Abstract
Purpose
To examine potential relationships between alcohol consumption and age-related macular degeneration (AMD) progression, including progression to late AMD and geographic atrophy (GA) enlargement rate.
Design
Post hoc analysis of cohorts within the Age-Related Eye Diseases Study 2 (AREDS2).
Participants
6670 eyes (of 3673 participants) with no late AMD at baseline; 1143 eyes (of 841 participants) with GA at ≥2 consecutive visits.
Methods
Color fundus photographs were collected at annual study visits and graded centrally for late AMD, GA area, and GA proximity. Alcohol consumption was calculated by food frequency questionnaire. Regression analyses of disease progression were performed according to alcohol consumption.
Main outcome measures
Progression to late AMD and its subtypes; GA area-based progression; GA proximity-based progression.
Results
Over mean follow-up of 3.8 years, 40.2% of eyes progressed to late AMD. In men, with alcohol tertile 1 (no regular consumption) as reference, hazard ratios for progression to late AMD were 0.69 (95% CI 0.55–0.87, p=0.0015) for tertile 2 and 0.85 (0.71–1.02, p=0.079) for tertile 3. In women, hazard ratios were 1.12 (0.95–1.31, p=0.17) and 0.85 (0.72–1.00, p=0.046), respectively. Over mean follow-up of 3.1 years, GA area-based progression was significantly faster in women than men, at 0.295 (95% CI 0.278–0.311) and 0.260 mm/year (0.241–0.279), respectively (p=0.007). In men, area-based progression differed significantly by alcohol tertile (p=0.0001), at 0.275 (0.248–0.303), 0.183 (0.143–0.223), and 0.280 mm/year (0.254–0.306) in tertiles 1–3, respectively. In women, the area-based rate did not differ significantly by alcohol tertile (p=0.11). In men only, CDC-defined heavy drinking was associated with faster progression (p=0.024), at 0.306 (0.262–0.349) vs 0.252 mm/year (0.233–0.270). In 808 eyes with non-central GA, GA proximity-based progression did not differ significantly by alcohol tertile (p=0.55).
Conclusions
Moderate alcohol consumption is associated with decreased risk of progression to late AMD in men. GA progression is faster in women, but its relationship with alcohol consumption is much stronger in men. In men, moderate consumption is associated with slower GA progression and higher consumption with faster progression. Although some of these associations may also relate to confounding, they might suggest that individuals with GA should avoid high alcohol consumption.
Précis
Relationships between alcohol consumption and age-related macular degeneration progression differ markedly by sex. In men, moderate consumption is associated with decreased risk of progression to late disease and high consumption with faster geographic atrophy expansion.
Introduction
Age-related macular degeneration (AMD) is the leading cause of legal blindness in industrialized countries.1 Of the two subtypes of late AMD, geographic atrophy (GA) is the defining lesion of late atrophic disease and is estimated to affect over five million people worldwide.2 In most cases, it arises outside the central macula (i.e., non-central GA) and expands gradually, with increasing loss of the central visual field.3–5 Its progression rate is often measured as change over time in GA area; indeed, this outcome measure is recognized as a clinically important endpoint by the United States Food and Drug Administration (FDA).6 For non-central GA, change over time in GA proximity to the central macula is an important complementary outcome measure, as it is related to the beneficial phenomenon of foveal sparing and has important implications for visual prognosis.7–11
From multiple previous studies, the environmental factors associated with altered risk of progression to late AMD are known to include diet and cigarette smoking.12–15 Alcohol consumption might represent another important modifiable factor for disease progression, either harmful or beneficial, but potential relationships between alcohol consumption and disease progression remain poorly understood. Previous studies have reported both harmful and protective associations.16,17 A systematic review and meta-analysis reported in 2019 that the highest alcohol consumption category was associated with increased risk of developing early AMD but not late AMD.16 However, the study binarized alcohol consumption so was unable to examine associations with moderate alcohol consumption or consider the possibility of non-linear relationships. In addition, we are aware of only one previous study, from our own research group, to consider possible relationships between alcohol consumption and altered GA progression rate.18
Our research group has previously analysed risk of progression to late AMD in the Age-Related Eye Diseases Study (AREDS) and AREDS2 datasets according to the alternative Mediterranean diet index (aMedi).15 Alcohol consumption formed a small part of these analyses, as it represents one of the nine aMedi components. However, these analyses were relatively limited in their nature, as the aMedi considers alcohol consumption as a binary variable only, with no distinction made between zero consumption and very high consumption.19 Similarly, we have previously analyzed GA progression in the AREDS2 dataset according to the aMedi.18 Indeed, in these analyses, alcohol consumption within the aMedi-specified interval was strongly associated with slower GA area-based progression. However, again, it was not clear whether this protective association applied to avoiding excess alcohol, avoiding abstinence, or both. In addition, no proximity-based analyses were performed. Finally, in all cases, no separate analyses for men and women were performed; these may be important, since both alcohol consumption and metabolism tend to differ between the sexes.20,21
The AREDS2 was a multicenter, randomized controlled clinical trial designed to assess the effects of oral micronutrient supplements on the course of AMD in people at moderate to high risk of progression to late disease.21 Detailed information was collected on alcohol consumption and diet. The aim of this study was to explore potential relationships between alcohol consumption and altered risk of AMD progression in the AREDS2, including progression to late AMD and its subtypes, GA area-based progression, and GA proximity-based progression.
Methods
Study Population and Procedures
The AREDS2 study design has been described previously.22 In brief, 4,203 participants (aged 50–85 years) were recruited between 2006 and 2008 at 82 US retinal specialty clinics. Inclusion criteria were the presence of either large drusen in both eyes or late AMD in one eye and large drusen in the fellow eye. Institutional review board approval was obtained at each site and written informed consent was obtained from all participants. The research was conducted under the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act.
The AREDS2 participants were randomly assigned to receive the supplements that lowered risk of AMD progression in the AREDS, either (i) alone, or with additional (ii) lutein/zeaxanthin, (iii) docosahexaenoic acid plus eicosapentaenoic acid, or (iv) the combination. At baseline and annual visits, eye examinations were performed, and digital stereoscopic color fundus photographs were captured and graded centrally at the Wisconsin Reading Center.23 The participants, investigators, and reading center personnel were masked to the treatment assignments. The randomized clinical trial lasted five years. Progression to late AMD (including GA/neovascular subtype) was defined by fundus photograph grades, together with history of intravitreal injections for neovascular AMD.22
Evaluation of Geographic Atrophy on Color Fundus Photographs
The definitions of GA and methods to measure GA area and other characteristics have been described previously.3,8,23 In brief, GA was defined as a lesion equal to or larger than drusen circle I-2 (diameter 433 um, area 0.146 mm2, i.e., 1/4 disc diameter and 1/16 disc area) at its widest diameter with at least two of the following features present: circular shape, sharp (well-demarcated) edges, and loss of the retinal pigment epithelium (RPE; partial or complete depigmentation of the RPE, typically with exposure of underlying choroidal vessels). Planimetry tools were used to demarcate the area of GA within the AREDS grid in square millimeters. In the case of multifocal GA, areas were summed to yield a single value for analysis. If the GA was non-central, GA proximity to the central macula (i.e., the smallest distance between the foveal center-point and the border of the GA closest to the foveal center-point) was documented in microns (using the closest border of the closest lesion, in the case of multifocal GA).3,8,23 Grading for GA area measurements and other features was performed independently at the image level, i.e., the reading center graders analyzed each image independently from other images in the full time-series of images for each eye and did not have access to any accompanying clinical information.
Evaluation of Alcohol Consumption and Modified Alternative Mediterranean Diet Index Score
The assessment of alcohol consumption and the aMedi in the AREDS2 has been described previously.15,18 In brief, a 131-item, semi-quantitative Harvard food frequency questionnaire (FFQ) was administered to all participants at randomization.24,25 Participants were asked how often, on average, they had consumed each food/beverage item during the preceding year. There were five questions related to alcohol consumption, one for each of the following five beverage categories (with the unit of consumption given in parentheses): regular beer (1 glass/bottle/can); light beer (1 glass/bottle/can); red wine (4 oz. glass); white wine (4 oz. glass); liquor (1 drink or shot). Each of these five questions had one of 10 possible responses: never; less than once per month; 1–3 per month; 1 per week; 2–4 per week; 5–6 per week; 1 per day; 2–3 per day; 4–5 per day; 6+ per day. The FFQ responses were used to quantify for each participant the consumption of alcohol (in gram/day) and of each food item (in number of medium-sized servings per week).
The aMedi consists of nine components, including alcohol: whole fruits, vegetables, whole grains, nuts, legumes, red meat, fish, monounsaturated fatty acid: saturated fatty acid ratio, and alcohol. Typically, as in our previous analyses15,18, the FFQ data are summed for each participant to obtain the consumption for each of the nine components. For each component, sex-specific consumption quartiles (1–4) are calculated, with quartile 4 representing highest consumption. The quartiles for red meat are reversed (i.e., quartile 4 with highest consumption scored 1, as least aMedi-adherent, and quartile 1 with lowest consumption scored 4). Alcohol consumption is considered in a binary way as consumption above or below specified intervals.19 Then, to calculate the aMedi score for each participant, the quartile values for the nine components are summed (range 8–36). However, for the current analyses, focused specifically on alcohol consumption, the aMedi was used for covariate adjustment rather than as the main variable of interest. For this reason, the aMedi was modified to exclude alcohol consumption. Specifically, the modified aMedi score used in this study was calculated for each participant by summing the quartile values for the eight other components only, without alcohol (range 8–32).
Study Populations and Statistical Methods
The outcome measures comprised: (i) progression to late AMD, and its individual subtypes, (ii) GA area-based progression, and (iii) GA proximity-based progression. For the analyses of risk of progression to late AMD and its individual subtypes, the study population comprised all eyes without late AMD at baseline in participants with at least two study visits. For the analyses of the GA area-based progression rate, the study population comprised all eyes that had GA measurements available at two or more study visits (without previous or simultaneous neovascular AMD). This included eyes that had GA at baseline (i.e., prevalent GA) and those in which GA developed during follow-up (i.e., incident GA). For the analyses of the GA proximity-based progression rate, the study population comprised only eyes where GA was non-central (i.e., with a proximity variable greater than 0) at the first time-point. In all analyses, participants were excluded if they had no FFQ or missing data for any covariate.
For the analyses of progression to late AMD, multivariable proportional hazards regression analyses were performed separately for progression to (i) late AMD, (ii) GA, and (iii) neovascular AMD, according to the alcohol consumption variable, using methods similar to those described previously.15 The unit of analysis was the eye. The proportional hazards assumption was tested in all cases. The models included terms for age, sex, smoking status, education level, total calorie consumption, modified aMedi, and randomized assignment to lutein/zeaxanthin vs no lutein/zeaxanthin. Adjustment for correlation between eyes was made in SAS by using the robust sandwich estimate for the covariance matrix in the Wald tests.
For the analyses of GA area-based progression, mixed-model repeated-measures regression was performed with square root of GA area as the outcome measure, using methods similar to those described previously.3,8,18 The unit of analysis was the eye. The square root transformation was used, to reduce the dependence of area-based progression rate on baseline lesion size.3,26,27 The models included the alcohol consumption variable, years from first time-point with GA, and their interaction term. The models also included terms for age, sex, smoking status, education level, total calorie consumption, modified aMedi, randomized assignment to lutein/zeaxanthin vs no lutein/zeaxanthin, and square root of GA area at first time-point with GA. To account for the correlation between both eyes of the same participant and between different visits of the same eye, an unstructured and a first-order autoregressive covariance structure (UN@AR(1)), respectively, was specified.
For the analyses of GA proximity-based progression, mixed-model repeated-measures regression was performed with the proximity variable as the outcome measure, using methods similar to those described previously.8,18 The unit of analysis was the eye. The models included the alcohol consumption variable, years from first time-point with GA, and their interaction term. The models also included terms for age, sex, smoking status, education level, total calorie consumption, modified aMedi, randomized assignment to lutein/zeaxanthin vs no lutein/zeaxanthin, and proximity at first time-point with GA. Again, the UN@AR(1) covariance structure was specified. In these analyses, on the rare occasion that proximity reached zero during follow-up, all subsequent time-points were censored.
For all the outcomes measures described above, the primary analyses were conducted by considering alcohol consumption in tertiles, with tertile 1 (as reference) representing no alcohol consumption, tertile 2 positive but lower consumption, and tertile 3 positive and higher consumption. The threshold between tertiles 2 and 3 was 6.0 gram/day of alcohol (equivalent to 3.0 standard drinks per week) for both men and women. In secondary analyses, alcohol consumption was considered as a binary variable, either below the United States Centers for Disease Control and Prevention (CDC) threshold definition of heavy drinking (as reference) or above it.28 The CDC defines heavy drinking as consuming 15 standard drinks or more per week (for men) or eight standard drinks or more per week (for women), i.e., equivalent to a weekly alcohol consumption of 210 gram or 112 gram, respectively.28 All analyses were performed for men and women together and for each sex separately, given differences in patterns of alcohol consumption and metabolism between the sexes.20,21 The analyses were performed with commercially available statistical software (SAS version 9.4; SAS Institute, Cary, NC).
Results
Progression to Late Age-Related Macular Degeneration and its Subtypes according to Alcohol Consumption
The study population for these analyses comprised 6670 eyes of 3673 participants. Their characteristics are shown in Table 1. The distribution of alcohol consumption, separately for men and women, is shown in Figure 1. Median alcohol consumption was 4.5 gram/day in men and 0.8 gram/day in women. Over mean follow-up of 3.8 years (standard deviation [SD] 2.0 years), the numbers of eyes that progressed to late AMD, GA, or neovascular AMD were 2683 (40.2%), 1347 (20.2%), and 1551 (23.3%), respectively.
Table 1.
Demographic and Clinical Characteristics of the Three Study Populations at Baseline.
| AMD progression cohort | GA area cohort | GA proximity cohort | ||
|---|---|---|---|---|
| Participants: n | 3673 | 841 | 638 | |
| Age: mean (SD) | 72.9 (7.7) | 74.9 (6.8) | 74.8 (6.9) | |
| Sex: n (%) | Female | 2090 (56.9) | 486 (57.8) | 377 (59.1) |
| Male | 1583 (43.1) | 355 (42.2) | 261 (40.9) | |
| Race: n (%) | White | 3549 (96.6) | 832 (98.9) | 631 (98.9) |
| Non-white | 124 (3.4) | 9 (1.1) | 7 (1.1) | |
| Smoking status: n (%) | Never | 1593 (43.4) | 342 (40.7) | 257 (40.3) |
| Former | 1843 (50.2) | 446 (53) | 343 (53.8) | |
| Current | 237 (6.5) | 53 (6.3) | 38 (6) | |
| Education: n (%) | HS or less | 1159 (31.6) | 302 (35.9) | 233 (36.5) |
| At least some college | 1729 (47.1) | 385 (45.8) | 287 (45) | |
| Post-graduate | 785 (21.4) | 154 (18.3) | 118 (18.5) | |
| Follow-up time (years): mean (SD) | 3.8 (2.0) | 3.1 (1.5) | 3.3 (1.5) | |
| Eyes: n | 6670 | 1143 | 808 | |
| GA prevalent at study baseline or incident during follow-up: n (%) | Incident | 713 (62.4) | 506 (62.6) | |
| Prevalent | 430 (37.6) | 302 (37.4) | ||
| GA central involvement at first appearance: n (%) | Central | 316 (27.6) | 0 (0.0) | |
| Non-central | 827 (72.4) | 808 (100.0) | ||
| GA configuration at first appearance: n (%) | Small (single patch <1DA) | 567 (49.6) | 420 (52.0) | |
| Multifocal | 267 (23.4) | 242 (30.0) | ||
| Horseshoe, Ring | 53 (4.6) | 50 (6.2) | ||
| Solid | 216 (18.9) | 70 (8.7) | ||
| Indeterminate | 40 (3.5) | 26 (3.2) | ||
| AMD severity category ≥ 7: n (%) | No | 1601 (24.0) | ||
| Yes | 5021 (75.3) | |||
| Cannot Grade | 48 (0.7) | |||
| GA area at first appearance (mm2): mean (SD) | 2.4 (3.3) | 2.0 (3.0) | ||
| GA proximity to fovea at first appearance (μm): mean (SD) | 428.7 (498.1) | 592.2 (498.0) | ||
AMD = age-related macular degeneration; DA = disc areas; GA = geographic atrophy; HS = high school; SD = standard deviation
Figure 1.
Histogram of alcohol consumption in the progression to late age-related macular degeneration study cohort, separately for men and women.
The results of the multivariable proportional hazards regression analyses, with alcohol consumption considered in tertiles, are shown in Table 2. The hazard ratios for progression to late AMD were 0.93 (95% confidence interval [CI] 0.81–1.06, p=0.29) for tertile 2 and 0.87 (95% CI 0.77–0.98, p=0.019) for tertile 3. Hence, higher alcohol consumption was significantly associated with decreased risk of progression to late AMD, compared to no alcohol consumption, though with a low magnitude of difference. The equivalent results for progression to GA and to neovascular AMD are also shown.
Table 2.
Results of Proportional Hazards Regression of Progression to Late Age-Related Macular Degeneration Outcomes according to Alcohol Consumption.
| Men & Women (n = 6670 Eyes of 3673 Participants) | Women Only (n = 3813 Eyes of 2090 Participants) | Men Only (n = 2857 Eyes of 1583 Participants) | |||||
|---|---|---|---|---|---|---|---|
| Outcome | Exposure | HR (95% CL) | P Value | HR (95% CL) | P Value | HR (95% CL) | P Value |
| Late AMD | Alcohol g/day – tertile 2 | 0.93 (0.82, 1.06) | 0.29 | 1.12 (0.95, 1.31) | 0.17 | 0.69 (0.55, 0.87) | 0.0015 |
| Alcohol g/day – tertile 3 | 0.87 (0.77, 0.98) | 0.019 | 0.85 (0.72, 1.00) | 0.046 | 0.85 (0.71, 1.02) | 0.079 | |
| Geographic Atrophy | Alcohol g/day – tertile 2 | 0.96 (0.79, 1.15) | 0.64 | 1.14 (0.91, 1.44) | 0.26 | 0.73 (0.53, 1.01) | 0.059 |
| Alcohol g/day – tertile 3 | 0.90 (0.76, 1.07) | 0.22 | 0.84 (0.66, 1.06) | 0.14 | 0.94 (0.73, 1.21) | 0.64 | |
| Neovascular AMD | Alcohol g/day – tertile 2 | 0.95 (0.80, 1.12) | 0.51 | 1.09 (0.89, 1.33) | 0.41 | 0.74 (0.56, 0.99) | 0.042 |
| Alcohol g/day – tertile 3 | 0.85 (0.73, 0.99) | 0.041 | 0.86 (0.70, 1.06) | 0.16 | 0.81 (0.64, 1.03) | 0.085 | |
| Late AMD | CDC-defined heavy drinking | 0.94 (0.79, 1.11) | 0.47 | 0.85 (0.66, 1.08) | 0.18 | 1.02 (0.80, 1.31) | 0.84 |
| Geographic Atrophy | CDC-defined heavy drinking | 0.99 (0.77, 1.27) | 0.93 | 0.78 (0.54, 1.13) | 0.19 | 1.22 (0.88, 1.71) | 0.24 |
| Neovascular AMD | CDC-defined heavy drinking | 0.92 (0.74, 1.15) | 0.47 | 0.98 (0.73, 1.33) | 0.91 | 0.85 (0.61, 1.19) | 0.34 |
AMD = age-related macular degeneration; CDC = Centers for Disease Control and Prevention; CL = confidence limits; HR = Hazard Ratio
The results of the analyses of men and women, considered separately, are also shown in Table 2. For men, the hazard ratios for progression to late AMD were 0.69 (95% CI 0.55–0.87, p=0.0015) for tertile 2 and 0.85 (95% CI 0.71–1.02, p=0.079) for tertile 3, with relatively similar results for the two late AMD subtypes. By contrast, for women, the hazard ratios for progression to late AMD were 1.12 (95% CI 0.95–1.31, p=0.17) for tertile 2 and 0.85 (95% CI 0.72–1.00, p=0.046) for tertile 3, again, with relatively similar results for the two late AMD subtypes. Hence, for men only, positive but lower alcohol consumption was significantly associated with decreased risk of progression to late AMD, compared to no alcohol consumption, but this was not true for higher alcohol consumption.
The results of the secondary analyses, with alcohol consumption considered above versus below the CDC definition of heavy drinking, are also shown in Table 2. For progression to late AMD, the hazard ratio associated with consumption above the CDC threshold was 0.93 (95% CI 0.78–1.10, p=0.40). Hence, alcohol consumption meeting the CDC definition of heavy drinking was not significantly associated with altered risk of progression to late AMD. This was also true for men and women, considered separately. The equivalent results for progression to GA and to neovascular AMD are also shown.
Geographic Atrophy Area-Based Progression according to Alcohol Consumption
The study population for these analyses comprised 1143 eyes of 841 participants. Their characteristics are shown in Table 1. The distribution of alcohol consumption, separately for men and women, is shown in Figure 2. Median alcohol consumption was 3.5 gram/day in men and 0.0 gram/day in women. Mean follow-up was 3.1 years (SD 1.5 years). The progression rate was significantly different in men and women (p=0.007), with faster rates in women, at 0.295 mm/year (95% CI 0.278–0.311 mm/year) in women and 0.260 mm/year (95% CI 0.241–0.279 mm/year) in men.
Figure 2.
Histogram of alcohol consumption in the geographic atrophy area-based study cohort, separately for men and women.
The results of the multivariable regression analyses, with alcohol consumption considered in tertiles, are shown in Table 3. The GA area-based progression rate differed significantly according to alcohol tertile (p=0.041). The estimates for the progression rates were 0.296 mm/year (95% CI 0.278–0.315 mm/year) for tertile 1, 0.258 mm/year (95% CI 0.232–0.283 mm/year, p=0.016) for tertile 2, and 0.272 mm/year (95% CI 0.250–0.295 mm/year, p=0.11) for tertile 3. In the analyses considering alcohol consumption above versus below the CDC definition of heavy drinking, no significant difference was observed (Table 3).
Table 3.
Geographic Atrophy Area-Based Progression Rates according to Alcohol Consumption.
| Men & Women (n = 1143 Eyes of 841 Participants) | Women Only (n = 666 Eyes of 486 Participants) | Men Only (n = 477 Eyes of 355 Participants) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Exposure | Estimate, mm/year (95% CL) | Pairwise P Value | Interaction P Value | Estimate, mm/year (95% CL) | Pairwise P Value | Interaction P Value | Estimate, mm/year (95% CL) | Pairwise P Value | Interaction P Value |
| Alcohol g/day tertile | |||||||||
| Tertile 1 | 0.296 (0.278, 0.315) | . | 0.041 | 0.309 (0.284, 0.333) | . | 0.11 | 0.275 (0.248, 0.303) | . | 0.0001 |
| Tertile 2 | 0.258 (0.232, 0.283) | 0.016 | . | 0.295 (0.262, 0.328) | 0.51 | . | 0.183 (0.143, 0.223) | 0.0002 | . |
| Tertile 3 | 0.272 (0.250, 0.295) | 0.11 | . | 0.260 (0.221, 0.299) | 0.037 | . | 0.280 (0.254, 0.306) | 0.81 | . |
| CDC-defined heavy drinking | |||||||||
| No | 0.279 (0.266, 0.292) | 0.69 | 0.69 | 0.298 (0.280, 0.316) | 0.18 | 0.18 | 0.252 (0.233, 0.270) | 0.024 | 0.024 |
| Yes | 0.287 (0.248, 0.326) | . | . | 0.248 (0.177, 0.319) | . | . | 0.306 (0.262, 0.349) | . | . |
CDC = Centers for Disease Control and Prevention; CL = confidence limits
Mixed-model, repeated-measures regression with square root of geographic atrophy (GA) area as dependent variable, according to alcohol consumption, years from GA first appearance/baseline, and their interaction term, with adjustment for age, sex, smoking, educational level, total calorie intake, and modified Alternative Mediterranean Diet Index score
The results of the analyses of men and women, considered separately, are also shown in Table 3. For men, the GA area-based progression rate differed highly significantly according to consumption tertile (p=0.0001), with large magnitudes of difference between the estimates. The estimates were 0.275 mm/year (95% CI 0.248–0.303 mm/year) for tertile 1, 0.183 mm/year (95% CI 0.143–0.223 mm/year, p=0.0002) for tertile 2, and 0.280 mm/year (95% CI 0.254–0.306 mm/year, p=0.81) for tertile 3. Hence, both no alcohol consumption and higher consumption were associated with faster GA area-based progression, while positive but lower alcohol consumption was associated with substantially slower progression, consistent with a non-linear relationship. In the analyses considering alcohol consumption above versus below the CDC definition of heavy drinking, consumption above the threshold was significantly associated with faster GA area-based progression (p=0.024), with estimates of 0.306 (95% CI 0.262–0.349 mm/year) and 0.252 mm/year (95% CI 0.233–0.270 mm/year), respectively.
By contrast, for women, a different pattern of results was observed. The progression rate did not differ significantly according to consumption tertile (p=0.11). The estimates were 0.309 mm/year (95% CI 0.284–0.333 mm/year) for tertile 1, 0.295 mm/year (95% CI 0.262–0.328 mm/year, p=0.51) for tertile 2, and 0.260 mm/year (95% CI 0.221–0.299 mm/year, p=0.037) for tertile 3. In the analyses considering alcohol consumption above versus below the CDC definition of heavy drinking, no significant difference was observed (p=0.18).
Geographic Atrophy Proximity-Based Progression according to Alcohol Consumption
The study population for these analyses comprised 808 eyes of 638 participants. Their characteristics are shown in Table 1. The distribution of alcohol consumption, separately for men and women, is shown in Figure 3. Median alcohol consumption was 4.8 gram/day in men and 0.0 gram/day in women. Mean follow-up was 3.3 years (SD 1.5 years). The GA proximity-based progression rate was not significantly different in men and women (p=0.80), with estimates of 91.5 μm/year (95% CI 78.7–104.4 μm/year) for men and 93.7 μm/year (95% CI 82.9–104.6 μm/year) for women.
Figure 3.
Histogram of alcohol consumption in the geographic atrophy proximity-based study cohort, separately for men and women.
The results of the multivariable regression analyses, with alcohol consumption considered in tertiles, are shown in Table 4. The GA proximity-based progression rate did not differ significantly according to alcohol tertile (p=0.57). The estimates were 95.3 μm/year (95% CI 82.9–107.5 μm/year) for tertile 1, 95.0 μm/year (95% CI 78.2–111.8 μm/year, p=0.98) for tertile 2, and 85.1 μm/year (95% CI 69.4–100.8 μm/year, p=0.32) for tertile 3. Similarly, in the analyses considering alcohol consumption above versus below the CDC definition of heavy drinking, no significant difference was observed (Table 4). In the analyses of men and women separately, similar results were obtained: no significant difference in the GA proximity-based progression rate was observed according to alcohol consumption tertile either for men or for women (Table 4). The same was true for alcohol consumption above versus below the CDC threshold.
Table 4.
Geographic Atrophy Proximity-Based Progression Rates according to Alcohol Consumption.
| Men & Women (n = 808 Eyes of 638 Participants) | Women Only (n = 488 Eyes of 377 Participants) | Men Only (n = 320 Eyes of 261 Participants) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Exposure | Estimate, μm/year (95% CL) | Pairwise P Value | Interaction P Value | Estimate, μm/year (95% CL) | Pairwise P Value | Interaction P Value | Estimate, μm/year (95% CL) | Pairwise P Value | Interaction P Value |
| Alcohol g/day tertile | |||||||||
| Tertile 1 | 95.3 (82.9, 107.5) | . | 0.57 | 94.2 (79.8, 108.6) | . | 0.87 | 98.0 (75.0, 120.9) | . | 0.61 |
| Tertile 2 | 95.0 (78.2, 111.8) | 0.98 | . | 94.7 (75.0, 114.4) | 0.97 | . | 98.7 (67.6, 129.7) | 0.97 | . |
| Tertile 3 | 85.1 (69.4, 100.8) | 0.32 | . | 86.8 (60.8, 112.8) | 0.63 | . | 84.2 (63.7, 104.7) | 0.38 | . |
| CDC-defined heavy drinking | |||||||||
| No | 90.8 (82.0, 99.6) | 0.24 | 0.24 | 92.2 (81.4, 103.0) | 0.28 | 0.28 | 89.5 (74.6, 104.3) | 0.42 | 0.42 |
| Yes | 108.3 (80.4, 136.3) | . | . | 120.9 (69.4, 172.3) | . | . | 104.9 (70.0, 139.8) | . | . |
CDC = Centers for Disease Control and Prevention; CL = confidence limits
Mixed-model, repeated-measures regression with geographic atrophy (GA) proximity to macular center-point as dependent variable, according to alcohol consumption, years from GA first appearance/baseline, and their interaction term, with adjustment for age, sex, smoking, educational level, total calorie intake, and modified Alternative Mediterranean Diet Index score
Discussion
Main Findings, Interpretation, and Clinical Implications
Detailed analyses of the AREDS2 have permitted a nuanced view of the complex relationship between alcohol consumption and AMD progression, by separate analyses of (i) late AMD and GA area-based progression, (ii) men and women, and (iii) different alcohol consumption thresholds (including non-binary categorization). Importantly, the relationship between alcohol consumption and AMD progression appears to differ substantially between men and women.
Regarding the risk of progression to late AMD, in men, positive but lower alcohol consumption was strongly associated with decreased risk of progression to late AMD, while higher consumption had no significant association. The pattern of results was relatively similar for each late AMD subtype considered separately. By contrast, in women, associations were generally absent or weak. It is possible that, in men, moderate alcohol consumption genuinely causes decreased risk of progression to late AMD. However, despite adjustment for several relevant variables in the statistical models, it is also possible that these findings are related at least partially to confounding, for example with socioeconomic status. This is because higher socioeconomic status is often associated with higher prevalence of regular alcohol consumption.29 For example, in the current study, the strength of the observed associations was weaker following adjustment for education level. The results might also relate partially to the possibility of the zero-alcohol consumption group containing some former heavy drinkers.
One important finding to emerge was that, in the AREDS2, the speed of GA area-based progression differs significantly between men and women. Following square root transformation, the rate was approximately 13% faster in women, despite adjustment for multiple relevant covariates. With one exception30, most previous studies have not observed any significant difference between the sexes, though the sample size in most previous studies has been much smaller than that of the current study.4 This observation may be important for prognostic information and clinical decision-making regarding potential treatments to slow GA progression.
Importantly, as for risk of progression to late AMD, the relationship between alcohol consumption and GA area-based progression appears strikingly different between men and women. For men, positive but lower alcohol was strongly associated with very substantially slower area-based progression, while both zero consumption and higher consumption were associated with faster progression. The differences were as high as approximately 40%, despite square root transformation and adjustment for multiple relevant variables, including education level, smoking status, aMedi, and GA characteristics. By contrast, the results for women demonstrated absent or only modest differences in rates of GA area-based progression, according to alcohol consumption. This contrast might be related to the lower prevalence of alcohol consumption and heavy drinking in women in this study and/or genuine differences in the biological relationships between alcohol consumption and GA progression. It is possible that, in men, modest levels of alcohol consumption might genuinely slow GA progression. However, the results might also be related partially to residual confounding and/or the zero-alcohol consumption group containing some former heavy drinkers. Similarly, high levels of alcohol consumption may causally speed up GA progression, though, again, a degree of residual confounding is possible. If the former were true, then individuals with GA might benefit from avoiding high levels of alcohol consumption.
Interestingly, in the analyses of GA proximity-based progression, alcohol consumption was not associated with altered rate of GA progression towards the central macula in either men or women. Hence, alcohol consumption does not appear to be important in altering the relative tendency towards foveal sparing in GA. The contrast between these results and those for GA area-based progression emphasizes that the risk and protective factors for GA progression may different for the foveal versus extrafoveal areas.5,8 Indeed, while GA area-based progression rates differed between men and women, proximity-based progression rates did not, such that the tendency to foveal sparing does not appear related to sex.
Comparison with Literature
Multiple previous studies have examined potential relationships between alcohol consumption and the risk of progression to late AMD. Almost all of these have been observational studies, except for one Mendelian randomization study.16,17,31 Among the observational studies, both harmful and protective associations have been reported.16,17 A systematic review and meta-analysis, published in 2019, combined the results of nine prospective cohort studies by comparing the risk of progression to late AMD in those with highest versus lowest alcohol consumption, defined as 1–2 vs <1 standard drinks/day.16 No significant difference was observed, with an adjusted risk ratio of 0.98 (95% CI 0.76–1.27). However, the meta-analysis was limited by the binarization of the alcohol variable and the absence of separate analyses either for men and women or for GA and neovascular AMD. It is also worth noting that the meta-analysis combined studies from multiple countries (specifically Australia, Denmark, Holland, South Korea, and the United States), between which patterns of alcohol consumption and metabolism are thought to differ. The meta-analysis also reported results for risk of progression to early AMD, based on 10 prospective cohort studies. Interestingly, for this outcome, highest vs lowest consumption of alcohol was associated with significantly increased risk, with an adjusted risk ratio of 1.29 (95% CI 1.16–1.43).
A more recent systematic review and meta-analysis, published in 2021, performed not only categorical analyses, with more than two categories, but also dose-response analyses.17 The categories were defined as no/occasional (as reference), light (0–11 g/day), moderate (12–23 g/day), and heavy (≥ 24 g/day) alcohol consumption. For risk of progression to late AMD, in the categorical analyses, based on meta-analysis of seven prospective cohort studies, no significant associations were observed. The pooled effect estimates were 1.03 (95% CI 0.79–1.33), 1.13 (95% CI 0.83–1.55), and 0.98 (0.63–1.53) for light, moderate, and heavy consumption, respectively. Similarly, in the dose-response analyses, no significant association was observed between alcohol consumption and risk of progression to late AMD, either with assumptions of linearity or non-linearity. Direct comparison between these results and those of the current study is limited, owing to differences in the thresholds for the comparisons. However, the results appear not too dissimilar in that, in the current study, no altered risk was observed for tertile 2 and only a modest degree of decreased risk for tertile 3. As for the 2019 meta-analysis16, this 2021 meta-analysis was limited by the absence of separate analyses either for men and women or for GA and neovascular AMD.17 The meta-analysis also reported results for risk of progression to early AMD: both moderate (1.19, 95% CI 1.03–1.37) and heavy (1.24, 95% CI 1.10–1.39) alcohol consumption were significantly associated with increased risk of early AMD, in comparison to no/occasional consumption. In the dose-response analyses for early AMD, a linear dose-response relationship was observed, with an effect estimate of 1.14 (95% CI 1.08–1.21) for an increase in alcohol consumption of 10 g/day. Overall, therefore, in both meta-analyses, the relationship between alcohol consumption and AMD progression appeared to be stronger for the development of early disease than late disease. This might represent a genuine difference in the risk factors operating at different disease stages but might also relate partially to lower power in the analyses for late disease.
Finally, one previous 2-sample Mendelian randomization study has evaluated the relationship between genetically predicted alcohol consumption and advanced AMD, based on cross-sectional data on AMD status from the International AMD Genomics Consortium.31 Genetically predicted alcohol consumption was not significantly associated with altered risk of advanced AMD: the odds ratio for advanced AMD was 1.57 (95% CI 1.03–2.40) per 1-SD increase of log-transformed alcoholic drinks per week. However, in subtype analysis, higher genetically predicted alcohol consumption was significantly associated with increased risk of GA (odds ratio 2.70, 95% CI 1.48–4.94) but not of neovascular AMD (odds ratio 1.49, 95% CI 0.95–2.34). The potential advantage of the Mendelian randomization approach, providing its complex assumptions are assumed to be met, is in avoiding confounding for the assessment of causal relationships. However, importantly, the study was limited in being unable to assess for potential non-linear relationships between alcohol consumption and AMD.
Regarding GA progression, we are aware of only one previous study to evaluate possible relationships between alcohol consumption and GA area-based progression rates.18 This was a previous analysis of the AREDS2 dataset by our own research group. The study was focused on the aMedi, of which alcohol consumption is one of the nine components. However, the aMedi considers alcohol consumption as a binary variable only, with high adherence defined as consumption within a specified interval, as described above.19 In these previous analyses, alcohol consumption within the aMedi interval was strongly associated with decreased rate of GA area-based progression, at 0.239 vs 0.288 mm/year. However, the current analyses provide much greater insight, by (i) separating zero consumption from high consumption and (ii) considering men and women separately. It is now clear that the significant difference observed in the previous analyses was driven principally by the findings for men rather than women. Finally, we are not aware of any previous studies to examine potential relationships between alcohol consumption and GA proximity-based progression or foveal sparing.
Strengths and Limitations
The main strengths of this study relate to the characteristics of the AREDS2 dataset, which comprises a large number of participants and long follow-up time, permitting analyses of both risk of progression to late AMD and rate of GA progression according to alcohol consumption. Additional advantages of the dataset include its clinical trial setting, with prospective and standardized collection of imaging and clinical data at frequent fixed time-points, reading center measurement of both GA area and proximity in all cases, and detailed dietary information by validated FFQ. Importantly, this allowed the models to be adjusted for diet and smoking, whose characteristics are often associated with alcohol consumption.32–36 The study was strengthened by its openness to non-linear relationships with alcohol consumption, as well as separate analyses into associations with heavy drinking, using the pre-specified sex-specific CDC definitions.28
The limitations of the study include its observational nature, such that causality cannot be assessed, together with post hoc hypothesis generation, potential inaccuracies in alcohol consumption measurement from FFQ, and the possibility of residual or unmeasured confounding (e.g., with physical activity). No information was available on alcohol consumption earlier in life, such that, for example, participants who had never drunk alcohol were considered in the same category as former heavy drinkers who were later abstinent. In addition, data were available on overall consumption rather than temporal and cultural patterns of drinking, such as binge drinking.28,37,38 Although assessing alcohol intake by self-report from surveys or FFQ is a highly validated and feasible way to quantify drinking, this approach may underestimate consumption.39 However, potential underestimation is reported as greatest in young men and middle-aged women, and much less in elderly age groups40, as studied in the AREDS2. In addition, accuracy is thought to be improved by using beverage-specific questions (i.e., separate questions for regular beer, light beer, red wine, etc.)41, as employed in this study. Overall, although a degree of underestimation is possible for some participants in this study, this was offset by the study design, including the detailed FFQ that includes beverage-specific questions, the tertile-based approach, and the relatively large size of the study population.
Another potential limitation is the use of color fundus photography, since GA is believed to be detected earlier, and perhaps with less variability of area measurements, on fundus autofluorescence (FAF) images.42 However, previous studies have demonstrated high correlation between color fundus photography and FAF images in measuring GA area and progression.42,43 This includes previous large-scale analyses of the AREDS2 GA dataset itself.42 In these analyses, the investigators analyzed 8070 instances of FAF-color fundus photograph image pairs from 2202 AREDS2 participants, including approximately 2000 instances with GA. GA area-based progression rates were extremely similar between the two modalities, with no significant difference. Hence, this factor is unlikely to have altered the results substantially. Finally, the generalizability of the results to populations of different race, alcohol consumption patterns, or dietary characteristics are unknown. In relation to race, the large majority of the AREDS2 participants were white, so that the applicability of the results to non-white individuals is unknown.
Conclusions
In the AREDS2, the relationship between alcohol and AMD progression differs substantially between men and women. This is true for both progression to late AMD and GA area-based progression. In men, positive but moderate alcohol consumption (i.e., up to three standard drinks per week) is associated with decreased risk of progression to late AMD. By contrast, in women, associations are absent or weak. Of note, GA progression appears faster in women than men, but its relationship with alcohol is much stronger in men. In men, positive but moderate consumption is associated with slower progression, while higher consumption (including CDC-defined heavy drinking) is associated with faster progression. By contrast, in women, again, associations are absent or weak. Interestingly, alcohol consumption appears more strongly related to area-based than proximity-based progression of GA, so does not appear linked to an altered tendency to foveal sparing. Despite adjustment for multiple variables including smoking, diet, and education level, it remains difficult to distinguish with absolute confidence between causation and confounding. If causation contributed to these results, then individuals with GA might benefit from avoiding high levels of alcohol consumption. In future studies of alcohol consumption and AMD progression, it is essential for analyses to consider men and women separately and to remain open to non-linear relationships.
Financial Support
This research was supported by the Intramural Research Program of the National Eye Institute, National Institutes of Health (NIH), Department of Health and Human Services, Bethesda, Maryland, including contract HHS-N-260-2005-00007-C and ADB contract N01-EY-5-0007 for the AREDS2. Funds were generously contributed to these contracts by the following NIH institutes: Office of Dietary Supplements; National Center for Complementary and Alternative Medicine; National Institute on Aging; National Heart, Lung, and Blood Institute; National Institute of Neurological Disorders and Stroke. The sponsor and funding organization participated in the design and conduct of the study, data collection, management, analysis, and interpretation, and preparation, review, and approval of the manuscript.
Conflicts of Interest
All authors declare no support from any other organization for the submitted work.
Abbreviations
- AMD
age-related macular degeneration
- aMedi
alternative Mediterranean diet index
- AREDS2
Age-Related Eye Diseases Study 2
- CDC
Centers for Disease Control and Prevention
- CI
confidence interval
- FAF
fundus autofluorescence
- FDA
Food and Drug Administration
- FFQ
food frequency questionnaire
- GA
geographic atrophy
- RPE
retinal pigment epithelium
- SD
standard deviation
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.GBD 2019 Blindness and Vision Impairment Collaborators; Vision Loss Expert Group of the Global Burden of Disease Study. Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the Right to Sight: an analysis for the Global Burden of Disease Study. Lancet Glob Health. Feb 2021;9(2):e144–e160. doi: 10.1016/S2214-109X(20)30489-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rudnicka AR, Jarrar Z, Wormald R, Cook DG, Fletcher A, Owen CG. Age and gender variations in age-related macular degeneration prevalence in populations of European ancestry: a meta-analysis. Ophthalmology. Mar 2012;119(3):571–80. doi: 10.1016/j.ophtha.2011.09.027 [DOI] [PubMed] [Google Scholar]
- 3.Keenan TD, Agron E, Domalpally A, et al. Progression of Geographic Atrophy in Age-related Macular Degeneration: AREDS2 Report Number 16. Ophthalmology. Dec 2018;125(12):1913–1928. doi: 10.1016/j.ophtha.2018.05.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fleckenstein M, Mitchell P, Freund KB, et al. The Progression of Geographic Atrophy Secondary to Age-Related Macular Degeneration. Ophthalmology. Mar 2018;125(3):369–390. doi: 10.1016/j.ophtha.2017.08.038 [DOI] [PubMed] [Google Scholar]
- 5.Keenan TDL. Geographic Atrophy in Age-Related Macular Degeneration: A Tale of Two Stages. Ophthalmology Science. 2023;3(3)doi: 10.1016/j.xops.2023.100306 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Csaky K, Ferris F 3rd, Chew EY, Nair P, Cheetham JK, Duncan JL. Report From the NEI/FDA Endpoints Workshop on Age-Related Macular Degeneration and Inherited Retinal Diseases. Invest Ophthalmol Vis Sci. Jul 1 2017;58(9):3456–3463. doi: 10.1167/iovs.17-22339 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Shen LL, Sun M, Ahluwalia A, Park MM, Young BK, Del Priore LV. Local Progression Kinetics of Geographic Atrophy Depends Upon the Border Location. Invest Ophthalmol Vis Sci. Oct 4 2021;62(13):28. doi: 10.1167/iovs.62.13.28 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Keenan TDL, Agron E, Keane PA, et al. Oral Antioxidant and Lutein/Zeaxanthin Supplements Slow Geographic Atrophy Progression to the Fovea in Age-Related Macular Degeneration. Ophthalmology. Jul 16 2024;doi: 10.1016/j.ophtha.2024.07.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shen LL, Sun M, Ahluwalia A, et al. Relationship of Topographic Distribution of Geographic Atrophy to Visual Acuity in Nonexudative Age-Related Macular Degeneration. Ophthalmol Retina. Aug 2021;5(8):761–774. doi: 10.1016/j.oret.2020.11.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sayegh RG, Sacu S, Dunavolgyi R, et al. Geographic Atrophy and Foveal-Sparing Changes Related to Visual Acuity in Patients With Dry Age-Related Macular Degeneration Over Time. Am J Ophthalmol. Jul 2017;179:118–128. doi: 10.1016/j.ajo.2017.03.031 [DOI] [PubMed] [Google Scholar]
- 11.Schmitz-Valckenberg S, Nadal J, Fimmers R, et al. Modeling Visual Acuity in Geographic Atrophy Secondary to Age-Related Macular Degeneration. Ophthalmologica. 2016;235(4):215–24. doi: 10.1159/000445217 [DOI] [PubMed] [Google Scholar]
- 12.Chakravarthy U, Wong TY, Fletcher A, et al. Clinical risk factors for age-related macular degeneration: a systematic review and meta-analysis. BMC Ophthalmol. Dec 13 2010;10:31. doi: 10.1186/1471-2415-10-31 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hagag AM, Kaye R, Hoang V, et al. Systematic review of prognostic factors associated with progression to late age-related macular degeneration: Pinnacle study report 2. Surv Ophthalmol. Mar–Apr 2024;69(2):165–172. doi: 10.1016/j.survophthal.2023.10.010 [DOI] [PubMed] [Google Scholar]
- 14.Kai JY, Zhou M, Li DL, et al. Smoking, dietary factors and major age-related eye disorders: an umbrella review of systematic reviews and meta-analyses. Br J Ophthalmol. Dec 18 2023;108(1):51–57. doi: 10.1136/bjo-2022-322325 [DOI] [PubMed] [Google Scholar]
- 15.Keenan TD, Agron E, Mares J, et al. Adherence to the Mediterranean Diet and Progression to Late Age-Related Macular Degeneration in the Age-Related Eye Disease Studies 1 and 2. Ophthalmology. Nov 2020;127(11):1515–1528. doi: 10.1016/j.ophtha.2020.04.030 [DOI] [PubMed] [Google Scholar]
- 16.Dinu M, Pagliai G, Casini A, Sofi F. Food groups and risk of age-related macular degeneration: a systematic review with meta-analysis. Eur J Nutr. Aug 2019;58(5):2123–2143. doi: 10.1007/s00394-018-1771-5 [DOI] [PubMed] [Google Scholar]
- 17.Zhang J, Mitsuhashi T, Matsuo T, Yorifuji T, Hamada J, Liu Y. Alcohol Consumption and Age-related Macular Degeneration: A Systematic Review and Dose-response Meta-analysis. Curr Eye Res. Dec 2021;46(12):1900–1907. doi: 10.1080/02713683.2021.1942070 [DOI] [PubMed] [Google Scholar]
- 18.Agron E, Mares J, Chew EY, Keenan TDL, Group AR. Adherence to a Mediterranean Diet and Geographic Atrophy Enlargement Rate: Age-Related Eye Disease Study 2 Report 29. Ophthalmol Retina. Sep 2022;6(9):762–770. doi: 10.1016/j.oret.2022.03.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lopez-Garcia E, Rodriguez-Artalejo F, Li TY, et al. The Mediterranean-style dietary pattern and mortality among men and women with cardiovascular disease. Am J Clin Nutr. Jan 2014;99(1):172–80. doi: 10.3945/ajcn.113.068106 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kerr WC, Lui CK, Ye Y, et al. Long-term trends in beverage-specific drinking in the National Alcohol Surveys: Differences by sex, age, and race and ethnicity. Alcohol Clin Exp Res (Hoboken). Jul 2024;48(7):1322–1335. doi: 10.1111/acer.15335 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kezer CA, Simonetto DA, Shah VH. Sex Differences in Alcohol Consumption and Alcohol-Associated Liver Disease. Mayo Clin Proc. Apr 2021;96(4):1006–1016. doi: 10.1016/j.mayocp.2020.08.020 [DOI] [PubMed] [Google Scholar]
- 22.AREDS2 Research Group, Chew EY, Clemons T, et al. The Age-Related Eye Disease Study 2 (AREDS2): study design and baseline characteristics (AREDS2 report number 1). Ophthalmology. Nov 2012;119(11):2282–9. doi: 10.1016/j.ophtha.2012.05.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Danis RP, Domalpally A, Chew EY, et al. Methods and reproducibility of grading optimized digital color fundus photographs in the Age-Related Eye Disease Study 2 (AREDS2 Report Number 2). Invest Ophthalmol Vis Sci. Jul 8 2013;54(7):4548–54. doi: 10.1167/iovs.13-11804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Michaud DS, Giovannucci EL, Ascherio A, et al. Associations of plasma carotenoid concentrations and dietary intake of specific carotenoids in samples of two prospective cohort studies using a new carotenoid database. Cancer Epidemiol Biomarkers Prev. Apr 1998;7(4):283–90. [PubMed] [Google Scholar]
- 25.Rimm EB, Giovannucci EL, Stampfer MJ, Colditz GA, Litin LB, Willett WC. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am J Epidemiol. May 15 1992;135(10):1114–26; discussion 1127–36. doi: 10.1093/oxfordjournals.aje.a116211 [DOI] [PubMed] [Google Scholar]
- 26.Feuer WJ, Yehoshua Z, Gregori G, et al. Square root transformation of geographic atrophy area measurements to eliminate dependence of growth rates on baseline lesion measurements: a reanalysis of age-related eye disease study report no. 26. JAMA Ophthalmol. Jan 2013;131(1):110–1. doi: 10.1001/jamaophthalmol.2013.572 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Behning C, Fleckenstein M, Pfau M, et al. Modeling of atrophy size trajectories: variable transformation, prediction and age-of-onset estimation. BMC Med Res Methodol. Aug 17 2021;21(1):170. doi: 10.1186/s12874-021-01356-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Centers for Disease Control and Prevention. Alcohol and Public Health. Data on Excessive Drinking. Accessed 8/13/2024, https://www.cdc.gov/drinklessbeyourbest/excessivedrinking.html
- 29.Xu Y, Geldsetzer P, Manne-Goehler J, et al. The socioeconomic gradient of alcohol use: an analysis of nationally representative survey data from 55 low-income and middle-income countries. Lancet Glob Health. Sep 2022;10(9):e1268–e1280. doi: 10.1016/S2214-109X(22)00273-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Caire J, Recalde S, Velazquez-Villoria A, et al. Growth of geographic atrophy on fundus autofluorescence and polymorphisms of CFH, CFB, C3, FHR1–3, and ARMS2 in age-related macular degeneration. JAMA Ophthalmol. May 2014;132(5):528–34. doi: 10.1001/jamaophthalmol.2013.8175 [DOI] [PubMed] [Google Scholar]
- 31.Kuan V, Warwick A, Hingorani A, et al. Association of Smoking, Alcohol Consumption, Blood Pressure, Body Mass Index, and Glycemic Risk Factors With Age-Related Macular Degeneration: A Mendelian Randomization Study. JAMA Ophthalmology. 2021/12/01/ 2021;139(12):1299. doi: 10.1001/jamaophthalmol.2021.4601 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Beard E, West R, Michie S, Brown J. Association between smoking and alcohol-related behaviours: a time-series analysis of population trends in England. Addiction. Oct 2017;112(10):1832–1841. doi: 10.1111/add.13887 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Jiang N, Lee YO, Ling PM. Association between tobacco and alcohol use among young adult bar patrons: a cross-sectional study in three cities. BMC Public Health. May 24 2014;14:500. doi: 10.1186/1471-2458-14-500 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Britton M, Derrick JL, Shepherd JM, et al. Associations between alcohol consumption and smoking variables among Latinx daily smokers. Addict Behav. Feb 2021;113:106672. doi: 10.1016/j.addbeh.2020.106672 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Joseph PV, Zhou Y, Brooks B, McDuffie C, Agarwal K, Chao AM. Relationships among Alcohol Drinking Patterns, Macronutrient Composition, and Caloric Intake: National Health and Nutrition Examination Survey 2017–2018. Alcohol Alcohol. Sep 10 2022;57(5):559–565. doi: 10.1093/alcalc/agac009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Butler L, Popkin BM, Poti JM. Associations of Alcoholic Beverage Consumption with Dietary Intake, Waist Circumference, and Body Mass Index in US Adults: National Health and Nutrition Examination Survey 2003–2012. J Acad Nutr Diet. Mar 2018;118(3):409–420 e3. doi: 10.1016/j.jand.2017.09.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Jani BD, McQueenie R, Nicholl BI, et al. Association between patterns of alcohol consumption (beverage type, frequency and consumption with food) and risk of adverse health outcomes: a prospective cohort study. BMC Med. Jan 12 2021;19(1):8. doi: 10.1186/s12916-020-01878-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Probst C, Kilian C, Sanchez S, Lange S, Rehm J. The role of alcohol use and drinking patterns in socioeconomic inequalities in mortality: a systematic review. Lancet Public Health. Jun 2020;5(6):e324–e332. doi: 10.1016/S2468-2667(20)30052-9 [DOI] [PubMed] [Google Scholar]
- 39.Gilligan C, Anderson KG, Ladd BO, Yong YM, David M. Inaccuracies in survey reporting of alcohol consumption. BMC Public Health. Dec 5 2019;19(1):1639. doi: 10.1186/s12889-019-7987-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Livingston M, Callinan S. Underreporting in alcohol surveys: whose drinking is underestimated? J Stud Alcohol Drugs. Jan 2015;76(1):158–64. [PubMed] [Google Scholar]
- 41.Casswell S, Huckle T, Pledger M. Survey data need not underestimate alcohol consumption. Alcohol Clin Exp Res. Oct 2002;26(10):1561–7. doi: 10.1097/01.ALC.0000034390.38886.14 [DOI] [PubMed] [Google Scholar]
- 42.Domalpally A, Danis R, Agron E, et al. Evaluation of Geographic Atrophy from Color Photographs and Fundus Autofluorescence Images: Age-Related Eye Disease Study 2 Report Number 11. Ophthalmology. Nov 2016;123(11):2401–2407. doi: 10.1016/j.ophtha.2016.06.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Yaspan BL, Williams DF, Holz FG, et al. Targeting factor D of the alternative complement pathway reduces geographic atrophy progression secondary to age-related macular degeneration. Sci Transl Med. Jun 21 2017;9(395)doi: 10.1126/scitranslmed.aaf1443 [DOI] [PubMed] [Google Scholar]