Associations of Complement Factor H and Smoking with Early Age-Related Macular Degeneration: The ALIENOR Study

Abstract

Purpose.: To assess the associations of complement factor H (CFH) Y402H polymorphism and smoking with specific features of early AMD (type, location, and area).

Methods.: The ALIENOR study is a population-based study of age-related eye diseases in 963 residents of Bordeaux (France), aged 73 years or more. AMD features were graded from nonmydriatic color retinal photographs. CFH Y402H was genotyped by using DNA extracted from blood. Statistical analyses included 796 subjects with complete data.

Results.: CFH CC genotype was strongly associated with late neovascular AMD (OR, 6.0; 95% confidence interval [CI], 1.5–23.5) but not with late atrophic AMD (OR, 0.9; 95% CI, 0.2–4.3). Among early characteristics, it was associated with central soft drusen (within 500 μm of the fovea), whether of intermediate (63–125 μm; OR, 2.7; 95% CI, 1.5–4.8), or large (>125 μm; OR, 5.9; 95% CI, 2.2–15.7) size, but not with pericentral soft drusen (500–3000 μm from the fovea). It was also strongly associated with a large central area of soft drusen (OR, 5.7; 95% CI, 1.7–19.2). Similarly, heavy smoking (>20 pack-years) was strongly associated with central large drusen (OR, 3.9; 95% CI, 1.6–9.6) and a large central area of drusen (OR, 3.5; 95% CI, 1.2–10.0), but not with pericentral soft drusen. By contrast, both CFH CC and smoking tended to be more strongly associated with pericentral pigmentary abnormalities.

Conclusions.: Location of abnormalities, together with type and area, may prove useful for the identification of subjects at high risk for late AMD.

Age-related maculopathy (AMD) is the leading cause of blindness in industrialized countries. ¹ In recent decades, major advances have been achieved in the understanding of this disease. ² Smoking, the Y402H polymorphism of complement factor H (CFH), and other genetic polymorphisms have been identified as major risk factors for this disease. ² With regard to natural history, clinical ^{3 –7} and population-based prospective studies ^{8 –13} have helped identify early retinal abnormalities (soft drusen and pigmentary abnormalities) that precede and predict the development of the late forms of AMD (neovascular AMD and geographic atrophy). Although these features clearly identify a subgroup of the population in which most of the cases of late AMD occur, several epidemiologic studies ^7,9,11,13,14 have suggested that this subgroup is heterogeneous and that additional characteristics of drusen and pigmentary abnormalities, such as type (distinct, indistinct, and reticular drusen), area, and location and their combinations, are useful for a further stratification of the risk of incident late AMD. 

Thus, early AMD, as currently defined in epidemiologic and clinical studies, probably represents a heterogeneous group of subjects, some of whom actually bear a very low risk of developing late AMD. This may be one of the reasons why associations of risk factors are often much weaker with early than with late AMD in epidemiologic studies. For instance, smoking is associated with a 2.5- to 4.5-fold increased risk of late AMD, ¹⁵ while its association with early AMD is usually weak, ^{16 –23} and did not reach statistical significance in many studies. ^{17,19,21 –23} Similarly, the subjects homozygous for the C allele of the CFH Y402H polymorphism generally exhibit a 6- to 14-fold increased risk of late AMD, ^{24 –32} whereas odds ratios (ORs) for early AMD are generally around 2. ^{24 –27,33 –35}  

In the Rotterdam Study, ORs for the CFH CC genotype increased with increasing stages of early AMD (from 1.16 for stage 1 to 4.58 for stage 3), which were defined according to their increasing prediction of incident late AMD. ²⁵ This suggests that high-risk early AMD is more closely related to the CC genotype than low-risk early AMD and that subjects with early AMD may be heterogeneous, not only with respect to their probability of developing late AMD, but also with regard to their association with risk factors. 

In the present study, we investigated the associations of several features of early AMD (type, location, and area) with CFH Y402H polymorphism and smoking in a population-based study of elderly French subjects, to identify specific early abnormalities that are associated with a high risk of late AMD. 

Methods

Study Goals

The ALIENOR (Antioxydants, Lipids Essentiels, Nutrition et maladies OculaiRes) Study is a population-based prospective study conducted to assess the associations of age-related eye diseases (AMD, glaucoma, cataract, and dry eye syndrome) with nutritional factors (in particular antioxidants, macular pigment, and fatty acids), determined from plasma measurements and estimation of dietary intakes. ³⁶ It also takes into account other major determinants of eye diseases, including gene polymorphisms, lifestyle, and vascular factors. 

Study Sample

Subjects of the ALIENOR Study were recruited from an ongoing population-based study on the vascular risk factors for dementia, the Three City (3C) Study. ³⁷ The 3C Study included 9294 subjects aged 65 years or more from three French Cities (Bordeaux, Dijon, and Montpellier), among which 2104 were recruited in Bordeaux. Subjects were contacted individually from the electoral rolls. They were initially recruited in 1999 to 2001 and have been followed-up approximately every 2 years since. Data collected at each examination included cognitive testing with diagnosis of dementia and assessment of vascular risk factors. In addition, at baseline fasting blood and DNA samples were collected and kept frozen at −80°C. 

The ALIENOR Study consists of an eye examination, which was proposed to all participants in the third follow-up (2006–2008) of the 3C cohort in Bordeaux and will be proposed in future follow-ups. Among the 1450 participants re-examined in 2006 to 2008, 963 (66.4%) participated in the ALIENOR Study. 

This research complied with the Declaration of Helsinki. Participants gave written consent for participation in the study. The design of the study was approved by the Ethics Committee of Bordeaux (Comité de Protection des Personnes Sud-Ouest et Outre-Mer III) in May 2006. 

Eye Examination

The eye examination took place in the Department of Ophthalmology of the University Hospital of Bordeaux. It included a recording of ophthalmic history, measures of visual acuity, refraction, two 45° nonmydriatic color retinal photographs (one centered on the macula, the other centered on the optic disc), measures of intraocular pressure and central corneal thickness, and a tear breakup time test. A self-completed questionnaire on risk factors specific to the eye and dry eye symptoms was filled at home and brought back on the day of the eye examination. 

Retinal photographs were performed using a high-resolution digital nonmydriatic retinograph (TRC NW6S; Topcon, Japan). Such high-resolution digital photographs have shown comparable grading qualities as previous film-based photographs. ^38,39 Photographs were interpreted in duplicate by two specially trained technicians. Inconsistencies between the two interpretations were adjudicated by a retina specialist for classification of AMD and other retinal diseases, or by a glaucoma specialist for classification of glaucoma. All cases of late AMD, other retinal diseases and glaucoma were reviewed and confirmed by specialists. Graders had no access to genetic determinations. 

Classification of AMD

Retinal photographs were interpreted according to the international classification ⁴⁰ and to a modification of the grading scheme used in the Multi-ethnic Study of Atherosclerosis (MESA) for drusen size, location, and area. ⁴¹ Late AMD was defined by the presence of neovascular AMD or geographic atrophy within the grid (3000 μm from the foveal center). Neovascular AMD included serous or hemorrhagic detachment of the retinal pigment epithelium (RPE) or sensory retina, subretinal or sub-RPE hemorrhages and fibrous scar tissue. Geographic atrophy was defined as a discrete area of retinal depigmentation, 175 μm in diameter or larger, characterized by a sharp border and the presence of visible choroidal vessels. Five cases of late AMD had no gradable photographs and were classified by using ophthalmic history of AMD and AMD therapy (in particular antiangiogenic agents and photodynamic therapy), and confirmed by their treating ophthalmologist. Because etiologies of neovascular and atrophic AMD may be different, ⁴² we separated two groups: subjects with neovascular AMD (with or without geographic atrophy) and subjects with late atrophic AMD (geographic atrophy without neovascular AMD). 

The definition of early AMD was based on the classification of two major European epidemiologic studies (Rotterdam Study ¹⁰ and Eureye ⁴³ ), to facilitate comparison with these studies. This classification has also been used in the INDEYE Study. ⁴⁴ It defines three grades of early AMD (in the absence of late AMD): early AMD 1 (soft distinct drusen without pigmentary abnormalities or pigmentary abnormalities without large drusen [>125 μm]); early AMD 2 (soft indistinct drusen and/or reticular drusen and/or soft distinct drusen associated with pigmentary abnormalities); early AMD3 (soft indistinct drusen and/or reticular drusen, associated with pigmentary abnormalities). Soft distinct and indistinct drusen were larger than 125 μm in diameter and with uniform density and sharp edges or decreasing density from the center outward and fuzzy edges, respectively. Pigmentary abnormalities were defined as areas of hyperpigmentation and/or hypopigmentation (without visibility of choroidal vessels). 

Because the early AMD3 group was too small in our study, it was merged with the early AMD 2 group. Subjects were thus classified according to their worse eye into one of the five exclusive groups: no AMD, early AMD 1, early AMD 2 (including initial early AMD 3), late atrophic AMD, and late neovascular AMD. Our definition of early AMD 2 (soft indistinct drusen and/or reticular drusen and/or soft distinct drusen associated with pigmentary abnormalities) is also similar to the definition of early AMD in the Blue Mountains Eye Study. ¹⁹  

In addition, detailed characteristics of early AMD abnormalities were studied, using a modification of the grading scheme used in the MESA for drusen size, location, and area. ⁴¹ This grading scheme includes a detailed grading protocol and standard photographs, which were used for the training of graders in the present study. It has shown high intra- and intergrader reproducibility. ⁴¹ The items graded according to this classification are presented in Table 1. For each item, the highest possible score was given. Thus, the categories were mutually exclusive. For instance, a subject showing both pericentral and central intermediate soft drusen were given a score of 3 (central intermediate soft drusen). A score of 0 for a given item therefore corresponds to the absence of abnormality. For the purpose of the present study, only definite characteristics were studied. Questionable abnormalities were considered as absent. 

Table 1.

 

Classification of Detailed Early AMD Characteristics, According to the Grading Scheme of the MESA ⁴¹

Characteristic, Presence, Location	Score
Intermediate soft drusen (63–125 μm)
None	0
Questionable	1
Pericentral	2
Central	3
Large soft distinct drusen (>125 μm)
None	0
Questionable	1
Pericentral	2
Central	3
Large soft indistinct drusen (>125 μm)
None	0
Questionable	1
Pericentral	2
Central	3
Reticular drusen
None	0
Questionable	1
Present	2
Large area of soft drusen (>500 μm in diameter)
None	0
Questionable	1
Pericentral	2*
Central	3†
Hyperpigmentation
None	0
Questionable	1
Pericentral	2
Central	3
Hypopigmentation
None	0
Questionable	1
Pericentral	2
Central	3

 Pericentral: 500–3000 μm from fovea; central: <500 μm from fovea.

*  Reaching 500 μm in area within the whole grid.

†  Reaching 500 μm in area within 500 μm from the fovea.

We generated two additional parameters: large soft drusen (combining large soft indistinct and distinct drusen) and pigmentary abnormalities (combining hypo- and hyperpigmentations). 

Genotyping

Subjects were genotyped for CFH single-nucleotide polymorphism rs1061170 (Y402H) in exon 9. Genotyping was performed on DNA extracted from leukocytes, using genotyping assays (Taqman; Applied Biosystems, Inc., [ABI], Foster City, CA), as described by the suppliers. 

Smoking Exposure

Participants who had smoked fewer than 100 cigarettes during their lifetime were defined as having never smoked. For participants who were current or past smokers, the number of pack-years of cigarettes smoked during their lifetime was calculated: number of smoking years × (mean number of cigarettes per day)/20. 

Statistical Analysis

Statistics were performed using SAS software, ver. 9.1 (SAS Institute Inc, Cary, NC). The associations of AMD with the CFH Y402H genotypes (TT, TC, and CC) and smoking exposure were estimated using polytomic nominal logistic regression. This method allows the estimate of one OR for each type of AMD, with subjects without AMD being the reference. For each type of early AMD characteristics, we also used polytomic nominal logistic regression to estimate the ORs for each location. Subjects without the given abnormality (score 0 or 1) constituted the reference group. Thus, for instance, the OR for central intermediate drusen compares subjects with central intermediate soft drusen with subjects without any intermediate soft drusen. 

We first studied the association of CFH and smoking with the different stages of AMD and then analyzed the association with all characteristics of early AMD, as defined in our grading scheme. Associations with reticular drusen could not be determined because of the small number of affected subjects (n = 15). 

Analyses were adjusted for age and sex. Because of limited sample size, it was not possible to test interactions between CFH and smoking. 

Results

Among the 1450 participants re-examined from 2006 through 2008, 963 (66.4%) participated in the ALIENOR Study, among which 879 (91.3%) had gradable photographs in at least one eye. Among them, 796 (90.5%) had complete data for CFH Y402H polymorphism and smoking and were included in the statistical analyses. In this population, the prevalence of AMD (any type) was 34.3% (273/796) and late AMD was found in 5.7% (45/796) of subjects. As shown in Table 2, we first compared the characteristics of three groups of subjects: those included in the present statistical analysis (n = 796), those excluded from statistical analysis because of missing data (n = 167), and those who did not participate in the ALIENOR Study (n = 487). Although participants included in the statistical analysis were younger and had different distribution according to sex than the two other groups, we did not detect significant differences in smoking, CFH Y402H polymorphism or AMD status (available only for the two first groups), among these groups. 

Table 2.

 

Comparison of Participants with Nonparticipants

	Participants Included in the Statistical Analysis (n = 796)	Participants Not Included in the Statistical Analysis* (n = 167)	Nonparticipants (n = 487)	P
Age, y ± SD	79.7 (4.3)	80.4 (4.5)	82.9 (5.1)	<0.0001
Sex				0.004
Female	505 (63.4)	91 (54.5)	334 (68.6)
Male	291 (36.6)	76 (45.5)	153 (31.4)
Smoking				0.06†
Never	517 (64.9)	97 (62.6)	345 (71.9)
<20 pack-years	144 (18.1)	34 (21.9)	70 (14.6)
≥20 pack-years	135 (17.0)	24 (15.5)	65 (13.5)
CFH Y402H genotype				0.97†
TT	360 (45.2)	39 (47.6)	189 (46.8)
TC	345 (43.3)	33 (40.2)	168 (41.6)
CC	91 (11.4)	10 (12.2)	47 (11.6)
Age-related maculopathy				0.81†
None	523 (65.7)	57 (68.7)	Not available
Early AMD 1	143 (18.0)	15 (18.1)
Early AMD 2	85 (10.7)	7 (8.4)
Late atrophic AMD	21 (2.6)	3 (3.6)
Late neovascular AMD	24 (3.0)	1 (1.2)

 Data are the number of subjects (% of the total group), unless otherwise specified.

*  Because of missing data.

†  Adjusted for age and sex.

As shown in Table 3, mean age was 80 years, subjects without AMD being younger (79.5 years) than subjects with late atrophic and neovascular AMD (83.2 and 82.8 years), as expected. Two thirds of subjects were women, as expected, given the age of the subjects. Distribution of the genotypes of the CFH Y402H was 45.2% of subjects with TT genotype, 43.3% with TC, and 11.4% with CC. Only 37 (4.7%) subjects were current smokers. It was therefore impossible to conduct statistical analyses in this group of subjects. For further analyses, smoking was classified according to number of pack-years, divided in two groups according to its median among smokers (20 pack-years). 

Table 3.

 

Characteristics of Participants According to AMD Status in the Alienor Study

	None (n = 523)	Early AMD 1 (n = 143)	Early AMD 2 (n = 85)	Late Atrophic AMD (n = 21)	Late Neovascular AMD (n = 24)	All (n = 796)
Age, y ± SD*	79.5 (4.3)	80.6 (4.4)	80.5 (4.5)	83.2 (4.7)	82.8 (4.3)	80.0 (4.4)
Male	209 (40.0)	46 (32.2)	22 (25.9)	10 (47.6)	4 (16.7)	291 (36.6)
CFH Y402H genotype
TT	250 (47.8)	65 (45.4)	29 (34.1)	12 (57.1)	4 (16.7)	360 (45.2)
TC	222 (42.4)	58 (40.6)	43 (50.6)	7 (33.3)	15 (62.5)	345 (43.3)
CC	51 (9.8)	20 (14.0)	13 (15.3)	2 (9.5)	5 (20.8)	91 (11.4)
Smoking
Never	334 (63.9)	100 (69.9)	56 (65.9)	10 (47.6)	17 (70.8)	517 (64.9)
Current	24 (4.6)	5 (3.5)	4 (4.7)	2 (9.5)	2 (8.4)	37 (4.7)
Former	165 (31.5)	38 (26.6)	25 (29.4)	9 (42.9)	5 (20.8)	242 (30.4)
Pack-years of smoking (±SD)
0	334 (63.9)	100 (69.9)	56 (65.9)	10 (47.6)	17 (70.8)	517 (64.9)
1–20	107 (20.4)	17 (11.9)	11 (12.9)	7 (33.3)	2 (8.3)	144 (18.1)
>20	82 (15.7)	26 (18.2)	18 (21.2)	4 (19.1)	5 (20.8)	135 (17.0)

 Data are the number of subjects (% of the total group), unless otherwise specified.

*  Mean (SD).

The associations of CFH Y402H genotypes and smoking exposure with AMD types are presented in Table 4, the reference group being subjects with no AMD. The CC genotype was associated with a sixfold increased risk for late neovascular disease and a twofold increased risk for early AMD 2. The risk of late neovascular AMD and early AMD 2 was also increased 4.2- and 1.7-fold in heterozygotes (TC), respectively. By contrast, we were unable to demonstrate a statistically significant association of the CFH Y402H genotypes with late atrophic AMD and early AMD 1. These associations were not affected by further adjustment for age and sex (Table 4). 

Table 4.

 

Associations of AMD with CFH Y402H Polymorphism and Smoking Exposure

	Early AMD 1 (n = 143)	Early AMD 2 (n = 85)	Late Atrophic AMD (n = 21)	Late Neovascular AMD (n = 24)
CFH Y402H Genotype
Unadjusted OR
TT (reference)	1.0	1.0	1.0	1.0
TC	1.0 (0.7–1.5)	1.7 (1.0–2.8)	0.7 (0.2; 1.7)	4.2 (1.38–12.9)
	0.98	0.05	0.38	0.01
CC	1.5 (0.8–2.7)	2.2 (1.1–4.5)	0.8 (0.2–3.8)	6.1 (1.6–23.6)
	0.17	0.03	0.79	0.008
Age- and sex-adjusted OR
TT (ref.)	1.0	1.0	1.0	1.0
TC	1.0 (0.7–1.5)	1.6 (1.0–2.7)	0.7 (0.2–1.7)	4.1 (1.3–12.5)
	0.92	0.06	0.38	0.01
CC	1.5 (0.8–2.7)	2.1 (1.0–4.4)	0.9 (0.2–4.3)	6.0 (1.5–23.5)
	0.18	0.04	0.91	0.01
Smoking Exposure
Unadjusted OR
Never smoked (ref.)	1.0	1.0	1.0	1.0
1–20 pack-years	0.5 (0.3; 0.9)	0.6 (0.3–1.2)	2.2 (0.8; 5.9)	0.4 (0.08–1.6)
	0.03	0.16	0.12	0.19
>20 pack-years	1.0 (0.6–1.7)	1.3 (0.7–2.3)	1.6 (0.5; 5.3)	1.2 (0.4–3.3)
	0.82	0.36	0.42	0.73
Age- and sex-adjusted OR
Never smoked (ref.)	1.0	1.0	1.0	1.0
1–20 pack-years	0.6 (0.3–1.1)	0.9 (0.4–1.8)	2.2 (0.7–6.7)	0.6 (0.1–3.0)
	0.10	0.71	0.17	0.58
>20 pack-years	1.3 (0.7–2.2)	2.1 (1.1–4.1)	1.8 (0.5–7.0)	2.8 (0.8–8.9)
	0.39	0.03	0.38	0.09

 Data are the OR (95% CI) estimated by using polytomic nominal logistic regression.

By contrast, associations with smoking were evident only after adjustment for age and sex. This finding was expected, since the prevalence of smoking decreases with age, which biases the associations toward the null. After such adjustment, heavy smokers (>20 pack-years) had a twofold increased risk of early AMD 2. Associations of smoking with early AMD 1, late atrophic, and late neovascular AMD did not reach statistical significance, although there was a tendency of increased risk for neovascular AMD in heavy smokers (OR, 2.8; 95% CI, 0.8–8.9). 

We then studied the associations of CFH Y402H and smoking with detailed characteristics of early abnormalities, after exclusion of subjects affected with late AMD. These characteristics do not completely overlap with the definitions of early AMD 1 and 2. In particular, the initial classification of early AMD 1 and 2 did not include criteria for intermediate soft drusen, area of soft drusen (defined as area of intermediate and large soft drusen), or location of drusen. Thus, for instance, 269 subjects with soft intermediate drusen and 30 subjects with large area of soft drusen (having only numerous soft intermediate drusen) were considered as having no AMD in the initial classification. By contrast, subjects with large soft drusen were, by definition, all classified as having early AMD 1 or 2. Although all indistinct soft drusen were classified in early AMD 2, subjects with soft distinct drusen were classified as early AMD 1 in the absence of pigmentary abnormalities, or as early AMD 2 in their presence. The item “soft distinct drusen” therefore groups some of the subjects classified in early AMD 1 and 2. Finally, although size and area of drusen are correlated, overlap was not complete among the different characteristics, with, for instance, 49 of 127 cases (38.5%) with a large area of drusen having no large drusen (but only numerous intermediate drusen). We studied whether some of these characteristics, which are not taken into account in most definitions of early AMD, may be more closely associated with CFH Y402H and smoking. 

Table 5 displays the age and sex-adjusted associations of CFH Y402H polymorphism exposure with characteristics of early abnormalities. Overall, the CC genotype was associated with a significantly increased risk of central soft drusen. The increased risk varied according to the characteristics of the drusen from a 2.7-fold increased risk (95% CI, 1.5–4.8) for central intermediate soft drusen to a 10.0-fold increase (95% CI, 2.4–42.2) for central distinct soft drusen. By contrast, pericentral drusen, whether intermediate, large, soft distinct, or soft indistinct, were not significantly associated with the CC genotype. 

Table 5.

 

Age- and Sex-Adjusted Associations of Early AMD Characteristics with CFH Y402H Polymorphism in the Alienor Study

	n	CFH Y402H Polymorphism
		TT (ref.)	TC	CC
Intermediate soft drusen
Central	169	1.0	1.4 (0.9–2.2)	2.7 (1.5–4.8)
Pericentral	260	1.0	1.2 (0.8–1.7)	0.6 (0.3–1.1)
Large soft drusen
Central	35	1.0	2.4 (1.0–5.6)	5.9 (2.2–15.7)
Pericentral	89	1.0	1.3 (0.8–2.1)	1.5 (0.7–3.1)
Distinct soft drusen
Central	15	1.0	2.5 (0.6–9.8)	10.0 (2.4–42.2)
Pericentral	67	1.0	1.2 (0.7–2.0)	1.9 (0.9–4.0)
Indistinct soft drusen
Central	23	1.0	2.1 (0.7–5.7)	4.0 (1.2–12.8)
Pericentral	45	1.0	1.8 (0.9–3.4)	0.8 (0.2–2.9)
Large area of soft drusen
Central	26	1.0	3.6 (1.3–9.9)	5.7 (1.7–19.2)
Pericentral*	91	1.0	1.5 (0.9–2.5)	2.6 (1.3–5.0)
Hyperpigmentation
Central	69	1.0	0.9 (0.5–1.6)	1.7 (0.8–3.7)
Pericentral	49	1.0	1.1 (0.6–2.2)	2.6 (1.3–5.0)
Hypopigmentation
Central	81	1.0	0.7 (0.4–1.1)	1.6 (0.8–3.1)
Pericentral	53	1.0	1.1 (0.6–2.0)	2.4 (1.0–5.3)
Pigmentary abnormalities†
Central	90	1.0	0.7 (0.4–1.1)	1.5 (0.8–3.0)
Pericentral	55	1.0	1.4 (0.7–2.6)	2.9 (1.3–6.5)

 Data are OR (95% CI) estimated by using polytomic nominal logistic regression. Bold characters represent statistically significant OR (P < 0.05).

 Central: within 500 μm of the fovea; pericentral: 500–3000 μm from the fovea.

*  Area greater than 500 μm in diameter within 3000 μm of the fovea, but smaller than 500 μm in the central 500 μm.

†  Any hypopigmentation and/or hyperpigmentation.

Drusen covering a large area (greater than a circle 500 μm in diameter) were also strongly associated with CFH Y402H polymorphism, particularly when this area was reached in the central 500 μm of the macula (OR, 5.7; 95% CI, 1.7–19.2), but also when this area was reached only on the whole 3000 μm of the macula (OR, 2.6; 95% CI, 1.3–5.0). 

By contrast, pigmentary abnormalities (hyper- or hypopigmentation) showed stronger associations with the CFH polymorphism when they were pericentral (OR, 2.9; 95% CI, 1.3–6.5), whereas central pigmentary abnormalities were not significantly associated with the CFH polymorphism (OR, 1.5; 95% CI, 0.8–3.0). 

As presented in Table 6, results were similar for smoking exposure. Heavy smokers had an increased risk of central soft drusen, either large (OR, 3.9; 95% CI, 1.6–9.6), indistinct (OR, 4.6; 95% CI, 1.5–13.4), or covering a large area (OR, 3.5; 95% CI, 1.2–10.0). By contrast, no significant associations were found with pericentral drusen. Finally, pericentral pigmentary abnormalities tended to be more strongly associated with smoking than central pigmentary abnormalities, although both associations were statistically significant (OR, 2.3; 95% CI, 1.0–5.0 and OR, 1.9; 95% CI, 1.0–3.0, respectively). 

Table 6.

 

Age- and Sex-Adjusted Associations of Early AMD Characteristics with Smoking Exposure

	n	Never Smoked (Ref.)	1–20 Pack-Years	>20 Pack-Years
Intermediate soft drusen
Central	169	1.0	1.0 (0.6–1.8)	1.7 (0.9–3.0)
Pericentral	260	1.0	0.9 (0.5–1.4)	1.2 (0.8–2.0)
Large soft drusen
Central	35	1.0	1.1 (0.4–3.7)	3.9 (1.6–9.6)
Pericentral	89	1.0	0.8 (0.4–1.5)	0.9 (0.4–1.8)
Distinct soft drusen
Central	15	1.0	0.6 (0.1–4.7)	2.4 (0.6–10.1)
Pericentral	67	1.0	0.4 (0.2–1.1)	1.0 (0.5–2.1)
Indistinct soft drusen
Central	23	1.0	2.2 (0.7–7.0)	4.6 (1.5–13.4)
Pericentral	45	1.0	0.9 (0.4–2.4)	0.9 (0.3–2.4)
Large area of soft drusen
Central	26	1.0	1.5 (0.5–5.0)	3.5 (1.2–10.0)
Pericentral*	91	1.0	1.1 (0.5–2.0)	1.3 (0.6–2.5)
Hyperpigmentation
Central	69	1.0	0.5 (0.2–1.3)	1.9 (0.9–3.8)
Pericentral	49	1.0	1.1 (0.4–2.7)	2.5 (1.1–5.7)
Hypopigmentation
Central	81	1.0	0.7 (0.3–1.4)	1.9 (1.0–3.7)
Pericentral	53	1.0	0.8 (0.3–2.0)	2.6 (1.2–5.7)
Pigmentary abnormality†
Central	90	1.0	0.6 (0.3–1.3)	1.9 (1.0–3.0)
Pericentral	55	1.0	0.8 (0.3–2.1)	2.3 (1.0–5.0)

 Data are OR (95% CI) estimated by using polytomic nominal logistic regression. Bold characters represent statistically significant OR (P < 0.05).

 Central: within 500 μm of the fovea; pericentral : 500–3000 μm from the fovea.

*  Area greater than 500 μm in diameter within 3000 μm of the fovea, but smaller than 500 μm in the central 500 μm.

†  Any hypopigmentation and/or hyperpigmentation.

Discussion

In the present study, both CFH Y402H polymorphism and smoking were associated with an increased risk for AMD. Moreover, associations of CFH Y402H polymorphism and smoking were limited to central soft drusen (whether intermediate, distinct, or indistinct) and drusen covering a large area. With regard to pigmentary abnormalities, associations tended to be stronger with pericentral abnormalities. 

Our results are consistent with two major population-based studies. In particular, in the Rotterdam Study, stages 1 and 2 of early AMD showed ORs of 1.2 and 2.0 for the CC genotype of CFH Y402H, respectively. ²⁵ This is very close to our findings (1.5 and 2.1 for early AMD 1 and 2, respectively), with similar definitions of these stages. In addition, in the Rotterdam study, a third stage of AMD was individualized, combining soft indistinct or reticular drusen with pigmentary abnormalities. It showed an OR of 4.6 for the CC genotype. We could not study this stage of early AMD in the present study, and because of the very low number of subjects in this group, we included them in the early AMD 2. In the Blue Mountains Eye Study, early AMD showed an OR of 2.4 for the CC genotype. ²⁴ Definition of early AMD in this study was similar to early AMD 2 in the present study, which showed an OR of 2.1 with the CC genotype. Comparisons with other previous clinical and epidemiologic studies are more difficult because of differing definitions of early AMD. However, most studies evidenced ORs around 2 for early abnormalities (in particular large drusen), ^26,27,33,35 which is consistent with our findings. 

With regard to smoking, previous epidemiologic studies have usually shown weak associations with early AMD, often not reaching statistical significance. ^{16 –23} This is also the case in our study, with an OR of 1.3 (95% CI, 0.7–2.2) and 2.1 (95% CI, 1.1–4.1) for early AMD 1 and 2, respectively. 

To our knowledge, there are no published studies on the associations of CFH and smoking with location and area of retinal abnormalities. Our results suggest that their associations are much stronger with central soft drusen and with soft drusen covering a large area. Indeed, the magnitude of ORs for these particular features with CFH and smoking were of the same magnitude as those expected for late AMD (i.e., approximately six for CC genotype and three for smoking). By contrast, large pericentral drusen showed no such association, with point estimates of OR close to unity. Furthermore, intermediate central drusen (63–125 μm; but not pericentral) also showed significant association with CFH. Thus, central soft drusen, whether intermediate or large, seem more closely associated with major risk factors of AMD, than pericentral drusen, even when large. 

In the present study, pericentral pigmentary abnormalities also tended to be more strongly associated with CFH and smoking than central pigmentary abnormalities. Our results therefore suggest that associations of major risk factors with features of early AMD may be heterogeneous and may be limited to some specific characteristics. 

Both large area and central location of drusen have been shown to be associated with increased incidence of late AMD. ^{3,7 –13} This suggests that these characteristics of drusen may be helpful in the stratification of subjects with early AMD. Further studies are needed to identify the associations of these particular features with incident late AMD and its risk factors and to better assess their prognostic value. In particular, few studies have described the risk of incident late AMD according to location of drusen, ^3,13 although central location of drusen was highly predictive in these studies. 

A potential limitation to our results is the questionable representativeness of the sample. Indeed, about two thirds of the participants in the 3C Study accepted the eye examination. In this subsample, subjects of younger age and higher socioeconomic status were overrepresented among subjects participating in the 3C Study. ³⁶ The included subjects may therefore be healthier and have different lifestyles, in particular concerning smoking, than the general population. This may also have affected the prevalence of eye diseases. However, for most parameters of interest in our study, in particular smoking, subjects included in the ALIENOR study were not different from those who did not participate. ³⁶ As described previously, ³⁶ the age- and sex-specific prevalence rates of AMD were also similar to those observed in other studies performed in Europe, ^43,45 and other industrialized countries. ⁴⁶ The frequencies of the genotypes of CFH Y402H polymorphism was also similar to those observed in previous studies in Caucasian subjects. ^24,25,34 Finally, although the distributions of smoking and CFH and the prevalence of ocular endpoints may be different from the general population, this is unlikely to bias the estimation of the associations of AMD with smoking and CFH. 

The small number of subjects affected by late AMD is another limitation. In particular, insufficient statistical power may explain the lack of association of CFH with atrophic AMD in our sample or of smoking with both late forms of AMD. This is reflected by the wide confidence intervals, which show that true ORs of 4 to 9 cannot be excluded (Table 3). The limited sample size also impeded assessment of interactions of CFH and smoking in their relation to AMD. Another concern is the high number of comparisons performed, in particular when studying specific features of early AMD. We therefore cannot exclude that some of the observed associations were due to chance finding, although our findings are generally consistent with previous studies in the field. 

Recent studies have evidenced associations of CFH polymorphism with other retinal features (in particular peripheral drusen and pigmentary abnormalities). ^35,47,48 Unfortunately, in the present study, the presence of drusen and pigmentary abnormalities were graded only within the central 3000 μm of the retina, and therefore these data are not available in our study. We also did not have information on ARMS2 A69S polymorphism, which represents a major risk factor for AMD. ² In the future, it will be interesting to study whether the retinal features most strongly associated with smoking and CFH, are also associated with ARMS2 A69S. 

In conclusion, the present study confirms the associations of CFH Y402H polymorphism and smoking with AMD and suggests that their associations are particularly strong with some features of early AMD (central soft drusen, large area of soft drusen, pericentral pigmentary abnormalities). The location of drusen, together with the type and area, may prove useful for the identification of subjects at high risk for late AMD, who may benefit the most from preventive strategies, increased ophthalmic follow-up, and early treatment. Further studies, in particular population-based prospective studies, will be needed to confirm these findings. 

Footnotes

 Supported by Laboratoires Théa, Clermont-Ferrand, France. The sponsor participated in the design of the study, but not in the collection, management, statistical analysis, and interpretation of the data or in the preparation, review, and approval of the present manuscript.

Footnotes

 Disclosure: C. Delcourt, Laboratoires Théa (F); M.-N. Delyfer, Laboratoires Théa (F); M.-B. Rougier, Laboratoires Théa (F); P. Amouyel, Laboratoires Théa (F); J. Colin, Laboratoires Théa (F); M. Le Goff, Laboratoires Théa (F); F. Malet, Laboratoires Théa (F); J.-F. Dartigues, Laboratoires Théa (F); J.-C. Lambert, Laboratoires Théa (F); J.-F. Korobelnik, Laboratoires Théa (F)