Five-Year Incidence of Age-Related Macular Degeneration and Its Risk Factors in Adult Chinese Population: The Tongren Health Care Study

Abstract

Purpose

To examine the 5-year incidence of age-related macular degeneration (AMD) and its associated factors in an adult Chinese population.

Methods

The Tongren Health Care Study included individuals attending regular health care check-up examinations in the Beijing Tongren Hospital. Baseline examinations were performed from 2014 to 2015, with 5-year follow-up examinations conducted between 2019 and 2020. Fundus photographs were graded according to the Beckman Initiative guidelines.

Results

A total of 5658 participants with gradable photographs at both examinations were included in the study, comprising 58.0% women, with a mean age of 54.9 ± 11.0 years. The 5-year incidence of any, early, intermediate, and late AMD were 6.1% (95% confidence interval [CI], 5.5%–6.8%), 5.0% (95% CI, 4.4%–5.6%), 3.4% (95% CI, 2.9%–3.9%), and 0.3% (95% CI, 0.2%–0.4%), respectively. In multivariate analysis, incident early AMD was associated with older age (P < 0.001; odds ratio [OR], 1.04; 95% CI, 1.02–1.06), female sex (P = 0.011; OR, 1.42; 95% CI, 1.08–1.86), and a higher estimated glomerular filtration rate (P = 0.020; OR, 1.15; 95% CI, 1.02–1.30), whereas having diabetes was a protective factor (P = 0.019; OR, 0.61; 95% CI, 0.41–0.92). Incident intermediate AMD was associated with older age (P < 0.001; OR, 1.05; 95% CI, 1.04–1.07), a higher high-density lipoprotein cholesterol level (P < 0.001; OR, 1.97; 95% CI, 1.38–2.83) and a lower triglyceride level (P = 0.008; OR, 0.77; 95% CI, 0.64–0.93).

Conclusions

A higher estimated glomerular filtration rate level was a risk factor for incident early AMD. A higher high-density lipoprotein cholesterol level and lower triglyceride level were risk factors for incident intermediate AMD. This finding may point to the role of renal circulation and lipid metabolism in incident AMD.

Translational Relevance

This community-based longitudinal study may provide a valuable understanding of AMD and its associated factors for targeted prevention and management strategies.

Introduction

Age-related macular degeneration (AMD) is one of the most common causes of irreversible vision impairment and blindness worldwide.¹ Previous longitudinal population-based studies have revealed interethnic and geo-epidemiological differences in the incidence of AMD.^2–11 The Beijing Eye Study and the Handan Eye Study showed that the incidence of AMD was lower in Chinese populations than in other ethnic groups.^6,10 These studies also revealed that older age, male sex, and short axial length were risk factors for early AMD in Chinese, whereas other systemic parameters, such as arterial hypertension, were not associated with the prevalence or incidence of AMD. More information is needed on which systemic parameters may be related to the incidence and prevalence of AMD and may as risk factors influence the course of the disease. In this study, we have aimed to investigate the 5-year incidence of AMD in a Chinese population and its associated systemic risk factors.

In this study, AMD was graded into early, intermediate, and late stages according to the guidelines proposed by the Beckman Initiative.¹² Previous epidemiological research on AMD were usually based on the Wisconsin Age-related Maculopathy Grading System, which divided AMD into early and late stages.^13,14 The biggest difference between the two grading systems is that hard drusen with the diameter of 63 to 125 µm is defined as early AMD in the Beckman system, although it is not considered to be AMD in the Wisconsin system. In the Wisconsin system, the drusen needs to be soft and indistinct for early AMD. Analyses of eyes with medium drusen (63–125 m) provided evidence that as soon as drusen of this size were present, there was evidence of increasing risk of AMD progression, regardless of the morphology of the drusen.¹² The Beckman grading system enables us to detect more early AMD cases that may progress into later stages of AMD. This factor is important as additional forms of therapy for early stages of the disease are discovered. As far as we know, this longitudinal study is the first using the Beckman grading system. The other studies based on this grading system were all cross-sectional studies, which only provided the prevalence of AMD, including the Yangxi Eye Study, the Australian National Eye Health Survey, the Hainan Study and the Ural Eye and Medical Study.^15–18

Methods

Population

The longitudinal Tongren Health Care Study included individuals who attended regular health care check-up examinations in the Beijing Tongren Hospital. Baseline examination was performed from 2014 to 2015, with a 5-year follow-up examination conducted in the period from 2019 to 2020. The demographic characteristics of the study participants have been reported in detail elsewhere.¹⁹ Briefly, the population of the Tongren Health Care Study initially included 11,806 participants; 10,342 participants successfully followed up 5 years later, with a corresponding rate of 87.6%. The mean age of participants of the Tongren Health Care Study was 44.9 years at baseline. Out of these, those with an age of ≥40 years were included in the present study. The study protocol was approved by the Beijing Tongren Hospital Ethics Committee (No. TRECKY2020-066) and was performed in accordance with the guidelines of the Declaration of Helsinki.

Data Collection

Trained physicians asked each participant questions from a standardized questionnaire, which included demographic data and information about the past medical history. The physical examination included the measurement of body height, body weight, blood pressure and heart rate. The body mass index was then calculated.

Blood samples were taken under fasting conditions. Biochemical examinations and full blood cell count were performed using the Hematology Analyzer Coulter LH 780 (Beckman Coulter Inc, Miami, FL, USA) within 4 hours after the blood sample collection. The serum concentration of glucose, creatinine and blood lipids (triglyceride, total cholesterol, high-density lipoprotein [HDL] cholesterol, low-density lipoprotein cholesterol) was measured. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation based on the serum creatinine concentration and body weight.²⁰

Arterial hypertension was defined as a systolic blood pressure of ≥140 mm Hg, a diastolic blood pressure of ≥90 mm Hg, a history of physician-based diagnosis of hypertension, or the use of antihypertensive drugs. Diabetes mellitus was characterized by a fasting blood glucose concentration of ≥7.0mmol/L, a history of previously diagnosed diabetes or the use of hypoglycemic agents including insulin. Chronic kidney disease (CKD) was defined by an eGFR of <60 mL/min/1.73 m².

Ophthalmic Examination and AMD Grading

The ophthalmic examinations included measurement of uncorrected visual acuity, noncontact tonometry, slit lamp examination of the anterior segment, and fundus photography, with 45° macula-centered digital retinal photographs taken from each individual. The retinal photographs were obtained by using two digital retinal cameras (Topcon TRC-NW7SF, Topcon, Tokyo, Japan or Canon CR6-45N, Canon, Tokyo, Japan).

Fundus lesions within a circle with a diameter of 6 mm and centered on the fovea were accessed. Retinal images were deemed to be gradable for AMD if two-thirds of the macular area was visible. AMD was graded according to the guidelines proposed by the Beckman Initiative.¹² Early AMD was defined as the presence of medium-sized drusen (diameter of 63–125 µm) without pigmentary abnormalities related to AMD. Intermediate AMD was characterized by large drusen (diameter of >125 µm) or pigmentary abnormalities associated with at least medium-sized drusen (diameter of 63–125 µm). Late AMD was defined as the presence of neovascular AMD or geographic atrophy. Neovascular AMD included serous or hemorrhagic detachment of the retinal pigment epithelium (RPE) or sensory retina, subretinal or sub-RPE hemorrhages, and subretinal fibrous scars. Geographic atrophy was defined as a discrete circular area of depigmentation of the RPE with a diameter of ≥175 µm. AMD was graded by individual, based on the eye with the more severe stage of AMD. A trained grader (Y.C.) initially graded all photographs in a masked manner. This was followed by a side-by-side grading of the baseline and follow-up photographs when lesions indicating AMD were found at either examination, without the grader knowing which slide was the more recent. For the diagnosis of late AMD and in cases of doubt, the images were reassessed for the specific lesion by two senior graders (Y.M. and C.C.X.). The assessment of the interobserver variability in assessing the photographs of 100 randomly selected participants showed an excellent Cohen kappa value of 0.850 (P < 0.001) for the presence of any AMD.

Incident late AMD was defined by the appearance of neovascular AMD or geographic atrophy involving the macular area in either eye of persons at the follow-up examination, while late AMD was not present at the baseline examination. Incident intermediate AMD was defined by the appearance of either large drusen or the copresence of both medium-sized drusen and pigmentary abnormalities in either eye of persons at the follow-up examination, while late AMD or intermediate AMD was not present at baseline. Incident early AMD was defined by the appearance of medium-sized drusen in either eye of persons at the follow-up examination, while early, intermediate or late AMD was not present at baseline.

Persons without late AMD in either eye at baseline were counted as at risk of incident late AMD. Persons without intermediate or late AMD in either eye at baseline were counted as at risk of incident intermediate AMD. Those without early, intermediate, or late AMD in either eye at baseline were counted as at risk of incident early AMD and incident any AMD.

Statistical Analysis

Statistical analysis was conducted using a statistical software program (SPSS 27.0 for Windows (SPSS Inc., Chicago, IL, USA). Numerical variables were presented as mean ± standard deviation and the Student t test was used to compare differences between groups. Categorical variables were described by frequencies and percentages and differences were tested by the χ² test. The incidence of AMD was presented as percentage with the 95% confidence interval (CI). Logistic regression analysis was used to investigate associations between incident AMD and risk factors. Odds ratios (OR) and their 95% CIs were calculated. In the first step, univariate analysis was performed. In the second step, age-adjusted analysis was performed. In the third step, multivariate analysis was performed, which included all the variables for which the P value in the age-adjusted analysis was <0.05. Two-sided P values of <0.05 were considered as statistically significant in all analyses.

Results

Study Cohort

Among 10,342 individuals of the whole longitudinal Tongren Health Care Study, 5871 individuals (57.9% women) with an age of ≥40 years at baseline were included in the present study. Their mean age was 55.4 ± 11.0 years. After excluding those with ungradable photographs either at baseline or in the follow-up study, 5658 individuals were included in the final analysis (58.0% women) with a mean age of 54.9 ± 11.0 years (range, 40–92 years). As compared with the analyzed participants, those who had ungradable photographs were at baseline more likely to be older, had a higher body mass index, higher systolic blood pressure, higher blood glucose concentration, lower eGFR, lower blood lipids concentration, and with higher prevalence of arterial hypertension, diabetes mellitus, and CKD (all P < 0.05) (Supplementary Table S1).

Incidence of AMD

The 5-year incidence of AMD, including any AMD, early, intermediate and late AMD, was 6.1% (95% CI, 5.5%–6.8%), 5.0% (95% CI, 4.4%%–5.6%), 3.4% (95% CI, 2.9%%–3.9%), and 0.3% (95% CI, 0.2%%–0.4%), respectively. Age- and sex-stratified data are shown in Table 1.

Table 1.

Five-Year Incidence of Age-Related Macular Degeneration (AMD) Stratified by Sex and Age Group

	Any AMD			Early AMD			Intermediate AMD			Late AMD
	Incident Cases/N.	Incidence (%, 95% CI)	OR (95% CI)	Incident Cases/No.	Incidence (%, 95% CI)	OR (95%CI)	Incident Cases/No.	Incidence (%, 95% CI)	OR (95%CI)	Incident Cases/No.	Incidence (%, 95% CI)	OR (95% CI)
Age groups of men (years)
40–49	20/757	2.6 (1.7–4.1)	1.00 (Ref)	19/757	2.5 (1.6–3.9)	1.00 (Ref)	6/786	0.8 (0.3–1.7)	1.00 (Ref)	0/794	0.0 (0.0–0.0)	/
50–59	41/752	5.5 (4.1–7.3)	1.96 (1.14–3.37)	36/752	4.8 (3.5–6.6)	1.80 (1.02–3.16)	26/816	3.2 (2.2–4.7)	4.13 (1.69–10.10)	0/852	0.0 (0.0–0.0)	/
60–69	28/385	7.3 (5.1–10.4)	2.37 (1.32–4.25)	22/385	5.7 (3.8–8.6)	1.94 (1.04–3.62)	17/441	3.9 (2.4–6.1)	4.77 (1.87–12.17)	5/486	1.0 (0.4–2.5)	/
≥70	16/199	8.0 (5.0–12.9)	2.71 (1.38–5.31)	7/199	3.5 (1.7–7.3)	1.20 (0.45–2.89)	18/219	8.2 (5.3–12.8)	10.43 (4.09–26.58)	2/240	0.8 (0.2–3.3)	/
Total	105/2093	5.0 (4.2–6.0)		84/2093	4.0 (3.3–5.0)		67/2262	3.0 (2.3–3.8)		7/2372	0.3 (0.1–0.6)
P for trend			0.001			0.226			<0.001			0.005
Age groups of women (years)
40–49	28/1131	2.5 (1.7–3.6)	1.00 (Ref)	23/1131	2.0 (1.4–3.1)	1.00 (Ref)	10/1161	0.9 (0.5–1.6)	1.00 (Ref)	0/1167	0.0 (0.0–0.0)	/
50–59	93/984	9.5 (7.8–11.5)	3.66 (2.38–5.62)	77/984	7.8 (6.3–9.7)	3.64 (2.27–5.85)	53/1087	4.9 (3.8–6.3)	5.70 (2.89–11.27)	0/1127	0.0 (0.0–0.0)	/
60–69	49/504	9.7 (7.5–12.7)	3.49 (2.17–5.61)	43/504	8.5 (6.4–11.4)	3.70 (2.21–6.21)	28/584	4.8 (3.3–6.9)	5.47 (2.64–11.33)	1/620	0.2 (0.0–1.1)	/
≥70	30/258	11.6 (8.3–16.3)	3.69 (2.17–6.27)	20/258	7.8 (5.1–11.8)	2.92 (1.58–5.38)	24/313	7.7 (5.2–11.3)	8.25 (3.91–17.42)	6/359	1.7 (0.8–3.7)	/
Total	200/2871	7.0 (6.1–8.0)		163/2871	5.7 (4.9–6.6)		115/3145	3.7 (3.1–4.4)		7/3273	0.2 (0.1–0.5)
P for trend			<0.001			<0.001			<0.001			0.006
Age groups of men and women (years)
40–49	48/1888	2.5 (1.9–3.4)	1.00 (Ref)	42/1888	2.2 (1.7–3.0)	1.00 (Ref)	16/1947	0.8 (0.5–1.3)	1.00 (Ref)	0/1961	0.0 (0.0–0.0)	/
50–59	134/1736	7.7 (6.6–9.1)	2.89 (2.07–4.05)	113/1736	6.5 (5.5–7.8)	2.77 (1.93–3.96)	78/1903	4.1 (3.3–5.1)	5.05 (2.94–8.68)	0/1979	0.0 (0.0–0.0)	/
60–69	77/889	8.7 (7.0–10.7)	2.98 (2.06–4.31)	65/889	7.3 (5.8–9.2)	2.85 (1.92–4.24)	45/1025	4.4 (3.3–5.8)	5.16 (2.90–9.17)	6/1106	0.5 (0.2–1.2)	/
≥70	46/457	10.1 (7.7–13.2)	3.28 (2.16–4.96)	27/457	5.9 (4.1–8.5)	2.13 (1.30–3.49)	42/532	7.9 (5.9–10.6)	9.06 (5.06–16.24)	8/599	1.3 (0.7–2.7)	/
Total	305/4970	6.1 (5.5–6.8)		247/4970	5.0 (4.4–5.6)		182/5407	3.4 (2.9–3.9)		14/5645	0.3 (0.2–0.4)
P for trend			<0.001			<0.001			<0.001			<0.001

CI, confidence interval; N, number at risk at baseline; OR, odds ratio.

A P value for trend of ≤0.05 indicates a significant increase in the incidence of AMD with age.

As the statistically most powerful risk factor for incident AMD, older age was related with a higher incidence of any AMD or every stage of AMD (all P for trend < 0.001) (Table 1; Figure 1). As compared with participants aged 40 to 49 years, those aged 50 to 59 years, 60 to 69 years, and ≥70 years had an increased risk for developing AMD of 2.89 fold (OR, 2.89; 95% CI, 2.07%–4.05), 2.98 fold (OR, 2.98; 95% CI, 2.06%–4.31), and 3.28 fold (OR, 3.28; 95% CI, 2.16%–4.96), respectively. If only incident early AMD was considered, the increased risk for those aged 50 to 59 years, 60 to 69 years, and ≥70 years was 2.77 fold, 2.85 fold, and 2.13 fold, respectively. If only incident intermediate AMD was considered, the increased risk for those aged 50 to 59 years, 60 to 69 years, and ≥70 years was 5.05 fold, 5.16 fold, and 9.06 fold, respectively. For late AMD, the number of incident cases was not sufficient for a meaningful statistical analysis.

Figure 1.

Distribution of the 5-year incidence of age-related macular degeneration (AMD) in the Tongren Health Care Study. Figures are shown as the incidence and its 95% confidence interval.

Risk Factors

An age-adjusted analysis indicated that the incidence of early AMD was higher in females than in males, and in those individuals without diabetes mellitus, with higher eGFR, higher HDL cholesterol level, and lower blood monocyte count (all age-adjusted P < 0.05) (Table 2). In multivariate regression analysis, a higher tendency to develop early AMD was associated with older age (P < 0.001; OR, 1.04), female sex (P = 0.011; OR, 1.42), higher eGFR level (P = 0.020; OR, 1.15), and lower prevalence of diabetes mellitus (P = 0.019; OR, 0.61) (Table 3; Figure 2).

Table 2.

Risk Factors Associated With the Incidence of Age-related Macular Degeneration (AMD) in Univariate and Age-adjusted Regression Analyses, Respectively

	Any AMD				Early AMD				Intermediate AMD
	Univariate Analysis		Age-Adjusted Analysis		Univariate Analysis		Age-Adjusted Analysis		Univariate Analysis		Age-Adjusted Analysis
Risk Factors	P Value	OR (95% CI)	P Value	OR (95% CI)	P Value	OR (95% CI)	P Value	OR (95% CI)	P Value	OR (95% CI)	P Value	OR (95% CI)
Age (years)	<0.001	1.03 (1.02–1.04)	–	–	<0.001	1.02 (1.01–1.03)	–	–	<0.001	1.05 (1.04–1.07)	–	–
Female sex	0.006	1.41 (1.10–1.79)	0.004	1.43 (1.12–1.82)	0.009	1.43 (1.09–1.87)	0.008	1.44 (1.10–1.89)	0.147	1.25 (0.92–1.70)	0.117	1.28 (0.94–1.74)
Body weight (kg)	0.246	0.99 (0.98–1.00)	0.539	1.00 (0.99–1.01)	0.435	1.00 (0.98–1.01)	0.647	1.00 (0.99–1.01)	0.006	0.98 (0.97–0.99)	0.043	0.99 (0.97–1.00)
Body height (cm)	0.012	0.98 (0.97–1.00)	0.247	0.99 (0.98–1.01)	0.053	0.99 (0.97–1.00)	0.264	0.99 (0.98–1.01)	0.005	0.97 (0.96–0.99)	0.324	0.99 (0.97–1.01)
Body mass index (kg/m2)	0.623	1.01 (0.97–1.05)	0.865	1.00 (0.97–1.04)	0.590	1.01 (0.97–1.05)	0.746	1.01 (0.97–1.05)	0.150	0.97 (0.92–1.01)	0.074	0.96 (0.91–1.00)
Hypertension	<0.001	1.55 (1.23–1.96)	0.112	1.23 (0.95–1.58)	0.016	1.38 (1.06–1.78)	0.249	1.18 (0.89–1.55)	0.014	1.45 (1.08–1.94)	0.623	0.92 (0.67–1.27)
Diabetes	0.769	0.95 (0.69–1.32)	0.109	0.76 (0.54–1.06)	0.100	0.71 (0.48–1.07)	0.015	0.60 (0.40–0.91)	0.107	1.36 (0.94–1.99)	0.863	0.97 (0.66–1.42)
Chronic kidney disease	0.593	1.00 (0.99–1.01)	0.668	1.00 (0.98–1.02)	0.825	0.93 (0.50–1.73)	0.271	0.70 (0.37–1.32)	0.688	1.00 (0.98–1.01)	0.744	1.00 (0.97–1.02)
Systolic blood pressure (mm Hg)	0.052	1.01 (1.00–1.01)	0.908	1.00 (0.99–1.01)	0.259	1.01 (1.00–1.01)	1.000	1.00 (0.99–1.01)	0.017	1.01 (1.00–1.02)	0.968	1.00 (0.99–1.01)
Diastolic blood pressure (mm Hg)	0.394	1.01 (0.99–1.02)	0.323	1.01 (1.00–1.02)	0.281	1.01 (1.00–1.02)	0.257	1.01 (1.00–1.02)	0.247	0.99 (0.98–1.01)	0.378	0.99 (0.98–1.01)
Blood glucose mmol/L)	0.821	0.99 (0.91–1.08)	0.224	0.94 (0.86–1.04)	0.241	0.94 (0.84–1.04)	0.073	0.90 (0.80–1.01)	0.959	1.00 (0.90–1.11)	0.224	0.93 (0.82–1.05)
eGFR (10 mL/min per 1.73 mm2)	0.024	0.92 (0.85–0.99)	0.023	1.13 (1.02–1.26)	0.487	0.97 (0.89–1.06)	0.023	1.15 (1.02–1.29)	<0.001	0.81 (0.74–0.89)	0.277	1.08 (0.94–1.23)
Blood monocyte count (109/L)	0.185	0.56 (0.24–1.32)	0.100	0.49 (0.21–1.15)	0.193	0.93 (0.83–1.04)	0.021	0.31 (0.12–0.84)	0.044	0.31 (0.10–0.97)	0.013	0.23 (0.07–0.74)
Hb (g/L)	0.216	1.00 (0.99–1.00)	0.286	1.00 (0.99–1.00)	0.379	1.00 (0.99–1.01)	0.427	1.00 (0.99–1.01)	0.044	0.99 (0.98–1.00)	0.088	0.99 (0.98–1.00)
Triglycerides (mmol/L)	0.153	0.92 (0.82–1.03)	0.104	0.91 (0.80–1.02)	0.208	0.92 (0.81–1.05)	0.164	0.91 (0.80–1.04)	0.054	0.85 (0.72–1.00)	0.028	0.82 (0.68–0.98)
Total cholesterol (mmol/L)	0.118	1.10 (0.98–1.24)	0.103	1.10 (0.98–1.24)	0.389	1.06 (0.93–1.21)	0.377	1.06 (0.93–1.21)	0.065	1.15 (0.99–1.33)	0.040	1.17 (1.01–1.35)
LDL cholesterol (mmol/L)	0.135	1.10 (0.97–1.26)	0.106	1.11 (0.98–1.27)	0.437	1.06 (0.92–1.23)	0.407	1.06 (0.92–1.23)	0.232	1.11 (0.94–1.30)	0.140	1.13 (0.96–1.33)
HDL cholesterol (mmol/L)	0.015	1.44 (1.07–1.94)	0.018	1.43 (1.06–1.91)	0.041	1.41 (1.02–1.95)	0.044	1.40 (1.01–1.93)	<0.001	2.05 (1.43–2.96)	<0.001	1.97 (1.38–2.83)

CI, confidence interval; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; OR, odds ratio.

Variables with a P value of ≤0.05 in the age-adjusted analysis are presented in bold type.

Table 3.

Risk Factors Associated With the Incidence of Any, Early, Intermediate, and Late Age-Related Macular Degeneration (AMD) in Multivariate Regression Analysis

	P Value	OR (95% CI)
Any AMD
Age (years)	<0.001	1.05 (1.03–1.06)
Female sex	0.004	1.43 (1.12–1.83)
eGFR (10ml/min per 1.73 mm2)	0.022	1.13 (1.02–1.26)
Early AMD
Age (years)	<0.001	1.04 (1.02–1.06)
Female sex	0.011	1.42 (1.08–1.86)
Diabetes	0.019	0.61 (0.41–0.92)
eGFR (10ml/min per 1.73 mm2)	0.020	1.15 (1.02–1.30)
Intermediate AMD
Age (years)	<0.001	1.05 (1.04–1.07)
HDL cholesterol (mmol/L)	<0.001	1.97 (1.38–2.83)
Triglycerides (mmol/L)	0.008	0.77 (0.64–0.93)
Late AMD
Age (years)	<0.001	1.13 (1.07–1.18)

CI, confidence interval; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; OR, odds ratio.

Figure 2.

Associations between the 5-year incidence of age-related macular degeneration (AMD) and systemic parameters after multivariate regression analysis. (A) Associations between the 5-year incidence of early AMD and systemic parameters including estimated glomerular filtration rate (eGFR), diabetes, female sex, and age. Continuous variables were converted into strata variables. eGFR was stratified into three levels: 60 to 75, 75 to 90 and >90 mL/min/1.73 m². Age was stratified into four levels: 40 to 49, 50 to 59, 60 to 69, and ≥70 years. (B) Associations between the 5-year incidence of intermediate AMD and systemic parameters, including high-density lipoprotein (HDL) cholesterol, triglycerides, and age. Continuous variables were converted into strata variables. HDL cholesterol was stratified into three levels: <1.04, 1.04 to 1.6, and >1.6 mmol/L. Triglyceride was stratified into three levels: <1.7, 1.7 to 2.3, and >2.3 mmol/L. Age was stratified into four levels: 40 to 49, 50 to 59, 60 to 69, and ≥70 years.

The incidence of intermediate AMD was higher in those individuals with lower body weight, lower blood monocyte count, lower serum concentration of triglycerides, and higher total cholesterol and HDL cholesterol levels (all age-adjusted P < 0.05) (Table 2). In multivariate regression analysis, participants with an older age (P < 0.001; OR, 1.05), higher HDL cholesterol level (P < 0.001, OR, 1.97) and lower triglyceride level (P = 0.008; OR, 0.77) were more likely to develop intermediate AMD within the study period of 5 years (Table 3; Figure 2).

Similarly, the incidence of any AMD was related with older age (P < 0.001; OR, 1.05), female sex (P = 0.004; OR, 1.43), and higher eGFR level (P = 0.022; OR, 1.13) (multivariate analysis) (Table 3).

The eGFR ranged from 4 to 127 mL/min/1.73 m² with a mean value of 90.5 ± 15.2 mL/min/1.73 m² in the study population. After controlling for age, gender and diabetes, there was a 1.15-fold increase in the probability for incident early AMD with every 10 mL/min/1.73 m² increase in eGFR (P = 0.020; OR, 1.15; 95% CI, 1.02–1.30) (Table 3). The association between eGFR and incident early AMD remained significant, if only participants without CKD (eGFR ≥60 mL/min/1.73 m²) were included into the analysis (P = 0.034; OR, 1.16; 95% CI, 1.01–1.32). For participants without CKD, compared with those with an eGFR of <75 mL/min/1.73 m², in individuals with an eGFR of 75 to 90 mL/min/1.73 m² and >90 mL/min/1.73 m², the risk of incident early AMD was increased by 1.63 fold (OR, 1.63; 95% CI, 1.03–2.60) and 1.92 fold (OR, 1.92; 95% CI, 1.18–3.13), respectively (P for trend = 0.014) (Table 4).

Table 4.

Multivariate-Adjusted Association of Biochemical Parameters and Incidence of Early and Intermediate Age-related Macular Degeneration (AMD)

	Groups by Biochemical Parameters	No.	P Value	OR (95% CI)	P for Trend
Early AMD
eGFR (ml/min/1.73m2 )*	60∼75	700	0.051	1 (reference)	0.014
	75∼90	1626	0.038	1.63 (1.03–2.60)
	>90	3159	0.009	1.92 (1.18–3.13)
Intermediate AMD
HDL (mmol/L)†	<1.04	1357	0.001	1 (reference)	0.001
	1.04∼1.6	3012	0.514	1.14 (0.76–1.71)
	>1.6	1267	0.002	2.00 (1.30–3.07)
Triglyceride (mmol/L)‡	<1.7	3981	0.029	1 (reference)	0.013
	1.7∼2.3	801	0.794	0.95 (0.63–1.43)
	>2.3	854	0.008	0.49 (0.29–0.83)

CI, confidence interval; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; OR, odds ratio.

^*Adjusted for age, sex, and diabetes.

^†Adjusted for age.

^‡Adjusted for age and total cholesterol.

For every mmol/L increase in serum HDL cholesterol concentration, the risk for developing intermediate AMD in 5 years increased 1.97 fold (P < 0.001; OR, 1.97; 95% CI, 1.38–2.83) (Table 3). Participants with a serum HDL cholesterol concentration of >1.6 mmol/L as compared with those with a serum HDL cholesterol concentration of <1.04 mmol/L had a 2-fold (OR, 2.00; 95% CI, 1.30, 3.07) higher risk of developing intermediate AMD (P for trend = 0.001) (Table 4). For every mmol/L increase in serum triglyceride concentration, the risk for developing intermediate AMD in 5 years decreased 0.77 fold (P = 0.008; OR, 0.77; 95% CI, 0.64–0.93) (Table 3). Compared with individuals with a serum triglyceride concentration of <1.7 mmol/L, the risk of developing intermediate AMD in participants with a serum triglyceride concentration of >2.3 mmol/L was reduced by 51% (OR, 0.49; 95% CI, 0.29–0.83) (P for trend = 0.013) (Table 4).

Discussion

The present study provides information on the 5-year incidence of AMD in a Chinese population. The incidence of any, early, intermediate and late AMD was 6.1%, 5.0%, 3.4%, and 0.3%, respectively. In multivariate analysis, incident early AMD was associated with older age (P < 0.001; OR, 1.04), female sex (P = 0.011; OR, 1.42), and a higher eGFR level (P = 0.020; OR, 1.01), whereas having diabetes mellitus was a protective factor (P = 0.019; OR, 0.49). Incident intermediate AMD was associated with older age (P < 0.001; OR, 1.05), a higher HDL cholesterol level (P < 0.001; OR, 1.97), and a lower triglyceride level (P = 0.008; OR, 0.77).

As the first longitudinal study using the Beckman grading system, this study provided the incidence of early, intermediate, and late AMD based on Beckman classification for the first time. Beckman classification is valuable in predicting the risk of late AMD and improving care and communication between eye care providers.¹² The main differences between the Beckman system and the Wisconsin system lie in two aspects. First, unlike in the Wisconsin system, only drusen size is considered in the Beckman system. Drusen morphology (soft or hard, distinct or indistinct) is no longer considered. Second, hard drusen with the diameter of 63 to 125 µm is defined as early AMD in the Beckman system, whereas it is not considered to be AMD in the Wisconsin system. As a result, the incidence of any AMD is higher in the present study (6.1%) compared with that in previous studies in Chinese population.^6,10 In other words, more individuals with early changes of AMD were detected using Beckman classification. This strategy might facilitate risk factor analysis and prevention of AMD progression.

According to literature, the incidence of AMD was lower in Chinese people, such as the Beijing Eye Study (early/late AMD, 4.2%/0.1%; age at baseline, ≥40 years)⁶ and the Handan Eye Study (early/late AMD, 4.2%/0.2%; age at baseline, ≥30 years)¹⁰ than in other ethnic groups, such as the Beaver Dam Eye Study (early/late AMD, 8.2%/0.9%; age at baseline, ≥43 years, United States),² the Blue Mountains Eye Study (early/late AMD, 8.7%/1.1%; age at baseline, ≥49 years, Australia),³ the Hisayama Study (early/late AMD, 8.5%/0.8%; age at baseline, ≥50 years, Japan),⁵ the Los Angeles Latino Eye Study (early/late AMD, 7.5%/0.2%; age at baseline, ≥40 years),¹¹ and the Singapore Malay Eye Study (early/late AMD, 5.9%/0.8%; age at baseline, ≥40 years).⁷

Because of different definitions for classifying early and intermediate AMD, it is difficult to compare the incidence of early and intermediate AMD in the present study with previous studies that used the Wisconsin system. However, the definition of late AMD was the same in the two grading systems; thus, the incidence of late AMD was comparable with previous studies.

In our study, female sex was associated significantly with incident early AMD (Beckman classification) after adjustment for age and other risk factors, although men and women did not differ significantly in the incidence of intermediate or late AMD. Previous studies had divergent results on gender differences in AMD incidence. In Asians and Latinos, men had a significantly higher incidence of early AMD in the Hisayama Study, the Singapore Malay Eye Study, the Handan Eye study, and the Los Angeles Latino Eye Study.^5,7,10,11 In Caucasians and Africans, female sex was associated with a higher incidence of early AMD in the Beaver Dam Eye Study and the Nakuru Eye Disease Cohort Study.^2,8 In the Beijing Eye Study, men and women did not differ significantly in the incidence of early AMD or late AMD.⁶ It remains unclear whether there is an actual gender difference based on ethnicity in the incidence of AMD.

In the present study, diabetes was a protective factor for early AMD (Beckman classification), whereas it had no association with intermediate AMD. Similarly, in the Ural Eye and Medical Study in Russia, higher AMD prevalence was correlated with lower diabetes prevalence.¹⁸ One study in South India also showed a possible protective effect of diabetes for AMD.^21,22 In the nondiabetic group, AMD was present in 22.7% subjects as compared with only 14.8% in subjects with diabetes (P < 0.001).²² However, diabetes was found to be a risk factor for AMD in the Los Angeles Latino Eye Study and the Nakuru Eye Disease Cohort Study in Kenya.^8,23 Other studies showed no relationship between diabetes and AMD risk. The early stages of diabetic retinopathy begin with alterations of the inner blood–retina barrier (BRB), whereas AMD involves the outer BRB (RPE). It has been suggested that, in diabetic macular edema, there is signaling from the damaged inner BRB that induces the upregulation of the transport function of the RPE (outer BRB),²⁴ thus delaying the development of the AMD. This may offer a likely explanation for the relationship between diabetes and lower incidence of early AMD in the present study.

Several studies have shown that the presence of CKD characterized by a decreased eGFR (<60 mL/min/1.73 m²) is independently associated with early, late and exudative AMD in the general population.^25–29 However, the relationship between eGFR and AMD in participants without CKD had remained unclear so far. The present study revealed that, in participants without CKD, with adjustment for systemic factors, including age, sex, and prevalence of diabetic mellitus, higher eGFR was a risk factor for incident early AMD. A similar relationship has been found between GFR and the risk of cardiovascular diseases. The curve that describes the relationship between GFR and the risk for cardiovascular disorders is U shaped, indicating that both a reduced GFR and an elevated GFR are cardiovascular risk factors.^30,31 We infer that the relationship between eGFR and AMD may also be U shaped. Recently, the prevalence of Gunn's dots was found to be associated with a higher eGFR.³² Indirect relationships between the ocular circulation and status and circulation of the kidney may be the reason for the association. Further studies are warranted to reveal the role of a high eGFR for these ocular characteristics.

Lipid metabolism may be involved in the pathogenic mechanism of AMD. However, conflicting results have been reported. The present study showed that a high plasma HDL cholesterol concentration was significantly associated with incident intermediate AMD (OR, 1.97; 95% CI, 1.4–2.8), whereas high triglyceride concentration was a protective factor (OR, 0.77; 95% CI, 0.64–0.93). It agrees with previous observations in the longitudinal ALIENOR Study and the Rotterdam Study showing the association between high HDL cholesterol level and the risk of incident AMD (OR, 1.2; 95% CI, 1.0–1.4 and OR, 1.20; 95% CI, 1.06–1.35, respectively).^9,33 A meta-analysis including 19 studies showed that a high HDL cholesterol level was associated with an increased AMD risk, whereas participants with a high triglyceride concentration showed a decreased AMD risk.³⁴ A 2017 meta-analysis using Mendelian randomization suggested that HDL cholesterol is a causal risk factor for AMD.³⁵ Another Mendelian analysis showed that higher genetically predicted HDL cholesterol level increased the risk of all AMD subtypes, whereas a high genetically predicted triglyceride level was associated with a decreased risk of different AMD subtypes.³⁶ However, in the Beijing Eye Study and Handan Eye Study, incident early AMD was not associated with HDL cholesterol concentration or triglyceride concentration in Chinese populations.^6,10 No relationship between incident AMD and blood lipids level was found in the Singapore Malay Eye Study either.⁷

Limitations of this study should be mentioned. First, this is a community-based instead of population-based study, so that an inclusion bias might have been possible. However, the Tongren Health Care Study mostly consisted of employees or retirees from enterprises, government offices, hospitals, academic institutes, and religious institutions who had free health care and who regularly participated in the health care examinations, without the need of an existing disease. Also, the AMD prevalence in our study population at baseline and follow-up examinations was comparable to the AMD prevalence in population-based investigations using the same diagnostic criteria.^15,17,18 Second, information about the refractive error or ocular axial length was not available, so that the multivariate analysis did not include axial length as a potentially confounding factor. Previous studies have shown that hyperopia or a small globe size was associated with a higher AMD incidence.⁶ Third, in contrast with the official Beckman guidelines for the definition of AMD, we included a minimum age of 40 years as the age limit for the definition of AMD, as has also been done in other studies.^2,6,7,11 Strengths of the Tongren Health Care Study were that the study population size was relatively large and that the study examined a relatively large number of systemic parameters.

In conclusion, the incidence of any, early, intermediate and late AMD was 6.1%, 5.0%, 3.4%, and 0.3%, respectively. Higher eGFR level was, beside older age and female sex, a risk factor for incident early AMD, while diabetes was a protective factor. Additionally, the incidence of intermediate AMD was associated with older age, higher HDL cholesterol level and lower triglyceride level. This may point to the role of renal circulation and lipid metabolism in incident AMD. This community-based longitudinal study provides a valuable understanding of the disease and its associated factors for targeted prevention and management strategies.