Systematic Review Process

Eligible studies were those that reported on the prevalence and/or incidence of AMD. We searched MEDLINE (1950 onwards), EMBASE (1960 onwards), and Web of Science (1960 onwards) electronic databases; the search was updated to October 2012. A combination of text words for AMD (age-related macula$ degeneration/age related maculopathy/senile macula$ degeneration) and epidemiologic terms (inciden$/prevalen$/Population$/Survey$) were combined with related subject headings in MEDLINE and EMBASE only (in addition to MESH/subject headings for population study designs). Studies that quantified the incidence of late AMD, geographic atrophy (GA), and/or neovascular AMD (NVAMD) from a nondiseased state at baseline (in 1 or both eyes), in population-based samples of European ancestry, with detailed methods of sampling, were included. Studies were excluded if they were carried out in other racial groups, where nonspecific volunteers or those from particular professions (except for 1 study of American watermen) were included, if they were hospital audits/surveys, or where self-reported diagnoses were made or clinical diagnosis was only ascertained in those with reduced vision (as these may represent a subgroup with disease). Papers reporting incidence rates based on counting eyes rather than individuals were also excluded. In total, 74 studies on incidence and prevalence were found and reviewed in detail (by A.R.R., Z.J., and C.G.O.), from which 14 publications based on 10 population studies contained relevant incidence data.

Data Extraction

AMD incidence rates (and associated 95% confidence limits and/or standard errors) were extracted, or calculated from the number of new cases reported over time (person-years), by age and sex, if available. The reported mean (or median) age at baseline or midpoint of the age range reported was used for analysis. If the age group was specified as younger than x , older than x , or x +, then the age band was taken to be the same width as other age groups reported in the same study. The calendar year at baseline, period of follow-up, and number of participants were also recorded where available. Whether International Classification System or Wisconsin Age-Related Maculopathy Grading System (or other systems) were used was recorded, as was whether 1 eye or both eyes were examined (with or without fundus imaging). The geographic location of the study was extracted and classified into 3 continental regions (America, Australasia, Europe). We recorded whether an individual was defined as a case of AMD on the basis of disease being present in either eye/worse eye/at least 1 eye/1 or both eyes or only in 1 eye randomly selected for ocular assessment. Late AMD refers to eyes with GA and/or NVAMD. Data on other potential confounders, such as smoking, were not routinely available. Data were extracted by 3 reviewers (Z.J., A.R.R., C.G.O.) with independent extraction on a subgroup of studies. Disagreements were resolved by discussion.

Meta-analysis of Incidence Studies

Most of the studies reported cumulative incidence as a percentage; only 1 study reported rates by person-years of follow-up. Assuming incidence rates follow a Poisson distribution, a meta-analysis of incidence rates requires the number of cases of AMD along with the person-years of follow-up for each study. For the other studies either the person-years or the number of cases of AMD had to be derived from the published estimates of cumulative incidence and average duration of follow-up. If the number of new cases of AMD or person-years of follow-up were not reported, they were estimated from the reported cumulative incidence and average duration of follow-up. An assumption was made that cases of AMD occurred halfway through the follow-up period and that non-cases were followed-up for the entire duration of the study to give an approximation for person-years of follow-up for each study. Study-specific incidence rates were combined using Bayesian Poisson meta-regression, adjusting for age, year of recruitment (to examine trends in incidence over time), continent (to examine geographic variations in incidence among populations of the same ancestry), and AMD classification system, to produce mutually adjusted rate ratios. Our model took into account that some studies had incidence rates for more than 1 age group and 3 population-based studies reported incidence rates at different time points (eg, 5-, 10-, 15-year incidence). As in our previous meta-analyses of prevalence, within-study repeated measures of incidence were modeled by a Poisson multilevel Bayesian meta-regression using WinBUGS software (MRC Biostatistics Unit, Cambridge, UK). Our model simultaneously took account of the following study level confounders: average age, sex, year of survey (to examine trends in incidence over time), continent (to examine geographic variations in incidence among populations of the same ancestry), and AMD classification system, to produce mutually adjusted rate ratios.

Estimating Age-Related Macular Degeneration Incidence From Age-Related Macular Degeneration Prevalence

We obtained the prevalence of late AMD, GA, NVAMD (for men, women, and sexes combined) for each year of age from 50 to 97 years (the oldest reported age in the studies included) from our earlier meta-analysis of 31 prevalence surveys (with 51 173 participants, including 1571 cases of late AMD, 455 with GA, and 404 with NVAMD). These prevalence estimates allow for study characteristics, such as age of the sample, examination methods, and definitions of disease (internationally recognized definitions being preferred), and represent the most up-to-date meta-analysis of AMD prevalence in white populations, similar to the older white population of the US. Age-sex-specific estimates of prevalence standardized to studies using fundus imaging and either International Classification System or Wisconsin Age-Related Maculopathy Grading System were applied to 2011 American white population estimates to give the prevalence of late AMD, GA, and NVAMD and absolute number of prevalent cases by 5-year age bands in men and women separately and combined. We give the 95% credible intervals of prevalence, which represents the range of values within which the true prevalence is expected to lie with 95% probability.

Owing to the relatively small number of longitudinal studies reporting incidence by AMD subtype and by sex (especially at older ages), we also derived estimates of incidence from age-sex-specific prevalence (standardized to studies using fundus imaging and either International Classification System or Wisconsin Age-Related Maculopathy Grading System) using a method previously reported for estimating incidence of major causes of eye disease (including “senile macular degeneration”) from the Framingham Eye Study. The method assumes that (1) the duration of disease is life-long after diagnosis, since the disease is considered to be irreversible; (2) mortality risk is the same in diseased and nondiseased individuals; and (3) disease incidence and population composition (in terms of risk factors for late AMD) remain stable over time. These assumptions refer to the population at large, not the study population. Calculation of incidence requires knowledge of the population at the beginning of a given age interval (available from the US Census Bureau ), the probability of dying (available from the National Centre for Health Statistics), and the probability of AMD (estimated from prevalence) for the same age interval. We used this approach to estimate the annual incidence per 1000 by age for late AMD, GA, and NVAMD in men and women separately and combined.

Data presented on prevalence and incidence are based on “either eye” definitions (including at least 1 eye, worse eye, 1 or both eyes) using the International Classification System or Wisconsin Age-Related Maculopathy Grading System along with fundus imaging.

Systematic Review and Meta-analysis of Incidence Studies

The incidence per 1000 person-years of late AMD, GA, and NVAMD by age for each population is shown in Figure 1 . Incidence is plotted on a logarithmic scale and shows a linear increase with age. A meta-analysis of incidence studies showed that the incidence rate for late AMD triples per decade increase in age (rate ratio 2.9; 95% credible interval 2.6, 3.2, Table 2 ). There was no evidence of a trend in late AMD incidence rates over time. There was no evidence of a difference between studies using International Classification System or Wisconsin Age-Related Maculopathy Grading System ( Table 2 ). While late AMD incidence rates appeared marginally higher in the US and Australia, compared to Europe ( Table 2 ), there were only a small number of studies and credible intervals were wide (and included the null value of 1). Hence, there was no strong evidence of geographic differences in late AMD incidence. However, while the same incident rates may apply, the age-sex structure of populations in these countries are heterogeneous; hence, the number of incident cases will differ markedly between countries. Incidence rates for late AMD were 38% higher in women than in men (rate ratio 1.4, 95% credible interval 1.1, 1.8). Exclusion of 1 small study carried out on American watermen, at the margins of the inclusion criteria, made no difference to the point estimate and marginally widened the confidence interval (odds ratio 1.38, 95% credible interval 1.08, 1.83). Incidence data for GA or NVAMD separately were available for 4 populations only. These data were insufficient to provide stable estimates of incidence by year of age by AMD subtypes.

Incidence per 1000 person-years of late age-related macular degeneration (AMD), geographic atrophy (GA), and neovascular age related macular degeneration (NVAMD) by age for each population. Square symbols represent the estimated incidence per 1000 person-years of late AMD (Top), GA (Middle), and NVAMD (Bottom) at a given age for each study. Data points from the same study are joined by a straight line. The size of each symbol is inversely proportional to the standard error of the estimate of incidence. The vertical axis is plotted on a logarithmic scale.

Incidence of Age-Related Macular Degeneration in the AMERICAN White Population

Applying age-specific prevalence estimates from a recent meta-analysis to the American white population aged 50–97 years gave an overall prevalence of late AMD of 2.3% (95% credible interval 1.7%, 3.2%). The Supplemental Table (available at AJO.com ) gives estimates by age, sex, and AMD subtype. Figure 2 shows the 2 methods used to estimate annual incidence rate; A from the Bayesian meta-regression of incidence studies in populations of white European ancestry and B incidence estimated by applying age-specific prevalence to the population demographics of the American white population. The 2 approaches yield similar rates (see absolute rate differences given in Figure 2 ), with a steeper rise with age with method B (approximately quadrupling per decade; rate ratio 3.9, 95% credible interval 3.8, 4.0). Incidence estimated from age-specific prevalence yielded a tighter 95% credible interval and we were able to estimate it over a wider age range, as well as by AMD subtype and sex. Overall annual incidence in American whites estimated from prevalence is 3.5 per 1000 (95% credible interval 2.5, 4.7) for late AMD in those aged 50–97 years. The corresponding rates for GA and NVAMD are 1.9 per 1000 (95% credible interval 1.3, 2.8) and 1.8 per 1000 (95% credible interval 1.2, 2.5), respectively ( Table 3 ). This corresponds to approximately 293 000 new cases of late AMD each year (95% credible interval 207 000, 400 000), and 160 000 (95% credible interval 107 000, 234 000) and 148 000 (95% credible interval 103 000, 207 000) for GA and NVAMD, respectively. The estimated total number of new cases rises rapidly from age 50 years to the mid-80s and then begins to drop ( Figure 3 ). In those aged 65 years and older the annual incidence rate for late AMD is 7.8 per 1000 (95% credible interval 5.5, 10.6), 4.3 per 1000 for GA (95% credible interval 2.9, 6.2), and 3.9 per 1000 for NVAMD (95% credible interval 2.7, 5.5). In those aged 80 years and older the annual incidence rates per 1000 are 19.4 (95% credible interval 13.8, 26.2), 10.9 (95% credible interval 7.2, 16.1), and 9.9 (95% credible interval 6.8, 14.0), respectively. Under 70 years of age the number of incident cases is similar in men and women. Beyond 70 years of age the number of new cases of AMD is consistently higher in women than in men ( Table 3 , Figure 3 ). For all age groups (except the youngest), women have a slightly higher annual incidence rate of late AMD (4 per 1000, 95% credible interval 2.3, 6.6) than men (2.4 per 1000, 95% credible interval 1.4, 4.2; Table 3 ). Overall, women have a marginally higher rate of GA than men, but within each 5-year age group incidence rates in Table 3 for GA appear similar in men and women and this is partly owing to rounding. In those aged 65 years and older the incidence rates of NVAMD within each 5-year age group are higher in women than in men.

Annual incidence per 1000 of late age-related macular degeneration against age from a meta-analysis of incidence studies in populations of white European ancestry (solid circles) and as estimated from age-specific prevalence (open squares) applied to the demographics of the American white population. Incidence on the vertical axis is on a logarithmic scale.

Table 3

Estimated Number of New Cases Each Year and Average Annual Incidence per 1000 of Age-Related Macular Degeneration (Late Age-Related Macular Degeneration, Geographic Atrophy, Neovascular Age-Related Macular Degeneration) by 5-Year Age Groups for Men and Women in the White American Population

Age Group (Years)	Number of New Cases per Year in 1000s (95% Cri)			Estimated Annual Incidence per 1000 (95% Cri)
	Late AMD	GA	NVAMD	Late AMD	GA	NVAMD
Men
50–54	1.4 (0.8, 2.6)	0.9 (0.5, 1.6)	0.8 (0.5, 1.5)	0.2 (0.1, 0.3)	0.1 (0.1, 0.2)	0.1 (0.1, 0.2)
55–59	2.5 (1.4, 4.6)	1.6 (1.0, 2.7)	1.4 (0.8, 2.5)	0.3 (0.2, 0.6)	0.2 (0.1, 0.3)	0.2 (0.1, 0.3)
60–64	4.4 (2.4, 8.0)	2.8 (1.7, 4.6)	2.5 (1.5, 4.2)	0.6 (0.3, 1.1)	0.4 (0.2, 0.7)	0.4 (0.2, 0.6)
65–69	6.4 (3.5, 11.5)	4.1 (2.6, 6.5)	3.5 (2.1, 5.8)	1.3 (0.7, 2.3)	0.8 (0.5, 1.3)	0.7 (0.4, 1.1)
70–74	9.0 (5.1, 16.4)	5.8 (3.7, 9.2)	4.8 (2.9, 8.2)	2.5 (1.4, 4.5)	1.6 (1.0, 2.5)	1.3 (0.8, 2.3)
75–79	13.3 (7.5, 23.7)	8.6 (5.4, 13.8)	6.9 (4.1, 12.0)	4.9 (2.8, 8.8)	3.2 (2.0, 5.1)	2.6 (1.5, 4.4)
80–84	18.8 (10.7, 32.7)	12.3 (7.6, 20.1)	9.7 (5.7, 17.0)	9.5 (5.4, 16.6)	6.2 (3.9, 10.2)	4.9 (2.9, 8.6)
85–89	20.5 (11.8, 34.6)	13.7 (8.2, 22.6)	10.6 (6.2, 18.7)	17.7 (10.2, 29.8)	11.8 (7.1, 19.5)	9.1 (5.3, 16.1)
90+	19.5 (11.9, 30.1)	13.6 (8.0, 22.4)	10.5 (6.1, 18.3)	33.4 (20.3, 51.5)	23.3 (13.7, 38.3)	18.0 (10.4, 31.4)
All ages	95.8 (55.1, 164.2)	63.3 (38.7, 103.5)	50.7 (29.8, 88.0)	2.4 (1.4, 4.2)	1.6 (1.0, 2.6)	1.3 (0.8, 2.2)
Women
50–54	1.6 (0.9, 3.0)	0.9 (0.5, 1.6)	1.0 (0.6, 1.8)	0.2 (0.1, 0.3)	0.1 (0.1, 0.2)	0.1 (0.1, 0.2)
55–59	3.0 (1.6, 5.4)	1.7 (1.0, 2.8)	1.8 (1.1, 3.2)	0.4 (0.2, 0.6)	0.2 (0.1, 0.3)	0.2 (0.1, 0.4)
60–64	5.3 (2.9, 9.8)	3.0 (1.8, 4.9)	3.3 (1.9, 5.5)	0.7 (0.4, 1.3)	0.4 (0.2, 0.7)	0.4 (0.3, 0.7)
65–69	8.0 (4.5, 14.6)	4.5 (2.9, 7.3)	4.8 (2.9, 8.0)	1.4 (0.8, 2.6)	0.8 (0.5, 1.3)	0.8 (0.5, 1.4)
70–74	12.2 (6.8, 21.9)	6.9 (4.4, 10.9)	7.0 (4.3, 12.0)	2.8 (1.6, 5.1)	1.6 (1.0, 2.5)	1.6 (1.0, 2.8)
75–79	19.7 (11.2, 35.0)	11.2 (7.1, 17.9)	11.2 (6.8, 19.2)	5.6 (3.2, 9.9)	3.2 (2.0, 5.1)	3.2 (1.9, 5.4)
80–84	32.6 (18.5, 55.7)	18.7 (11.7, 30.5)	18.2 (10.9, 31.5)	10.8 (6.1, 18.5)	6.2 (3.9, 10.1)	6.0 (3.6, 10.5)
85–89	43.5 (25.3, 71.7)	25.8 (15.8, 42.6)	24.5 (14.5, 42.3)	19.9 (11.6, 32.8)	11.8 (7.2, 19.5)	11.2 (6.6, 19.3)
90+	53.5 (33.2, 80.9)	34.1 (20.7, 55.3)	31.7 (18.9, 53.5)	37.4 (23.2, 56.5)	23.8 (14.5, 38.6)	22.2 (13.2, 37.3)
All ages	179 (105, 298)	107 (66.0, 174)	104 (61.9, 177)	4.0 (2.3, 6.6)	2.4 (1.5, 3.8)	2.3 (1.4, 3.9)
Men and women
50–54	3.0 (2.0, 4.3)	1.5 (0.9, 2.4)	1.6 (1.0, 2.3)	0.2 (0.1, 0.2)	0.1 (0.1, 0.1)	0.1 (0.1, 0.1)
55–59	5.5 (3.8, 7.9)	2.8 (1.8, 4.2)	2.8 (1.9, 4.0)	0.3 (0.2, 0.5)	0.2 (0.1, 0.3)	0.2 (0.1, 0.2)
60–64	10.1 (6.9, 14.1)	5.1 (3.4, 7.3)	5.1 (3.5, 7.0)	0.7 (0.5, 1.0)	0.3 (0.2, 0.5)	0.3 (0.2, 0.5)
65–69	15.1 (10.5, 21.2)	7.7 (5.2, 10.8)	7.5 (5.3, 10.1)	1.4 (1.0, 2.0)	0.7 (0.5, 1.0)	0.7 (0.5, 0.9)
70–74	22.7 (15.8, 31.6)	11.6 (8.1, 16.4)	11.1 (8.0, 14.9)	2.9 (2.0, 4.0)	1.5 (1.0, 2.1)	1.4 (1.0, 1.9)
75–79	35.8 (24.9, 49.7)	18.5 (12.8, 26.4)	17.4 (12.4, 23.7)	5.7 (4.0, 8.0)	3.0 (2.1, 4.2)	2.8 (2.0, 3.8)
80–84	56.2 (39.3, 77.8)	29.7 (20.1, 43.2)	27.5 (19.4, 38.0)	11.3 (7.9, 15.6)	6.0 (4.0, 8.7)	5.5 (3.9, 7.6)
85–89	70.4 (49.6, 95.7)	38.7 (25.8, 57.3)	35.2 (24.3, 50.1)	21.0 (14.8, 28.6)	11.6 (7.7, 17.1)	10.5 (7.3, 15.0)
90+	74.1 (54.2, 97.5)	44.3 (28.8, 65.9)	39.9 (26.9, 57.0)	36.7 (26.9, 48.4)	22.0 (14.3, 32.7)	19.8 (13.4, 28.3)
All ages	293 (207, 400)	160 (107, 234)	148 (103, 207)	3.5 (2.5, 4.7)	1.9 (1.3, 2.8)	1.8 (1.2, 2.5)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. The Importance of A Priori Statistical Planning in Controlled Clinical Trials
2. Quantitative Trait Locus Analysis of SIX1-SIX6With Retinal Nerve Fiber Layer Thickness in Individuals of European Descent
3. Comparison of Contrast Sensitivity and Through Focus in Small-Aperture Inlay, Accommodating Intraocular Lens, or Multifocal Intraocular Lens Subjects
4. Reply

Stay updated, free articles. Join our Telegram channel

Join