h) Movies/ Concerts/Theater; (i) Restaurant visits; (j) Spend time with family, relatives, and friends; (h) Attending Courses/ Workshops; (i) Religious assemblies; (j) Association work; (k) Play musical instrument; (l) Needlework; and (m) Other Hobbies/Activities. For the activity Read newspaper, 97.1% reported to do this daily; for Watching TV/Listen to the radio, 98.3% did it every day. Due to the extreme skew, the data for these two questions were discarded from all analyses. To avoid subjective classification of the items into mental, social, and physical indices, a questionnaire that involved a rating of each of the items along these three Figure 1. Design of the analyses over three time periods. Those having dementia within every time period were compared with the nondemented controls.

4 496 ERIKSSON SÖRMAN ET AL. dimensions (on a scale 1 10) was sent to a sample of participants from Sample 5 (mean age = 67 years). In total, 18 people responded. Based on these ratings, activities were categorized as predominantly Mental activities (Read books, Read magazines, Movies/Concerts/ Theater, Play musical instrument, Needlework, Hunting/ Fishing) or predominately Social activities (Travel/Trips, Restaurant visits, Spend time with family, relatives, and friends, Attending Courses/Workshops, Religious assemblies, Association work). The only activity that was rated as predominantly physical was Sports/Exercise/Walking in the forest, and therefore no index to represent a physical dimension was generated. Finally, a total activity index was computed as the sum of scores for the entire set of activities, including both Sports/Exercise/Walking in the forest and Other Hobbies/Activities. Cronbach s alpha for the Social index =.49, for the Mental index =.23 and for the Total index =.59. Dementia. Dementia was diagnosed in accordance with the criteria of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (American Psychiatric Association, 2000). The diagnoses were ascertained based on the prospective longitudinal measures at hand at each test occasion, including a comprehensive analysis of various general and specific health measures, results from neuropsychological testing, and clinical course. The latter is characteristic for each of the dementia disorders. The analyses were performed by a clinical and research geriatric psychiatrist covering the time period , and dementia status between and up to every test occasion was available on an annual basis for all participants included in the analyses. Repeated prospective and retrospective analyses were performed under the study period, and data were blindly assessed in order to increase the validity of the cognitive status. For inclusion in the Betula study, a nondementia status was required. Testing the accuracy of diagnosing a state of progressive dementia, the data of the clinical course over the time period were essential. In 2010, the accuracy of a diagnosis of progressive dementia was determined. All participants in the Betula study, for which data on the clinical course were available, were compared with the cognitive status at inclusion, which was supposed to be a nondementia status. Improperly included participants in the Betula study at all test occasions (T1 T5) were only 19 (0.4%), showing a high validity of the present method of diagnosis. All dementia cases (n = 357) during the follow-up were included in this present investigation. Cases of AD (n = 198), vascular dementia (n = 127), Lewy body dementia (n = 8), frontal lobe dementia (n = 3), Parkinson dementia (n = 8), and unspecified dementia, NUD (n = 13), were all categorized as having dementia (1) in comparison with the nondemented (0). Potential confounders. Potential confounders included in the analyses were age, gender, and years of education. A sum of four self-reported diseases/conditions known to be cardiovascular risk factors was also included: incidence of heart attack/cardiovascular problems (yes/no), stroke (yes/no), high blood pressure (yes/no) and diabetes (yes/ no). Moreover, smoking status (yes currently smoking or used to almost daily/no), alcohol use (yes/no never drank/ no have quit), marital status (married/cohabiting, single, divorced, or widowed), general stress (yes/no), feelings of depression (yes/no), and APOE genotype were considered. All the potential confounders were measured at baseline, that is, the same time the participant answered the questionnaire about leisure activity. APOE genotyping was performed by using polymerase chain reaction (for detailed procedure, see Nilsson et al., 2006). The genotype distribution among the participants, that is, ɛ2/ɛ2 (n = 4), ɛ2/ ɛ3 (n = 136), ɛ2/ɛ4 (n = 25), ɛ3/ɛ3 (n = 605), ɛ3/ɛ4 (n = 232), and ɛ4/ɛ4 (n = 19), were available for 1,021 of the total sample (n = 1,475). The APOE genotype was categorized based on the presence of at least one ɛ4 allele (carriers = 251, and noncarriers = 745). The ɛ2/ɛ4 combination was not included in the analyses because it is suggested that the ɛ2 could be protective, whereas ɛ4 is a risk factor in relation to dementia (Corder et al., 1994). Statistical Analysis We tested differences in baseline characteristic for participants who remained dementia free and those who develop dementia by student s t-test for continuous variables and chi-square test for categorical variables. To examine the association between leisure activity and incident dementia, we used Cox proportional hazards regression models with dementia as dependent variable and with the following independent variables: (a) Total activity, (b) Mental, and (c) Social activities. To analyze the effects of leisure activity over time, we estimated models both for total time period and for the three time periods separately. First, analyses were adjusted for age, gender, and years of education (Model 1); second, diseases, smoking, alcohol use, marital status, general stress, and feelings of depression were added (Model 2); and third, APOE ε4 status (carrier/noncarrier) was also controlled for (model 3). Time to event was calculated as date of entry into the study to the date of dementia diagnosis, death, or date of end follow-up (2010), depending on whichever came first. The results were presented as hazard ratios (HRs) with a 95% confidence interval (CI). All participants included in the analyses had provided information about leisure activities at baseline, but there were a few cases with partially missing data (in total, the rate of missing values was 0.9%). Under the assumption that data were missing at random, the Expectation- Maximization Method (Dempster, Laird, & Rubin, 1977)

5 LEISURE ACTIVITY IN OLD AGE AND RISK OF DEMENTIA 497 was used for imputation of data. None of the imputation generated any changes of means, and further, there was not any substantial difference regarding SD for the variables. (There were some missing responses on questions of education [n = 15] and general stress [n = 10]. Regarding missing on education, missing was replaced by the mean for the age group. Concerning missing on general stressed [yes/no], the participants had rated low values [< 5] on a general stress scale [ranging from 1 10] and was therefore coded as having no general stress [= 0].) Finally, addressing multicollinearity is important when using regressions. In our case, none of the predictor variables were highly correlated. (Education, which perhaps could be assumed to have strong correlations with some select activities, revealed quite modest relationships with the Mental [r =.124, p <.01], Social [r =.156, p <.01], and Total [r =.207, p <.01] indexes. Age was negatively correlated with the Mental [r =.148, p <.01], Social [r =.172, p <.01], and Total [r =.288, p <.01] indexes. Other factors such as Sum of diseases and Feelings of depression were weakly correlated with the indexes.) Results The dementia-free cohort at baseline included 1,475 persons aged 65 years or older. During the follow-up period, 357 incident dementia cases were identified. The mean time from baseline examination to dementia onset was 6.30 (SD = 3.55) years. Out of those who received a diagnosis of dementia, 162 persons developed dementia within the first time period (1 5 years after baseline) and 1,118 were dementia free. Over the second time period (6 10 years), additional 139 persons had developed dementia and 682 individuals were dementia free. Finally, over the third time period (11 15 years), 56 persons developed dementia and 404 remained dementia free. Participant characteristics, as a function of dementia status, are provided in Table 1. Participants who developed incident dementia were more often women and older, compared with those who remained dementia free. Moreover, those who did develop dementia were less often married, used less alcohol, and were more often carrier of the APOE ɛ4 allele. Regarding the percentage of engagement in different leisure activities, Spend time with family, relatives and friends was the most common Social activity, and 74% of the participants were engaged in this, sometime per week or every day. Among the Mental activities, Reading magazines was the most commonly reported activity (56.9%). Many participants were also frequently engaged in Sports/Exercise/ Walking in the forest (55.0%) and Other Hobbies/Activities (41.3%), activities that were not categorized as either Social or Mental. The frequencies and distribution of the individual leisure activities are provided in Table 2. To examine whether leisure activities was related to dementia, separate Cox regression analyses were performed first on the total sample and thereafter on each time period, separately. The results of these analyses are presented in Table 3. Starting with the Cox regressions of the total sample (n = 1475), the first step of the analyses (Model 1) demonstrated that only Total activity (HR = 0.96, 95% CI: , p =.013) was associated with dementia. When including more potential confounders (Model 2), still only the Total activity index was related to dementia (HR = 0.97, 95% CI: , p =.019). When the analyses were further adjusted for APOE (Model 3; n = 996), the Total activity index remained significant (HR = 0.96, 95% CI: , p =.019). In addition, Social activity (HR = 0.94, 95% CI: , p =.048) turned out as significant. Worth mentioning in this regard is that in both the first and the second step of the analyses, Social activity nearly reached significance (Model 1: p =.052; Model 2: p =.070) Turning to the analyses of separate time periods, beginning with the first (1 5 years after baseline), we can see that there were associations between two of the indexes in the first step (Model 1), controlling only for age, gender, and education. For Social activities, the HR was 0.92 (95% CI: , p =.045), and for Total activity, the HR was 0.95 (95% CI: , p =.016), demonstrating a decreased risk of dementia associated with higher frequency 1 5 years before onset. The Mental index was not a significant predictor. When controlling for other potential confounders (Model 2), the analyses revealed that Social activity (HR = 0.90, 95% CI: , p =.014) and Total activity (HR = 0.95, 95% CI: , p =.016) were still significantly associated with dementia. When the analyses were further adjusted for APOE (Model 3, n = 826), Social activity (HR = 0.86, 95% CI: , p =.006) was still significantly associated with dementia together with higher frequency of Total activity (HR = 0.93, 95% CI: , p =.010). For the second time period, none of the indexes were associated with dementia in any model. Also in the third time period, no significant relationships were observed for the three activity indexes. Additional analyses combining the second and third time periods after baseline revealed no significant difference for indexes in any model, thus confirming that the observed association between leisure activity and dementia is present only using the shorter time frame. Finally, to examine the possibility that the activity and dementia association were more pronounced in subgroups such as APOE carriers and AD cases, separate analyses of these groups were performed. However, the results revealed no indication of such a pattern. Discussion The aim of this study was to examine the association between participation in various leisure activities in old

6 498 ERIKSSON SÖRMAN ET AL. Table 1. Characteristics for Participants Analyzed Within Each Time Period 1 5 years 6 10 years years Total sample Characteristic With dementia (n = 162) Without dementia (n = 1,118) With dementia (n = 139) Without dementia (n = 682) With dementia (n = 56) Without dementia (n = 404) With dementia (n = 357) Without dementia (n = 1,118) Age, mean (SD) (5.45) (7.04)** ± ± 6.22** ± ± 5.09* (5.95) (7.04)** Female (%) * * ** Years of education, mean (SD) 8.17 (3.94) 8.02 (3.10) 7.98 (3.10) 7.99 (3.10) 7.67 (2.54) 8.20 (3.17) 8.02 (3.43) 8.02 (3.10) Sum Diseases, mean (SD) 1.19 (0.90) 1.00 (0.89)* 0.89 (0.93) 0.86 (0.82) 0.84 (0.91) 0.75 (0.78) 1.02 (0.92) 1.00 (0.89) Feelings of depression (%) Currently smoking or used to almost daily (%) ** Alcohol (%) Yes ** * ** No, never drank * * ** No, have quit ** General stress (%) Marital Status (%) Married ** ** ** Single Divorced Widowed ** ** ** Apolipoprotein E-ɛ4 carrier (%) ** ** ** Social activity Index, mean (SD) 3.78 (1.83) 4.29 (1.95)** 4.29 (2.08) 4.37 (1.90) 4.13 (1.97) 4.46 (1.90) 4.03 (1.96) 4.29 (1.95)* Mental activity Index, mean (SD) 4.06 (1.99) 4.32 (1.88) 4.30 (1.71) 4.45 (1.84) 4.57 (1.74) 4.58 (1.81) 4.24 (1.85) 4.32 (1.88) Total activity Index, mean (SD) 9.62 (3.82) (3.83)** (3.86) (3.61) (3.32) (3.47) (3.81) (3.83)* Notes. Missing cases only within apolipoprotein E-ɛ4 genotype: 1 5 years = 454, 6 10 years = 202, years = 42, total = 479 missing. *p <.05, **p <.01.

7 LEISURE ACTIVITY IN OLD AGE AND RISK OF DEMENTIA 499 age ( 65 years) and risk of incident dementia. A significant relationship was observed for the Total activity and Social activity indexes but not for Mental activity in the analyses of the entire time period. In the analyses of separate time periods, significant relationships were observed only across the first time period (1 5 years after baseline). Specifically, in this time period, Total activity and Social activity were significant predictors of dementia when controlling for all confounding factors Table 2. Frequency Distribution of Individual Leisure Activities for the Sample (n = 1475) Never Occasionally or few times a month Sometime per week or every day (%) (%) (%) Social activities Travel/Trips Restaurant visits Time with family, relatives, and friends Religious assemblies Attending Courses/ Workshops Association work Mental activities Read magazines Movies/Concerts/Theater Hunting/Fishing Read books Play musical instrument Needlework Other Activities Sports/Exercise/Walking in the forest Other Hobbies/Activities (including age, gender, education, APOE, health, and lifestyle factors) in line with the overall analyses. However, there was no evidence that leisure activities serve to protect against incident dementia over an extended time frame (6 10 years or years) for any of the included indexes. Combining time periods two and three did not change the results. Thus, the results from this study are not in accordance with what would have been expected on the basis of an active model of the cognitive reserve hypothesis. The present demonstration of an relationship between activity level and dementia 1 5 years before dementia onset may indicate that there are short-term effects of engagement in leisure activities on risk of developing dementia, even if should be acknowledged that the effects were quite small (as indicated by the CIs). Alternatively, these effects could reflect reverse causality such that that reduced engagement in activities is an early symptom of dementia, despite a thorough screening for dementia. It is warranted to note that a considerable proportion of the evidence that late-life mental and social activities may protect against dementia was from studies involving a relatively short follow-up period (e.g., Akbaraly et al., 2009; Fabrigoule et al, 1995; Helmer et al., 1999; Wilson et al., 2002a). What still needs to be clarified is whether activities performed earlier in life has a more substantial impact on development of dementia. Given the pathological alterations in the hippocampus and regions nearby are present long before the diagnoses of dementia and AD (Bäckman, Jones, Berger, Laukka, & Small, 2005), activity patterns and related neural networks that are established early in life may be of more importance to maintain cognitive health in old age. Results from animal studies indicate possible Table 3. Associations Between the Activity Indexes and Dementia, Presented in Hazard Ratio (HR) First time period Second time period Third time period Total time period (n = 1,280) (n = 821) (n = 460) (n = 1,475) Activity HR CI (95%) HR CI (95%) HR CI (95%) HR CI (95%) Model 1 Mental activity index Social activity index 0.92* Total activity index 0.95* * Model 2 Mental activity index Social activity index 0.90* Total activity index 0.95* * Model 3 a (n = 826) (n = 619) (n = 418) (n = 996) Mental activity index Social activity index 0.86* * Total activity index 0.93* * Notes. CI = confidence interval. Model 1. Adjusting for age, gender, and education. Model 2. Adjusting for age, gender, education, diseases, smoking, alcohol use, marital status, general stress, and feelings of depression. Model 3. Adjusting for age, gender, education, diseases, smoking, alcohol use, marital status, general stress, feelings of depression, and APOE genotype. a Less participants due to missing data on APOE. *p <.05.

8 500 ERIKSSON SÖRMAN ET AL. support for this, in that environmental enrichment seems to have an positive effects on regions critical to learning and memory, for example, the neocortex and hippocampus (Frick & Benoit, 2010). However, if this can be applied to humans and serve as protection in old age is as yet uncertain, but some studies have revealed a relationship between midlife or early life activities and dementia (e.g., Andel et al., 2008; Carlson et al., 2008; Friedland et al., 2001; Fritsch, Smyth, Debanne, Petot, & Friedland, 2005; Kondo, Niino, & Shido, 1994; Rovio et al., 2005). For example, using prospective activity data, a study of Swedish twin pairs, discordant for dementia, showed greater participation in early and middle life leisure activities to be protective of dementia and AD in later life (Crowe, Andel, Pedersen, Johansson, & Gatz, 2003). Thus, there is some support of the hypothesis that activity level earlier in life may serve to buffer against dementia, findings that should be attended to in future studies. Despite several strengths of this present study, including the long follow-up and control of major confounding factors, including APOE genotype, a few limitations should be acknowledged. First, we were not able to investigate the effects of longitudinal changes in activity over time in relation to dementia due to relatively low number of cases of dementia with repeated measures on the activity questionnaire. Second, we cannot rule out the possibility that more demanding activities, such as board games (Wang et al., 2006) or those targeted by intensive training programs, might have more substantial effects. In particular, the questionnaire did not allow for examining the influence of physical activities, which would have been of interest because there are studies that have demonstrated a relation between physical activity in old age and dementia (see Abbott et al., 2004; Larson et al., 2006). Finally, even though estimates of internal consistency as an index of reliability of activity scales may be biased toward lower values to the extent that frequent engagement in one or several activities partly restrict the time available for other activities, the low estimates observed at present, in particular for mental activity, are of concern. In conclusion, the results of this present study provide little support for the hypothesis that frequent engagement in leisure activities among elderly serves to protect against dementia diseases across a longer time frame. The finding of a relationship using a short follow-up period of 1 5 years after baseline could indicate short-term protective effects but could also reflect reverse causality. Funding The Betula Project is supported by grants from the Swedish Research Council. Correspondence Correspondence should be addressed to Daniel Eriksson Sörman, PhD, Department of Psychology, Umeå University, S Umeå, Sweden. daniel.erikssonsorman@psy.umu.se. References Abbott, R. D., White, L. R., Ross, G. W., Masaki, K. H., Curb, J. D., & Petrovitch, H. (2004). Walking and dementia in physically capable elderly men. Journal of the American Medical Association, 292, doi: /jama Akbaraly, T. N., Portet, F., Fustinoni, S., Dartigues, J. F., Artero, S., Rouaud, O.,... Berr, C. (2009). Leisure activities and the risk of dementia in the elderly: Results from the Three-City Study. Neurology, 73, doi: /wnl.0b013e3181b7849b Alzheimer s Disease International. (2010). World Alzheimer Report 2010 The global economic impact of dementia. London. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC. Andel, R., Crowe, M., Pedersen, N. L., Fratiglioni, L., Johansson, B., & Gatz, M. (2008). Physical exercise at midlife and risk of dementia three decades later: A population-based study of Swedish twins. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 63, doi: /gerona/ Bäckman, L., Jones, S., Berger, A. K., Laukka, E. J., & Small, B. J. (2005). Cognitive impairment in preclinical Alzheimer s disease: A meta-analysis. Neuropsychology, 19, doi: / Bird, T. D. (2008). Genetic aspects of Alzheimer disease. Genetics in Medicine: official journal of the American College of Medical Genetics, 10, doi: /gim.0b013e31816b64dc Carlson, M. C., Helms, M. J., Steffens, D. C., Burke, J. R., Potter, G. G., & Plassman, B. L. (2008). Midlife activity predicts risk of dementia in older male twin pairs. Alzheimer s & Dementia: the Journal of the Alzheimer s Association, 4, doi: /j.jalz Churchill, J. D., Galvez, R., Colcombe, S., Swain, R. A., Kramer, A. F., & Greenough, W. T. (2002). Exercise, experience and the aging brain. Neurobiology of Aging, 23, doi: / S (02) Corder, E. H., Saunders, A. M., Risch, N. J., Strittmatter, W., Schmechel, D., Gaskell, P.,... Pericak-Vance, M. V. (1994). Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nature Genetics, 7, doi: /ng Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D., Gaskell, P., Small, G. (1993). Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer s disease in late onset families. Science (New York), 261, doi: /science Crowe, M., Andel, R., Pedersen, N. L., Johansson, B., & Gatz, M. (2003). Does participation in leisure activities lead to reduced risk of Alzheimer s disease? A prospective study of Swedish twins. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 58, P249 P255. doi: /geronb/58.5.p249 Dempster, A. P., Laird, N. M., & Rubin, D. (1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society, Series B, 39(1), doi: / Esiri, M. M., & Chance, S. A. (2012). Cognitive reserve, cortical plasticity and resistance to Alzheimer s disease. Alzheimer s Research & Therapy, 4, 7. doi: /alzrt105 Fabrigoule, C., Letenneur, L., Dartigues, J. F., Zarrouk, M., Commenges, D., & Barberger-Gateau, P. (1995). Social and leisure activities and risk of dementia: A prospective longitudinal study. Journal of the American Geriatrics Society, 43, Farrer, L. A., Cupples, L. A., Haines, J. L., Hyman, B. T., Kukull, W. A., Mayeux, R.,... van Duijn, C. M. (1997). Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. Journal of the American Medical Association, 278, doi: / jama Fratiglioni, L., Paillard-Borg, S., & Winblad, B. (2004). An active and socially integrated lifestyle in late life might protect against dementia. Lancet Neurology, 3, doi: /s (04)

9 LEISURE ACTIVITY IN OLD AGE AND RISK OF DEMENTIA 501 Frick, K. M., & Benoit, J. D. (2010). Use it or lose it: Environmental enrichment as a means to promote successful cognitive aging. The Scientific World Journal, 10, doi: /tsw Friedland, R. P., Fritsch, T., Smyth, K. A., Koss, E., Lerner, A. J., Chen, C. H.,... Debanne, S. M. (2001). Patients with Alzheimer s disease have reduced activities in midlife compared with healthy control-group members. Proceedings of the National Academy of Sciences of the United States of America, 98, doi: /pnas Fritsch, T., Smyth, K. A., Debanne, S. M., Petot, G. J., & Friedland, R. P. (2005). Participation in novelty-seeking leisure activities and Alzheimer s disease. Journal of Geriatric Psychiatry and Neurology, 18, doi: / He, Y. L., Zhang, X. K., & Zhang, M. Y. (2000). Psychosocial risk factors for Alzheimer s disease. Hong Kong Journal of Psychiatry, 10, 2 7. Helmer, C., Damon, D., Letenneur, L., Fabrigoule, C., Barberger-Gat-eau, P., Lafont, S.,... Dartigues, J. F. (1999). Marital status and risk of Alzheimer s disease: A French population-based cohort study. Neurology, 53, doi: /wnl Karp, A., Paillard-Borg, S., Wang, H. X., Silverstein, M., Winblad, B., & Fratiglioni, L. (2006). Mental, physical and social components in leisure activities equally contribute to decrease dementia risk. Dementia and Geriatric Cognitive Disorders, 21, doi: / Kemppainen, N. M., Aalto, S., Karrasch, M., Någren, K., Savisto, N., Oikonen, V.,... Rinne, J. O. (2008). Cognitive reserve hypothesis: Pittsburgh Compound B and fluorodeoxyglucose positron emission tomography in relation to education in mild Alzheimer s disease. Annals of neurology, 63, doi: /ana Kondo, K., Niino, M., & Shido, K. (1994). A case-control study of Alzheimer s disease in Japan significance of life-styles. Dementia (Basel, Switzerland), 5, doi: / Larson, E. B., Wang, L., Bowen, J. D., McCormick, W. C., Teri, L., Crane, P., & Kukull, W. (2006). Exercise is associated with reduced risk for incident dementia among persons 65 years of age and older. Annals of Internal Medicine, 144, doi: /j x Laurin, D., Verreault, R., Lindsay, J., MacPherson, K., & Rockwood, K. (2001). Physical activity and risk of cognitive impairment and dementia in elderly persons. Archives of Neurology, 58, doi: /archneur Nilsson, L. G., Adolfsson, R., Bäckman, L., Cruts, M., Nyberg, L., Small, B. J., & Van Broeckoven, C. (2006). The influence of APOE status on episodic and semantic memory: Data from a population-based study. Neuropsychology, 20, doi: / Nilsson, L. -G., Adolfsson, R., Bäckman, L., de Frias, C., Molander, B., & Nyberg, L. (2004). Betula: A prospective cohort study on memory, health and aging. Aging, Neuropsychology, and Cognition, 11, doi: / Nilsson, L. -G., Bäckman, L., Nyberg, L., Erngrund, K., Adolfsson, R., Bucht, G.,... Wilblad, B. (1997). The Betula prospective cohort study: Memory, health, and aging. Aging, Neuropsychology and Cognition, 4(1), doi: / Roe, C. M., Mintun, M. A., D Angelo, G., Xiong, C., Grant, E. A., & Morris, J. C. (2008). Alzheimer disease and cognitive reserve: Variation of education effect with carbon 11-labeled Pittsburgh Compound B uptake. Archives of Neurology, 65, doi: /archneur Rovio, S., Kåreholt, I., Helkala, E. L., Viitanen, M., Winblad, B., Tuomilehto, J. (2005). Leisure-time physical activity at midlife and the risk of dementia and Alzheimer s disease. Lancet Neurology, 4, doi: /s (05) Saczynski, J. S., Pfeifer, L. A., Masaki, K., Korf, E. S., Laurin, D., White, L., & Launer, L. J. (2006). The effect of social engagement on incident dementia: The Honolulu-Asia Aging Study. American Journal of Epidemiology, 163, doi: /aje/kwj061 Scarmeas, N., Levy, G., Tang, M. X., Manly, J., & Stern, Y. (2001). Influence of leisure activity on the incidence of Alzheimer s disease. Neurology, 57, doi: /wnl Scarmeas, N., & Stern, Y. (2003). Cognitive reserve and lifestyle. Journal of Clinical and Experimental Neuropsychology, 25, doi: /jcen Stern, C., & Munn, Z. (2010). Cognitive leisure activities and their role in preventing dementia: A systematic review. International Journal of Evidence-based Healthcare, 8, doi: /j x Stern, Y. (2002). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8, doi: /s Verghese, J., Lipton, R. B., Katz, M. J., Hall, C. B., Derby, C. A., Kuslansky, G.,... Buschke, H. (2003). Leisure activities and the risk of dementia in the elderly. The New England Journal of Medicine, 348, doi: /nejmoa Wang, H. X., Karp, A., Winblad, B., & Fratiglioni, L. (2002). Latelife engagement in social and leisure activities is associated with a decreased risk of dementia: A longitudinal study from the Kungsholmen project. American Journal of Epidemiology, 155, doi: /aje/ Wang, H. X., Xu, W., & Pei, J. J. (2012). Leisure activities, cognition and dementia. Biochimica et Biophysica Acta, 1822, doi: /j.bbadis Wang, J. Y., Zhou, D. H., Li, J., Zhang, M., Deng, J., Tang, M.,... Chen, M. (2006). Leisure activity and risk of cognitive impairment: The Chongqing aging study. Neurology, 66, doi: /01. wnl a Williams, J. W., Plassman, B. L., Burke, J., Holsinger, T., & Benjamin, S. (2010). Preventing Alzheimer s disease and cognitive decline. Rockville, MD: Agency for Healthcare Research and Quality. Wilson, R. S., Bennett, D. a, Bienias, J. L., Aggarwal, N. T., Mendes De Leon, C. F., Morris, M. C.,... Evans, D. A. (2002a). Cognitive activity and incident AD in a population-based sample of older persons. Neurology, 59, doi: /01. WNL A1 Wilson, R. S., Krueger, K. R., Arnold, S. E., Schneider, J. A., Kelly, J. F., Barnes, L. L.,... Bennett, D. A. (2007). Loneliness and risk of Alzheimer s disease. Archives of General Psychiatry, 64, doi: /archpsyc Wilson, R. S., Mendes De Leon, C. F., Barnes, L. L., Schneider, J. A., Bienias, J. L., Evans, D. A., & Bennett, D. A. (2002b). Participation in cognitively stimulating activities and risk of incident Alzheimer disease. jama: the journal of the American Medical Association, 287, doi: /jama Yoshitake, T., Kiyohara, Y., Kato, I., Ohmura, T., Iwamoto, H., Nakayama, K.,... Fujishima, M. (1995). Incidence and risk factors of vascular dementia and Alzheimer s disease in a defined elderly Japanese population: The Hisayama Study. Neurology, 45, doi: /wnl