Leisure-Time Physical Activity Is Associated With Reduced Risk of Dementia-Related Mortality in Adults With and Without Psychological Distress: The Cohort of Norway

My main leisure time activity was whitewater canoeing. My doctors did nothing to get me back to that activity so I was under lots of psychological distress. Your doctors are supposed to do no harm, mine left me with continuing disability and a higher risk for dementia.

Original Research ARTICLE

Front. Aging Neurosci., 25 May 2018 | https://doi.org/10.3389/fnagi.2018.00151

 

Ekaterina Zotcheva^1*, Geir Selbæk^2,3,4, Espen Bjertness⁴, Linda Ernstsen¹ and Bjørn H. Strand^2,4,5

• ¹Department of Public Health and Nursing, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology, Trondheim, Norway
• ²Norwegian National Advisory Unit on Ageing and Health, Vestfold Hospital Trust, Tønsberg, Norway
• ³Center for Old Age Psychiatric Research, Innlandet Hospital Trust, Ottestad, Norway
• ⁴Faculty of Medicine, University of Oslo, Oslo, Norway
• ⁵Department of Chronic Diseases and Ageing, Norwegian Institute of Public Health, Oslo, Norway

Background: Leisure-time physical activity (PA) has been proposed as a protective factor against dementia, whereas psychological distress is associated with an increased risk of dementia. We investigated the associations of leisure-time PA and psychological distress with dementia-related mortality, and whether the association between leisure-time PA and dementia-related mortality differs according to level of psychological distress.

Methods: 36,945 individuals from the Cohort of Norway aged 50-74 years at baseline (1994–2002) were included and followed up until January 1st 2015. Leisure-time PA and psychological distress were assessed through questionnaires, whereas dementia-related mortality was obtained through the Norwegian Cause of Death Registry. Adjusted Cox regression analyses were used to estimate hazard ratios (HR) and 95% confidence intervals (95%CI).

Results: Compared to inactivity, leisure-time PA was associated with a decreased risk of dementia-related mortality; low intensity leisure-time PA (HR = 0.73, 95% CI 0.59–0.89); high intensity leisure-time PA (HR = 0.61, 95%CI 0.49-0.77). A statistically significant difference in dementia-related mortality risk was observed between low and high intensity leisure-time PA (p < 0.05). Psychological distress was associated with an increased risk of dementia-related mortality (HR = 1.45, 95% CI 1.16–1.81). Among non-distressed, leisure-time PA was associated with a decreased dementia-related mortality risk; low intensity leisure-time PA (HR = 0.77, 95% CI 0.61–0.97); high intensity leisure-time PA (HR = 0.65, 95% CI 0.51–0.84). The same applied for those with psychological distress; low intensity leisure-time PA (HR = 0.57, 95% CI 0.35–0.94); high intensity leisure-time PA (HR = 0.42, 95% CI 0.22–0.82). The interaction between leisure-time PA and psychological distress on dementia-related mortality was not statistically significant (p = 0.38).

Conclusions: Participating in leisure-time PA was associated with a reduced risk of dementia-related mortality, whereas psychological distress was associated with an increased risk of dementia-related mortality. Leisure-time PA appears to be equally strongly related with dementia-related mortality among those with and without psychological distress, underlining the importance of leisure-time PA for various groups of middle-aged and older adults.

Introduction

Physical activity (PA) is widely considered one of the key lifestyle factors associated with reduced risk of mortality (Arem et al., 2015) and a number of non-communicable diseases (Reiner et al., 2013). Furthermore, a large body of research indicates that participating in regular PA may reduce the risk of cognitive decline (Blondell et al., 2014) and dementia (Rosness et al., 2014; Guure et al., 2017). A recently published meta-analysis of prospective studies demonstrated a dose-response relationship between volume and intensity of PA and risk of dementia (Xu et al., 2017). For every 10 metabolic equivalent of task hours (MET-h) increase per week, there was a 10% decrease in the risk of all-cause dementia (Xu et al., 2017). As populations across the globe age, the prevalence of dementia is predicted to increase substantially, from approximately 47 million in 2015, to 132 million by 2050 (Prince et al., 2015). Hence, measures to prevent or delay new cases of dementia are required, and research on protective and risk factors for dementia has intensified during the last decade (Livingston et al., 2017).

Both early- and late-life depression has been linked to an increased risk of dementia (Byers and Yaffe, 2011). Likewise, psychological distress, characterized by general symptoms of anxiety and depression, has been associated with a higher risk of dementia (Skogen et al., 2015) and dementia-related mortality (Rosness et al., 2016). However, PA has repeatedly been associated with reduced symptoms of depression and anxiety in non-clinical populations (Rebar et al., 2015). A recent randomized exercise trial revealed that increases in moderate-to-vigorous PA predicted reductions in psychological distress in older adults (Awick et al., 2017), and a meta-analysis of randomized controlled studies showed that exercise has a significant antidepressant effect among individuals with depression (Schuch et al., 2016). Thus, existing research indicates that PA is beneficial for both mental and cognitive health.

Despite the aforementioned associations between PA, psychological distress, and dementia, little is known on how PA is associated with dementia risk among individuals with psychological distress. Khatri et al. (2001) randomized clinically depressed older adults to either an anti-depressive medication group, an aerobic exercise group, or a combined exercise and medication group. After 4 months, the researchers found that aerobic exercise improved cognitive function among depressed older adults, and that these improvements corresponded to decreases in depressive symptoms (Khatri et al., 2001). However, the latter study did not investigate risk of incident dementia. The aim of the present study was to investigate the associations of leisure-time PA and psychological distress with dementia-related mortality in middle-aged and older adults, and to examine whether the association between leisure-time PA and dementia-related mortality differs according to level of psychological distress.

Methods

Study Population

The present study is part of the Gene-Environment Interaction in Dementia (GENIDEM) project, which aims to investigate environmental and genetic factors for dementia. The study population comprised 36,945 individuals free from symptoms or diagnosis of heart disease from the Cohort of Norway (CONOR) aged 50–74 years at baseline (1994–2002) linked with the Norwegian Cause of Death Registry and the National Education Database by means of a unique personal identification number. CONOR is a multipurpose study set up to study aetiological factors for a large variety of diseases, and includes 10 epidemiological cohorts from different geographical areas in Norway (Naess et al., 2008). In the present study, three epidemiological cohorts were excluded, as they did not include the CONOR Mental Health Index (CONOR-MHI) scale used to assess psychological distress.

Baseline data collection in CONOR was carried out from 1994 to 2002, following a standardized procedure, where participants received a letter of invitation containing an information brochure by mail 2 weeks prior to a health examination. The health examination included a physical checkup including measurements of participants' height, weight, and blood pressure, as well as drawing blood samples. After completing the examination, the participants returned a self-report questionnaire assessing health- and lifestyle-related variables by mail (Naess et al., 2008). Study participants were followed from baseline until death, emigration, or January 1st 2015, whichever occurred first, with a maximum follow-up time of 20.4 years (mean 15.3). Mean age at follow-up was 76.5 years (max 94.1 years).

All participants included in CONOR gave their written informed consent. The participants' names and personal ID numbers were omitted before data were made available for research purposes. The study was approved by the Norwegian Data Inspectorate and the Regional Committees for Medical Research Ethics, and was conducted in accordance with the Declaration of Helsinki.

Physical Activity

Participants were asked to define the intensity and duration of their leisure-time PA in an average week during the past year. The leisure-time PA question was divided into intensity categories of “low intensity” (not causing perspiration or panting) and “high intensity” (causing perspiration and/or panting) leisure-time PA, each with four alternatives related to average hours per week: “none,” “ <1 h,” “1–2 h,” and “≥3 h.” To ensure sufficient statistical power, all individuals participating in any low intensity leisure-time PA were placed in one group and all individuals participating in any high intensity leisure-time PA were placed in one group, irrespective of hours of weekly leisure-time PA. Participants who replied “none” to both categories of leisure-time PA were considered inactive. Participants who replied “none” on one category of but did not provide an answer to the other category of leisure-time PA, or who had missing answers on both categories were excluded. In the present study, participants were categorized into the following three categories of leisure-time PA based on intensity: “inactive,” “low intensity leisure-time PA,” and “high intensity leisure-time PA.”

Psychological Distress

Psychological distress was assessed with the CONOR Mental Health Index (CONOR-MHI). CONOR-MHI consists of seven items assessing general symptoms of depression and anxiety, and is based on a modification of the General Health Questionnaire (GHQ) and the Hopkins Symptom Check List (HSCL-10) (Søgaard et al., 2003). The items on the CONOR-MHI are shown below. Each item has four answer categories: “no,” “a little,” “a good amount,” and “very much,” and are given values 1–4. The CONOR-MHI score is calculated by dividing the total score (range 7–28) on all seven items by seven, resulting in a range of 1–4, where 1 represents low level of psychological distress, and 4 represents high level of psychological distress. In records containing one missing value, the value was replaced with the sample mean value for each item, whereas records with two or more missing items were excluded.

CONOR Mental Health Index

Have you, in the course of the last two weeks, felt:

Nervous and unsettled?

Troubled by anxiety?

Secure and calm (inverse score)?

Irritable?

Happy and optimistic (inverse score)?

Sad/depressed?

Lonely?

A study comparing the CONOR-MHI to the HSCL-10 and the Hospital Anxiety and Depression Scale (HADS) showed a strong correlation with both scales, r = 0.70 and r = 0.76, respectively (Søgaard et al., 2003). The same study showed that a cut-off at ≥2.15 on the CONOR-MHI has a sensitivity and specificity of, respectively, 41 and 98% for caseness of HADS-anxiety, 38 and 96% for HADS-depression, and 66 and 95% for HSCL-10 (Søgaard et al., 2003). Based on the cut-off of ≥2.15 on the CONOR-MHI, participants in the present study were categorized into two groups: “no psychological distress” or “psychological distress.”

Dementia-Related Mortality

Dementia-related mortality was used as a proxy for dementia illness, and was obtained from death certificates from the Norwegian Cause of Death Registry (Rosness et al., 2014). Dementia was identified when it served either as the underlying, immediate, or accompanying cause of death, according to the International Statistics Classification of Diseases and Related Health Problems, 10th revision, codes F00-F03 and G30.0-G30.9. In the present study population, a total of 919 dementia-related deaths were registered during the follow-up period. For validity purposes, we re-ran analyses with cases restricted to those with dementia as underlying cause.

Covariates

The following variables that could possibly affect the association between leisure-time PA, psychological distress, and dementia-related mortality were identified based on prior studies: sex, education, diabetes, smoking, body mass index (BMI; weight in kilograms divided by height in meters squared), and hypertension. Demographic and health-related covariates were obtained from the results of the physical checkup and the self-report questionnaire at baseline. Attained educational level was obtained by coupling the participants' unique identification number to the National Education Database. The education variable was then divided into three groups: “high” (university degree/college, corresponding to 13 or more years of schooling), “medium” (secondary qualifications, corresponding to 10 years of schooling), and “low” (elementary school, corresponding to 7 years of schooling) (Strand et al., 2014). Participants who reported currently or previously suffering from diabetes were categorized as diabetic. Smoking habits were dichotomized into daily or non-daily smoker. BMI was divided into four categories: underweight (BMI < 18.5), normal weight (BMI 18.5–24.9), overweight (BMI 25–29.9), and obesity (BMI ≥30). Participants were categorized as hypertensive at a systolic pressure of ≥160 mm Hg and/or a diastolic pressure of ≥100 mm Hg, according to the National Institutes of Health guidelines (National Institutes of Health, 2015).

Statistical Analyses

Adjusted Cox proportional hazards regression was used to estimate hazard ratios (HR) and 95% confidence intervals (95% CI) for the associations of leisure-time PA and psychological distress with dementia-related mortality. Attained age was used as the time variable in the regression models, and thereby all the models were finely adjusted by age. Emigration or non-dementia-related mortality were censored. For leisure-time PA, the “inactive” group served as the reference group, whereas for psychological distress, “no psychological distress” served as the reference group. Leisure-time PA served as the main exposure in the regression models. In the first regression model (model 1), the analyses were adjusted for sex. Next, education, smoking, and psychological distress were added to the model (model 2). The final model (model 3) included the covariates from model 1 and 2, in addition to diabetes, BMI, and hypertension.

To investigate whether psychological distress modified the association between leisure-time PA and dementia-related mortality, an interaction term between leisure-time PA and psychological distress was added to the regression models. In addition, relative excess risk due to interaction (RERI) estimates were calculated to investigate possible additive interaction between leisure-time PA and psychological distress. To obtain psychological distress-specific associations between leisure-time PA and dementia-related mortality, the regression models 1, 2, and 3 were stratified by psychological distress. In these analyses, psychological distress was removed as a covariate from the regression models. Stata version 14 was used for all statistical analyses.

Results

Baseline characteristics of the study sample by leisure-time PA group are presented in Table 1. Participants in the leisure-time PA groups had higher education, less psychological distress, hypertension, and diabetes, and were less likely to smoke compared to inactive participants. The high intensity leisure-time PA group consisted of less women than the low intensity leisure-time PA group and the inactive group.

Leisure-time PA was associated with a decreased risk of dementia-related mortality when compared to inactivity; low intensity leisure-time PA (HR = 0.68, 95% CI 0.56–0.84); high intensity leisure-time PA (HR = 0.54, 95% CI 0.43–0.68) in a model adjusted by sex (Table 2, Model 1). The results were slightly attenuated, but remained statistically significant after additional adjustment for education, smoking and psychological distress; low intensity leisure-time PA (HR = 0.76, 95% CI 0.62–0.93); high intensity leisure-time PA (HR = 0.63, 95% CI 0.50–0.80) (Model 2). Additional adjustment for diabetes, hypertension and BMI did not attenuate the results noteworthy; low intensity leisure-time PA (HR = 0.73, 95% CI 0.59–0.89); high intensity leisure-time PA (HR = 0.61, 95% CI 0.49–0.77) (Model 3). A statistically significant difference in dementia-related mortality hazard ratios was observed between low and high intensity leisure-time PA (P < 0.05). Psychological distress was associated with an increased risk of dementia-related mortality (HR = 1.45, 95% CI 1.16–1.81) in the fully adjusted model 3.