Reports have suggested relationships of body weight with age-related maculopathy (ARM), particularly its nonneovascular (dry) forms, but results are inconsistent and prospective data are scarce.
To examine prospectively relationships of body mass index (BMI; calculated as weight in kilograms divided by the square of height in meters) with visually significant dry and neovascular ARM during an average of 14.5 years of follow-up.
Incident ARM was assessed by medical record confirmation of self-reported ARM among the 21 121 men participating in the Physicians' Health Study who (1) were followed up for at least 7 years, (2) were free of visually significant ARM at baseline, and (3) had information on BMI and cigarette smoking. We used proportional hazards regression models to estimate rate ratios (RRs) and 95% confidence intervals (CIs) for visually significant dry ARM (256 cases) and neovascular ARM (84 cases) within 4 categories of BMI: lean (<22.0), normal (22.0-24.9), overweight (25.0-29.9), and obese (≥30.0).
Adjusting for age, randomized aspirin and beta carotene assignments, and cigarette smoking, the incidence for visually significant dry ARM was lowest in men with a normal BMI. Compared with these men, the RRs (95% CIs) were as follows: 1.43 (1.01-2.04) for lean, 1.24 (0.93-1.66) for overweight, and 2.15 (1.35-3.45) for obese men. Although there was no significant relationship of BMI with the diagnosis of neovascular ARM, due to the small number of cases these analyses could not rule out an important relationship.
Obesity is a risk factor for visually significant ARM in men, in particular for dry ARM. However, the relationship of BMI with dry ARM appears to be J-shaped, and the leanest individuals also appear to be at increased risk.
AGE-RELATED maculopathy (ARM), a degenerative condition affecting the central regions of the retina and choroid, is the leading cause of blindness among older adults in developed countries, including the United States.1-5 Because its prevalence increases dramatically with age, ARM is likely to become an even greater public health problem in the future. Age-related maculopathy is not treatable in most cases; thus, a preventive approach is desirable. To this end, epidemiologic studies have been undertaken to identify potentially modifiable factors associated with the development of ARM.
Although much progress has been made in understanding the disease, the causes of ARM remain largely unknown. At least as early as 1937, some investigators proposed that ARM might be related to systemic arterial vascular disease.6-8 In conjunction with the retinal pigment epithelium and Bruch membrane, choroidal vessels underlying the macular region of the retina are thought to be involved in clearing metabolic waste material from the eye. It is plausible that a dysfunction in these vessels may be associated with increased risk of ARM. Histologic studies of eyes with ARM have shown a decreased density and cellularity of choroidal capillaries and thickening of intercapillary pillars9-11;angiographic studies have demonstrated slowed filling of the choroidal capillaries.12
Consistent with the idea that vascular disease may be a risk factor for ARM, some,13,14 though not all,15,16 epidemiologic studies have found a higher risk of ARM in subjects with cardiovascular disease (CVD). In addition, a number of studies have found that several CVD risk factors are also associated with ARM. For example, cigarette smoking, a well-known cause of CVD, has been consistently identified as a risk factor for ARM, including in 2 large prospective follow-up studies.17,18 To date, however, studies of other potentially modifiable CVD risk factors that might also increase risk of ARM, such as overweight and obesity,19-22 have yielded inconsistent results,19,23 and there are few prospective studies.20
The Physicians' Health Study (PHS) was a randomized trial of aspirin and beta carotene in prevention of CVD and cancer in men,24,25 in which information on diagnoses of ARM was also collected.17 In the present study we examined prospectively the relationship of body mass index (BMI) with the incidence of ARM during an average of 14.5 years of follow-up. Since it remains unclear whether the factors predisposing to the dry and neovascular forms of ARM are the same, and previous associations between BMI and ARM appeared strongest for dry ARM,20,22 we looked separately at the relationships of BMI with dry and neovascular ARM.
The PHS was a randomized, double blind, placebo-controlled trial that tested the balance of benefits and risks of alternate-day low-dose aspirin and beta carotene on cardiovascular disease and cancer, as well as cataract and macular degeneration.24-26 The PHS population, consisting of 22 071 apparently healthy male US physicians aged 40 to 84 years at study entry, was free of cancer (except possibly basal or squamous cell skin cancer), myocardial infarction, stroke, transient cerebral ischemia, current renal or liver disease, peptic ulcer, and gout. The active treatment phase of both arms of the trial has ended,24,25 but observational follow-up of the PHS cohort is ongoing.
At study entry, participants completed a mailed questionnaire on which they reported their height and weight as well as other characteristics such as blood pressure, diabetes mellitus, high cholesterol, cigarette smoking, alcohol consumption, and vitamin use. Relationships of cigarette smoking,17 alcohol consumption,27 and vitamin use28 with ARM in the PHS have been described in previous reports.
Participants were followed up prospectively every 6 months during the first year and then annually with mailed questionnaires on which they were asked to update some exposure data and report diagnoses of health events. Weight was updated annually starting with the 8-year questionnaire. In a validation study of 123 male health professionals from a similar cohort, the correlation between self-reported and technician measurement of weight was 0.97.29
Beginning with the 7-year follow-up questionnaire, we asked participants about the diagnosis of macular degeneration. On all subsequent questionnaires, information was updated with incident cases of ARM, including the month and year of diagnosis and the name and address of the diagnosing eye doctor, and signed permission to review medical records. For each report of ARM, we sent to the identified ophthalmologist or optometrist a letter containing a brief questionnaire to obtain information on the date of diagnosis, the best-corrected visual acuity at the time of diagnosis, the date when visual acuity first reached 20/30 or worse if later than the date of the initial diagnosis, and the chorioretinal lesions that were observed at diagnosis (drusen; retinal pigment epithelial changes including atrophy, hypertrophy, and retinal pigment epithelial detachment; geographic atrophy; subretinal neovascular membrane; and/or disciform scar). In the presence of other anomalies, the eye doctor was asked to judge whether, in the absence of the other abnormality, he or she would expect the visual acuity to be 20/30 or worse because of ARM. For the present study, ARM was defined as the presence of 1 or more typical lesions associated with a visual acuity loss to 20/30 or worse from these lesions. The visual acuity criterion was included in the definition to reduce the possibility of surveillance bias and because we were interested in determinants of visually significant disease. We defined neovascular macular degeneration as the presence of a retinal pigment epithelial detachment, subretinal neovascular membrane, or disciform scar that was not due to other causes (eg, histoplasmosis or choroidal rupture). Dry ARM was defined as confirmed ARM with vision loss as described above, but with no signs of neovascular macular degeneration.
We excluded 950 subjects who were not followed up for at least 7 years(the first time that ARM was assessed) or had missing information on BMI or cigarette smoking, or who reported a diagnosis of ARM that was made before study entry. After these exclusions, 21 121 participants were followed up from their date of study entry until the date of diagnosis of ARM, death, or December 1997, whichever came first.
In all analyses, we classified individuals rather than eyes, because the same examiner presumably made assessments at the same time for both eyes of each participant (ie, classification of the 2 eyes was not independent). We considered a participant to have ARM at the time it was diagnosed in at least 1 eye. We initially fit separate models for dry and neovascular ARM because risk factors for the 2 forms of the disease may differ, and previous studies indicated that BMI might be a stronger risk factor for the dry form of the disease.20,22 Additional models were also fit for the combined end point of all visually significant ARM. We examined relationships for categories of BMI formed by means of cutoff points defined a priori. We calculated each participant's BMI at the time of each weight assessment as his weight in kilograms divided by the square of his height in meters. We formed 4 categories of BMI (<22.0, 22.0-24.9, 25.0-29.9, and ≥30.0). The upper 2 categories correspond to the definitions of overweight and obesity as adopted by a number of organizations such as the US National Heart, Lung, and Blood Institute and the World Health Organization.30 For clarity of presentation, we have defined the lower 2 categories as lean (<22.0) and normal (22.0-24.9).
In initial analyses, we obtained age- and smoking-adjusted rate ratios(RRs) of ARM by category of BMI in proportional hazards regression models adjusting for age, cigarette smoking (never, past, current <20 cigarettes per day, current ≥20 cigarettes per day), and, because subjects were participants in a randomized trial, randomized aspirin and beta carotene assignments. In these analyses, we allowed BMI to vary over time as a time-varying covariate in the proportional hazards models, using the nearest past BMI measurement available for each participant. In additional models, we further adjusted for height in categories of less than 170 cm, 171 to 178 cm, 179 to 183 cm, and 184 cm or more, as well as other potential risk factors including alcohol consumption and vitamin supplement use. To investigate the possibility of residual confounding, we also fit models in which we adjusted for pack-years of cigarette smoking as described previously,17 as well as alcohol consumption (1 or more drinks per day, 1 to 6 drinks per week, 1 to 3 drinks per month, and rarely or never), vitamin E (never, past, and current) and vitamin C (never, past, and current) supplement use, and mean daily servings of vegetables (sum of servings of broccoli, brussels sprouts, carrots, spinach, dark green lettuce, yellow squash, yams or sweet potatoes, tomato juice, and tomatoes), fruits (sum of servings of orange juice, cantaloupe, peaches, apricots, and nectarines), and cold breakfast cereal. Dietary information was obtained by means of a brief food frequency questionnaire.31 Finally, we explored whether relationships of BMI with ARM were different in younger vs older men by fitting separate proportional hazards models for those who were aged 75 years and older and those who were younger than 75 years. Similarly, we investigated whether the effect of BMI appeared to differ according to smoking status by fitting separate models for never, past, and current smokers.
Visually significant dry ARM was confirmed in 256 participants during a total follow-up of 305 827 person-years (mean, 14.5 years). In addition, 84 participants developed neovascular ARM during 307 341 person-years(mean, 14.6 years) of follow-up. Table 1 displays the prevalence of potential confounding and intermediate factors within categories of baseline BMI. The prevalence of diabetes, hypertension, and cigarette smoking was higher among the men with higher BMI, while the prevalence of vitamin supplement use and daily alcohol consumption was lower.
For visually significant dry ARM, with adjustment for age and cigarette smoking, the relationship with BMI appeared to be J-shaped (Table 2). Compared with men with a normal BMI, lean men had an RR(95% confidence interval [CI]) of ARM of 1.43 (1.01-2.04), overweight men had an RR (95% CI) of 1.24 (0.93-1.66), and obese men had an RR (95% CI) of 2.15 (1.35-3.45). A likelihood ratio test comparing the model with 4 categories of BMI with one with a single variable indicating the trend across BMI categories was significant (P = .008), confirming the nonlinear nature of the BMI and ARM relationship.
Adjustment for height as well as vitamin supplement use and alcohol consumption did not change the magnitude of the association between BMI and dry ARM (Table 2). Further adjustment for frequency of consumption of fruits, vegetables, or breakfast cereal, as well as pack-years rather than 3 categories of cigarette smoking, also had little impact on the RR estimates for obesity (RR [95% CI], 2.18 [1.29-3.68] for obese vs normal), although the RR for the lean men was slightly attenuated(Table 2). The association between BMI and dry ARM (RR, 2.05; 95% CI, 1.26-3.34, for obese vs normal) also persisted in models controlling for the potential intermediate variables diabetes mellitus and hypertension (Table 2).
There was no significant association of BMI with the neovascular form of ARM in these data (Table 3), but the small number of neovascular cases limited the analyses. Lean men had an RR (95% CI) of neovascular ARM of 1.03 (0.56-1.88); overweight men had an RR (95% CI) of 0.81 (0.49-1.34); and obese men had an RR (95% CI) of 1.15(0.45-2.94) compared with men whose BMI was normal.
Results of analyses for the combined end point of all visually significant ARM (dry or neovascular) were similar to those for dry ARM and were most consistent with a J-shaped relationship, although the magnitude of the RRs was attenuated compared with the estimates for dry ARM. Compared with men with a normal BMI, lean men had an RR (95% CI) of ARM of 1.30 (0.95-1.76), overweight men had an RR (95% CI) of 1.08 (0.84-1.39), and obese men had an RR (95% CI) of 1.92(1.27-2.89).
In subgroup analyses (Table 4), the relationship of obesity with dry ARM persisted in men younger than 75 years as well as those aged 75 years and older. However, in these models, the increased risk of dry ARM among the lean men was most apparent for those younger than 75 years. Similarly, obese men appeared to have an increased risk of dry ARM regardless of smoking status, while the increased risk of dry ARM among the lean men was found primarily among the subgroup who had never smoked.
These prospective data from a large cohort of men indicate that obesity is a risk factor for visually significant dry ARM. In addition, the leanest men (those with a BMI <22.0) were also at higher risk of dry ARM. This relationship of BMI with ARM was independent of age and cigarette smoking and did not appear to be explained by an increased risk of diabetes or hypertension, plausible intermediate variables. Although there were no significant findings with regard to neovascular ARM, the number of participants with the neovascular form of ARM was too small to rule out an important effect of body weight.
Strengths of the present study include its large size, prospective design, careful data collection, and low loss to follow-up. We were not able to examine subjects directly, however, so identification of subjects with ARM relied on their seeking medical attention, being diagnosed, and then reporting their diagnoses on study questionnaires. Consequently, underascertainment of ARM is a concern, and the results of the present study must be interpreted in light of this limitation. However, underascertainment of outcomes does not bias results in a follow-up study if the specificity of the diagnosis is high.32 Specificity is likely to be high in the present study because all self-reports of ARM were confirmed by review of medical records, and this method of case detection has been associated with a high specificity(>99%) in a similar cohort of nurses.18 Differential rates of ARM ascertainment by exposure category is of greater concern and could cause bias in either direction. Unfortunately, we did not have information on the frequency with which subjects had their eyes examined, so it was not possible to perform a quantitative assessment of the degree to which detection bias may have influenced our findings. However, to decrease the likelihood of such bias, we limited our analysis to cases of ARM with associated vision loss, since it is less likely that participants with decreased vision would fail to seek medical attention. In addition, when we controlled in some models for factors, such as diabetes mellitus and hypertension, that could lead to more frequent ophthalmic visits, the estimates for BMI were not changed substantially.
All subjects in the present study were male physicians, and therefore this was not a random sample of the total US population. However, if valid, the findings would not be generalizable to other men only if the basic biological mechanisms involved in the development of ARM were somehow different among physicians as compared with other men,32 which seems very unlikely. On the other hand, since many basic differences do exist between men and women, the findings may not be generalizable to women, although other studies have shown similar relationships of BMI with ARM in women and men.20,22 A prospective study of the relationship of BMI with ARM in a large cohort of women would add important information on this issue.
Body weight, a strong predictor of coronary heart disease risk and death from CVD,33-36 was of borderline significance in the Eye Disease Case-Control Study of risk factors for neovascular macular degeneration,21 and the increased risk appeared to be limited to those with BMI of 30 or more. Of 3 cross-sectional studies that have looked at the relationship of BMI with ARM, 1 found no relationship with the combined end point of atrophic or neovascular macular degeneration,23 and 1 observed an increased risk of retinal pigment abnormalities but not neovascular ARM in women with higher BMI.19 In a subsequent report on incidence and progression of ARM during 5 years, these investigators found a significant relationship of higher BMI with increases in retinal pigment in both men and women, but still found no association with other signs of the disease or with neovascular macular degeneration.20 The possibility of a nonlinear association was apparently not investigated in these studies. In a recent cross-sectional study from Australia, a J- or U-shaped association was observed between BMI and early ARM, diagnosed by fundus photographs.22 Similarly, in the present study, the relationship of BMI with visually significant ARM appeared to be J-shaped, with the highest incidence among obese men with a BMI of at least 30 and a somewhat less elevated incidence among the leanest men with a BMI less than 22. Also similar to other authors,20,23 we could identify no significant relationship of BMI with neovascular ARM in the present study, but the number of participants with this late form of macular degeneration was relatively small, and consequently the CIs were too wide to rule out an important effect. Further study of whether BMI is a risk factor for neovascular ARM in studies of sufficient size is needed.
Although these results add to the existing evidence that certain CVD risk factors such as BMI are also related to the development of ARM, the relationship of BMI with ARM appears to be more complex and nonlinear. In particular, results of this as well as 1 previous study in which all subjects were examined are most consistent with an increased risk of dry ARM not only in those who are obese but also in the leanest men; therefore, the relationship of BMI with ARM, if causal, may be mediated by mechanisms other than vascular disease per se, which has a monotonic relationship with BMI. For example, obesity is related to higher levels of oxidative stress, which has been implicated as a probable contributing cause of ARM. It is important to consider that the mechanisms underlying the relationships of ARM with obesity on the one hand and leanness on the other could be different. A relationship of leanness with ARM is more difficult to explain biologically, but similar relationships have been seen, for example, with mortality.36,37 Although the increased risk of mortality among the leanest individuals is largely attributable to the adverse effects of cigarette smoking in at least some studies,36,37 this does not appear to be the case for ARM, for which the relationship of leanness with ARM was, if anything, strongest among never-smokers. One could speculate that deficiencies in 1 or more important nutrients in the diets of the leanest men could have led to the higher risk of ARM we observed among this group. For example, certain micronutrients have been shown previously to predict the development of macular degeneration.38 However, when we explored the possibility that the relationship of BMI with ARM might be explained by differences in consumption of fruits, vegetables, or cereal, we did not observe any indication of substantial confounding effects. Whether the relationship of BMI with ARM might be explained by differential intake of specific nutrients will require further study. Another remaining possibility is that our findings are due to residual confounding by some other factor. One possibility in this regard is that the leanest men may have had a family history of vascular disease that motivated them to remain lean. Indeed, in the PHS, the prevalence of a history of myocardial infarction in either parent before age 60 years was U-shaped and highest among the lean and the obese men. However, controlling for family history of vascular disease did not have any impact on the J-shaped relationship of BMI with ARM in this study(data not shown). Residual confounding by other factors not measured in the current study remains a possible explanation.
In conclusion, results from this prospective study indicate that BMI is an independent predictor of the dry form of visually significant ARM. The relationship appears to be J-shaped, and both obese and lean men appear to have an elevated risk of dry ARM compared with men whose BMI is normal. If indicative of a causal relationship, these data imply that interventions to reduce the prevalence of obesity, which would also result in numerous other health benefits, could help to lessen the incidence of ARM. Our finding of an apparent excess risk of ARM among the leanest men warrants further study.
Accepted for publication February 12, 2001.
Corresponding author: Debra A. Schaumberg, ScD, MPH, Division of Preventive Medicine, Brigham and Women's Hospital, 900 Commonwealth Ave E, Boston, MA 02215 (e-mail: firstname.lastname@example.org).
National Advisory Eye Council, Vision Research: A National Plan, 1999-2003. Rockville, Md National Institutes of Health1998;Report 98-4120
HB Statistics on Blindness in the Model Reporting Area 1969-70. Washington, DC US Dept of Health, Education, and Welfare1973;Report NIH 73-427
J A study of the prevalence of eye disease in the elderly in an English community. Trans Ophthalmol Soc U K.
196- 203Google Scholar
BC Prevalence of ocular disease in a population study of subjects 65 years old and older. Am J Ophthalmol.
1982;94181- 189Google Scholar
K Prevalence of ocular disease and blindness in a rural area in the eastern region of Iceland during 1980 through 1984. Acta Ophthalmol Suppl.
1987;18240- 43Google Scholar
HP Pathogenesis of disciform degeneration of the macula. Arch Ophthalmol.
1937;18561- 585Google ScholarCrossref
JD Pathogenesis of disciform detachment of the neuroepithelium. Am J Ophthalmol.
1967;63suppl1- 139Google Scholar
AL Changes in the choriocapillaris associated with senile macular degeneration. Ann Ophthalmol.
1977;9753-756, 759- 762Google Scholar
RSvan der Schaft
PT Morphometric analysis of Bruch's membrane, the choriocapillaris, and the choroid in aging. Invest Ophthalmol Vis Sci.
1994;352857- 2864Google Scholar
TR Pathogenesis: senile changes of the choriocapillaris of the posterior pole. Trans Am Acad Ophthalmol Otolaryngol.
1965;69652- 661Google Scholar
van der Schaft
PT Histologic features of the early stages of age-related macular degeneration: a statistical analysis. Ophthalmology.
1992;99278- 286Google ScholarCrossref
AC Choroidal perfusion abnormality with age-related Bruch's membrane change. Am J Ophthalmol.
1990;109211- 217Google Scholar
MO Factors associated with age-related macular degeneration: an analysis of data from the first National Health and Nutrition Examination Survey. Am J Epidemiol.
1988;128700- 710Google Scholar
SL Senile macular degeneration: a case-control study. Am J Epidemiol.
1983;118213- 227Google Scholar
DL The Beaver Dam Eye Study: the relation of age-related maculopathy to smoking. Am J Epidemiol.
1993;137190- 200Google Scholar
SE Relation of smoking to the incidence of age-related maculopathy: the Beaver Dam Eye Study. Am J Epidemiol.
1998;147103- 110Google ScholarCrossref
JE A prospective study of cigarette smoking and risk of age-related macular degeneration in men. JAMA.
1996;2761147- 1151Google ScholarCrossref
SE A prospective study of cigarette smoking and age-related macular degeneration in women. JAMA.
1996;2761141- 1146Google ScholarCrossref
T The relationship of cardiovascular disease and its risk factors to age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology.
1993;100406- 414Google ScholarCrossref
SC The relation of cardiovascular disease and its risk factors to the 5-year incidence of age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology.
1997;1041804- 1812Google ScholarCrossref
The Eye Disease Case-Control Study Group, Risk factors for neovascular age-related macular degeneration. Arch Ophthalmol.
1992;1101701- 1708Google ScholarCrossref
JJ Plasma fibrinogen levels, other cardiovascular risk factors, and age-related maculopathy: the Blue Mountains Eye Study. Arch Ophthalmol.
1998;116583- 587Google ScholarCrossref
PT Age-related macular degeneration is associated with atherosclerosis: the Rotterdam Study. Am J Epidemiol.
1995;142404- 409Google Scholar
Steering Committee of the Physicians' Health Study Research Group, Final report on the aspirin component of the ongoing Physicians' Health Study. N Engl J Med.
1989;321129- 135Google ScholarCrossref
et al. Lack of effect of long-term supplementation with beta carotene on the incidence of malignant neoplasms and cardiovascular disease. N Engl J Med.
1996;3341145- 1149Google ScholarCrossref
CH Low-dose aspirin and risks of cataract in a randomized trial of US physicians. Arch Ophthalmol.
1991;109252- 255Google ScholarCrossref
et al. A prospective study of alcohol consumption and the risk of age-related macular degeneration. Ann Epidemiol.
1999;9172- 177Google ScholarCrossref
et al. Prospective cohort study of antioxidant vitamin supplement use and the risk of age-related maculopathy. Am J Epidemiol.
1999;149476- 484Google ScholarCrossref
WC Validity of self-reported waist and hip circumferences in men and women. Epidemiology.
1990;1466- 473Google ScholarCrossref
National Institutes of Health, Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. Obes Res.
1998;6suppl 251S- 209S[published correction appears in Obes Res
. 1998;6:464]Google ScholarCrossref
JE Intake of vegetables rich in carotenoids and risk of coronary heart disease in men: the Physicians' Health Study. Int J Epidemiol.
2001;30130- 135Google ScholarCrossref
S Modern Epidemiology. 2nd Philadelphia, Pa Lippincott-Raven1998;
JW Body mass index and patterns of mortality among Seventh-Day Adventist men. Int J Obes.
1991;15397- 406Google Scholar
et al. Body mass index and body girths as predictors of mortality in black and white men. Am J Epidemiol.
1992;1351137- 1146Google Scholar
CD Body mass index at the age of 18 and its effects on 32-year-mortality from coronary heart disease and cancer: a nested case-control study among the entire 1932 Dutch male birth cohort. J Clin Epidemiol.
1989;42513- 520Google ScholarCrossref
Jr Body weight and mortality: a 27-year follow-up of middle-aged men. JAMA.
1993;2702823- 2828Google ScholarCrossref
D Overweight, underweight, and mortality: a prospective study of 48, 287 men and women. Arch Intern Med.
1996;156958- 963Google ScholarCrossref
et al. for the Eye Disease Case-Control Study Group, Dietary carotenoids, vitamins A, C, and E, and advanced age-related macular degeneration. JAMA.
1994;2721413- 1420[published correction appears in JAMA
. 1995;273:622]Google ScholarCrossref