Physical Activity and the 15-Year Incidence of Age-Related Macular Degeneration

Abstract

Purpose.: There is uncertainty in the published literature as to whether physical activity should be advocated for age-related macular degeneration (AMD) prevention. We aimed to assess prospectively the association between physical activity and the 15-year incidence of AMD in older adults.

Methods.: We assessed AMD from retinal photographs. Participants provided details of walking exercise and the performance of moderate or vigorous activities, which were used to calculate metabolic equivalents (METs).

Results.: After adjusting for age, adults aged ≥75 years in the highest tertile (the most physically active) compared to those in the lowest tertile (least physically active) were 79% less likely to have incident late AMD over the 15 years (odds ratio [OR], 0.21; 95% confidence intervals [CI], 0.05–0.95). However, after further adjusting for sex, body mass index, smoking, fish consumption, and white cell count, this association was no longer statistically significant (OR, 0.26; 95% CI, 0.06–1.28). Significant associations were not found in those aged <75 or with the 15-year cumulative incidence of early AMD.

Conclusions.: Physical activity did not influence the risk of AMD over 15 years in older adults, independent of diet, smoking, white cell count, and body mass index.

Introduction

Age-related macular degeneration (AMD) is a progressive, chronic disease of the central retina, and it is estimated that by 2020, at least 80 million people will be affected by AMD globally.¹ Advanced AMD, including neovascular AMD and geographic atrophy, is associated with substantial, progressive visual impairment.² Current treatment options are limited to patients with late-stage neovascular or intermediate AMD.^3,4 For the majority of older adults with early or no AMD, there is no strategy for disease prevention except the avoidance of cigarette smoking.^5,6 Preventive strategies through dietary modulation also are attractive strategies. The Age-Related Eye Disease Study (AREDS) showed less progression from early to late stage AMD with high-dose zinc/antioxidant supplement versus placebo.⁷ The AREDS-2 also showed benefits from dietary consumption of lutein/zeaxanthin,^8–10 but not for long-chain omega-3 fatty acids as supplements.^11,12 However, associations of single nutrients and AMD often are inconsistent across studies and impossible to totally disentangle from other aspects of diet.¹³ The Carotenoids in Age-Related Eye Disease Study (CAREDS) looked at overall dietary patterns and found that women who scored higher on a modified version of the Healthy Eating Index had lower odds for early AMD.¹³ 

The health benefits of regular physical activity are well established due to its protective influence against obesity, type 2 diabetes, elevated blood pressure, inflammation, coronary heart disease events, and all-cause mortality risk.^14–17 Given that some of these vascular risk factors (e.g., obesity and inflammation) are also postulated to have a role in the pathogenesis of AMD,² there is potential for physical activity to modify the risk of AMD developing in the longer term. Indeed, physical activity was shown to be associated with a lower risk of AMD in previous studies,^13,18,19 including the Beaver Dam Eye Study.²⁰ However, the Eye Disease Case-Control Study reported a nonsignificant association between physical activity and AMD risk.²¹ Of those that observed a significant association, only CAREDS¹³ and a United States cohort study of runners had accounted for the influence of diet on the relationship between exercise and AMD.¹⁹ However, the CAREDS examined the relationship in women only and not in men,¹³ and the latter study of runners did not have clinically verified diagnosis of AMD (i.e., only self-reported AMD); therefore, no information was available on the stage of AMD in this cohort.¹⁹ Given the shortfalls of the previous cohort studies that have examined the relationship between level of physical activity and risk of AMD, further research is needed to contribute toward the evidence base that will help determine whether exercise should or should not be advocated for AMD prevention. 

Therefore, we used a relatively large cohort of older adults aged 49 years and over to determine prospectively whether physical activity influences the risk of early or late AMD developing over 15 years, independent of potential confounders, such as smoking, fish consumption, white cell count (marker of systemic inflammation), and body mass index (BMI). 

Methods

Study Population

The Blue Mountains Eye Study (BMES) is a population-based cohort study of common eye diseases and other health outcomes in a suburban Australian population located west of Sydney. Study methods and procedures have been described previously.²² Participants were noninstitutionalized residents aged 49 years or older invited to attend a detailed baseline eye examination after a door-to-door census of the study area. Baseline examinations of 3654 residents aged >49 years were conducted during 1992 to 1994 (BMES-1, 82.4% participation rate). Surviving baseline participants were invited to attend examinations after 5 (1997–1999, BMES-2), 10 (2002–2004, BMES-3), and 15 (2007–2009, BMES-4) years. At BMES-2, -3, and -4, 2334 (75.1% of survivors), 1952 (75.6% of survivors), and 1149 (55.4% of survivors) participants with complete data were re-examined, respectively. The University of Sydney and the Western Sydney Area Human Ethics Committees approved the study, and written, informed consent was obtained from all participants at each examination, and adhered to the tenets of the Declaration of Helsinki. 

Assessment of AMD

Incidence of AMD was the main outcome, 5, 10, or 15 years later. We took two 30° stereoscopic color retinal photographs of the macula of both eyes, which were graded for presence of early and late AMD using the Wisconsin AMD Grading System.^23,24 Intergrader and intragrader reliability showed good agreement in identifying individual lesions.²⁵ The detailed methodology of AMD ascertainment in this population has been reported extensively.^23,24 Incident early AMD was defined as the absence of late AMD and presence of either large (>125-μm diameter) indistinct soft or reticular drusen, or large distinct soft drusen and retinal pigmentary abnormalities (hyperpigmentation or hypopigmentation) at BMES-2, -3, or -4 in either eye of persons free of early AMD in both eyes at BMES-1.²⁴ Similarly, incident late AMD was defined as the appearance of neovascular AMD or geographic atrophy at BMES-2, ‐3, or ‐4 in either eye of persons without AMD lesions in both eyes at BMES-1.²⁴ Incident any AMD was defined as having early or late AMD at BMES-2, ‐3, or ‐4. A retinal specialist (PM) adjudicated all uncertain retinal pathology and confirmed all late AMD cases. 

Assessment of Physical Activity

Participants provided details of walking exercise and the performance of moderate or vigorous activities,²⁶ which were used to calculate metabolic equivalents (METs) over 1 week.²⁷ Participants answered either yes/no/don't know to the following questions: (1) In the last 2 weeks did you walk for recreation or exercise for at least 10 minutes continuously? (2) In the past 2 weeks did you do any vigorous activity or exercise which made you breathe harder or puff and pant? (e.g., carrying loads, heavy gardening, chopping wood, laboring—at home, during work or anywhere else), and (3) In the past 2 weeks did you do any other leisure time physical activities that you haven't already mentioned? (e.g., more moderate activities, such as lawn bowls, gardening). Participants who answered “yes” then were asked how many times in the last 2 weeks, and the estimated time (in hours and minutes) that they spent walking, doing vigorous activity, and/or leisure time activities. The MET calculations were based on the International Physical Activity Questionnaire scoring protocol.²⁷ The METs data collected at all 4 examinations: BMES-1 (1992–1994), BMES-2 (1997–1999), BMES-3 (2002–2004), and BMES-4 (2007–2009) were used in the current analyses. Specifically, tertiles of METs were assessed in relation to 15-year cumulative incidence of AMD. 

Assessment of Covariates

For the current study, only baseline covariate information was used in the analyses; that is, covariate information collected during BMES-1 (1992–1994). The covariates that were included in the final, multivariable model were those that were identified previously as risk factors for AMD in the BMES cohort, including smoking,²⁸ white cell count,²⁹ and fish consumption.³⁰ Smoking status was determined from history as never smoked, past smoker, and current smoker (which included those who had ceased smoking within the last 12 months). Fasting blood samples collected at BMES-1 also were processed for white cell count. We extracted separate data on the frequency of consuming fish, including salmon, tuna, and sardines from the food frequency questionnaire (collected in BMES-1, during 1992–1994).³¹ We also accounted for baseline BMI, given that it is associated with physical activity and also could be associated with AMD risk. The BMI was calculated as weight divided by height squared (kg/m²). 

Statistical Analyses

We used SAS statistical software (SAS Institute, Cary, NC, USA) version 9.2 for analyses, including t-tests, χ²-tests, and logistic regression. All analyses were stratified by age, given that previously published data showed that the association between physical activity and advanced AMD was modified by age (i.e., present only in those aged >65, but not <65 years).²⁰ Additionally, we observed a significant interaction between age and physical activity on the incidence of late AMD over 15 years (P for interaction = 0.04). Associations between physical activity (in METs) and incident AMD (study outcome) were examined in discrete linear logistic regression models. We first adjusted for potential confounders, including age, sex, smoking, and fish consumption. These variables are likely to obscure or accentuate the relationship between physical activity and AMD, but are unlikely to be an intermediate variable or mediator of any observed relationship between the independent and dependent variable. We then adjusted for potential mediators–BMI and white cell count. This is because it is plausible that physical activity could be linked to AMD via adiposity (BMI) and/or systemic inflammation (white cell count); that is, intermediary pathways.^32–34 Specifically, Model 1 adjusted for age; model 2 further adjusted for sex, smoking, and fish consumption (i.e., potential confounders); and model 3 further adjusted for BMI (as a continuous variable) and white cell count (i.e., potential mediators). We analyzed physical activity as tertiles and used the lowest tertile of METs as the referent group. We used discrete time logistic regression modelling to investigate the associations between tertiles of total METs and risk of incident AMD, where change in METs in the 15-year study period was taken in to account. Findings from all analyses are expressed as adjusted odds ratios (OR) with 95% confidence intervals (CI). Study power calculations were performed with the following assumptions: significance level α set to 0.05 and 28 cases of incident late AMD among those aged 75+ over the 15-year follow-up period. These calculations showed that our study was underpowered, as we only had 70% power to detect a modest association between physical activity and 15-year cumulative incidence of late AMD among those aged 75+ years. 

Results

Of the 3654 participants examined at baseline, 2037 had complete information on AMD during follow-up and also could be included in incidence analyses as they did not have any AMD lesions at baseline. The remaining 1617 were not included in incidence analyses (nonparticipants) because they had incomplete AMD information at follow-up, had either early or late AMD signs at baseline, or could not be followed up at BMES-2, because they had either died or refused to participate. Participants compared to nonparticipants were more likely to be older, lighter, smokers, and have higher white cell count, but were less physically active (Supplementary Table S1). 

Table 1 shows the baseline characteristics of the 2037 participants aged <75 and 75. Those aged 75+ years compared to those aged <75 were more likely to have early or late AMD over the 15 years, but were less likely to smoke, eat fish, and exercise, and had lower BMI. After adjusting for age, adults aged ≥75 in the highest tertile (the most physically active) compared to those in the lowest tertile (least physically active) were 79% less likely to have incident late AMD over the 15 years (Table 2). However, after further multivariable-adjustment (Model 3) this association became marginally nonsignificant (P for trend = 0.08, Table 2). Significant associations between physical activity and late AMD were not observed among those aged <75 years (Table 3). Additionally, nonsignificant associations also were observed between physical activity and early AMD in those aged <75 (P for trend = 0.13) and ≥75 (P for trend = 0.25). 

Table 1

 

Baseline Characteristics of Blue Mountains Eye Study Participants Stratified by Age

Characteristics	Study Participants, n = 2037		P Value
	Aged <75, n = 1801	Aged ≥75, n = 236
Males	779 (43.3%)	102 (43.2%)	0.99
Current smoking	235 (13.4%)	11 (4.8%)	0.0002
BMI, kg/m2	26.5 (4.4)	25.0 (3.8)	<0.0001
White cell count, × 109/L	6.4 (1.7)	6.5 (1.7)	0.16
Physical activity, METs	792 (1569)	495 (1386)	0.01
Fish consumption, ≥1 serve/wk	1092 (60.6%)	121 (51.3%)	0.006
15-y cumulative incidence of AMD
Early	226 (14.0%)	42 (29.2%)	<0.0001
Late	56 (3.1%)	28 (11.9%)	<0.0001

 Data are presented as means ± SD or n (%), except for METs, which is presented as median (interquartile range).

Table 2

 

Association Between Physical Activity and 15-Year Cumulative Incidence of Late AMD Among Participants Aged 75+ Years, Presented as Adjusted OR and 95% CI

Physical Activity, METs	Late AMD
	Model 1*	Model 2†	Model 3‡
First tertile, ≤198, n = 84	1.0 (reference)	1.0 (reference)	1.0 (reference)
Second tertile, ≥297–≤1299, n = 71	0.94 (0.39–2.24)	1.11 (0.45–2.75)	1.03 (0.40–2.65)
Third tertile, >1386, n = 72	0.21 (0.04–0.95)	0.25 (0.05–1.17)	0.26 (0.06–1.28)
P for trend	0.04	0.07	0.08

*  Adjusted for age.

†  Adjusted for age, sex, smoking, and fish consumption.

‡  Adjusted for age, sex, smoking, fish consumption, white cell count, and BMI.

Table 3

 

Association Between Physical Activity and 15-Year Cumulative Incidence of Late AMD Among Participants Aged <75 Years, Presented as Adjusted OR and 95% CI

Physical Activity, METs	Late AMD
	Model 1*	Model 2†	Model 3‡
First tertile, ≤198, n = 548	1.0 (reference)	1.0 (reference)	1.0 (reference)
Second tertile, ≥200–≤1356, n = 560	1.08 (0.56–2.12)	1.31 (0.66–2.62)	1.43 (0.71–2.90)
Third tertile, >1386, n = 663	0.82 (0.41–1.64)	0.99 (0.47–2.06)	1.04 (0.48–2.24)
P for trend	0.32	0.33	0.68

*  Adjusted for age.

†  Adjusted for age, sex, smoking, and fish consumption.

‡  Adjusted for age, sex, smoking, fish consumption, white cell count, and BMI.

Discussion

It has not been established clearly whether lower AMD risk is also an additional benefit of moderate to vigorous physical activity. This cohort study showed that physical activity reduced the risk of developing late AMD by 79% over 15 years in those aged 75+ years. However, this association did not persist after further adjusting for potential confounders, such as diet and smoking. Physical activity did not influence the risk of developing early AMD, nor were there any significant associations observed in those aged <75. 

This study showed a marginally nonsignificant trend for increasing physical activity and decreased risk of incident AMD over 15 years among participants aged 75+ years. Our finding is in agreement with the Eye Disease Case Control Study, which showed that patients with neovascular AMD were less physically active, but the difference failed to reach statistical significance.²¹ Given that the observed association between physical activity and 15-year incidence of late AMD did not persist after further adjusting for smoking, fish intake, adiposity, and white cell count (marker of systemic inflammation), our findings suggested that the influence of physical activity on macula health could be explained by one or more of these factors. Indeed, there is evidence to suggest that habitual physical activity benefits health by reducing abdominal fat, weight, and systemic inflammation.²⁰ Adiposity and inflammation have been hypothesized to have a role in the pathogenesis of AMD.^32–34 It also is likely that older adults who are more physically active also are more likely to lead a healthier lifestyle; that is, consume more fish, fruits, and vegetables, and also are more likely to be biologically younger.²⁰ A healthful diet with adequate consumption of fish was shown previously to have a protective effect on late AMD,^35,36 and, thus, could be another mechanism that explains the reduced AMD risk in older adults who are more physically active. Alternatively, it could be that we had a small number of incident late AMD cases; hence, there could have been insufficient power to detect a modest association with physical activity after adjusting for all covariates. Therefore, additional larger cohort studies with sufficient information on potential confounders are needed to confirm or refute these findings from the BMES. 

In the BMES, a more marked association was observed with the cumulative incidence of late rather than early AMD. This concurs with findings from the Beaver Dam Eye Study which also showed an association between regular physical activity and late, but not early AMD.²⁰ We also have observed inconsistent associations between dietary factors (e.g., fish and dairy foods) and risk of early and late AMD in the BMES.^35,37 It is hypothesized that different pathogenetic mechanisms could be operating at the different AMD stages, which could explain the differential observations with physical activity.²⁰ Additionally, associations were observed only in those aged 75 years and over and not among the younger participants in the BMES. This could be partly due to the greater number of incident cases of late AMD in the older versus younger age group, and, therefore, greater study power to detect a modest association with physical activity. 

Strengths of this study included its prospective design, long-term follow-up of a stable and representative population-based sample, extensive information on other lifestyle risk factors, such as diet and smoking, and high quality stereoscopic retinal photography with validated grading to assess macular conditions.³⁷ However, this study has some noteworthy limitations. First, the number of participants who had incident late AMD was small; hence, confidence intervals for some of the associations are wide. Indeed, our study was not adequately powered as we only had 70% power to detect a modest association between physical activity and incidence of late AMD among participants aged 75+ years. Therefore, future studies that are adequately powered are warranted to examine the temporal association between physical activity and late AMD. Additionally, we only had self-reported measures of physical activity rather than objective measurements of physical activity levels. Nevertheless, self-reported physical activity in large population surveys is common practice,³⁸ as the costs, logistics, and expertise required to obtain accelerometer-measured METs can be prohibitive. Finally, participants compared to nonparticipants differed in certain study characteristics, such as age, BMI, smoking status, and level of physical activity. Hence, our findings must to be interpreted with caution as these differences between participants and nonparticipants could have influenced observed associations. 

In summary, this prospective observational study in community-living older adults shows that physical activity was not independently associated with risk of developing AMD over 15 years, after accounting for diet, smoking, white cell count, and body mass. This suggests that these factors and not level of physical activity, which could be responsible for the protective associations observed with exercise. Nevertheless, it is possible that our cohort study is underpowered to detect a modest association between physical activity levels and incidence of AMD. Hence, further investigation by larger population-based studies is warranted to confirm whether or not physical activity could be prescribed as a preventive strategy in minimizing the risk of developing AMD among older adults. 

Acknowledgments

Supported by Australian National Health and Medical Research Council Grants 974159, 991407, 211069, and 262120 (Blue Mountain Eye Study), and Westmead Millennium Institute (Blue Mountain Eye Study), and by a Macular Disease Foundation Australia and Blackmores Dr Paul Beaumont Fellowship (BG). The authors alone are responsible for the content and writing of the paper. 

Disclosure: B. Gopinath, None; G. Liew, None; G. Burlutsky, None; P. Mitchell, None