Allergy Is a Protective Factor Against Age-Related Macular Degeneration

Abstract

Purpose.: To investigate the role of allergy on AMD.

Methods.: Age-related macular degeneration staging was performed for 3585 individuals (1878 from Cologne, Germany, and 1707 from Nijmegen, The Netherlands). Interviewer-assisted questionnaires were evaluated for the factors smoking, use of corticosteroids, and history of allergy, including causative allergens. Serum complement component C3d and C3 levels were measured and the C3d:C3 ratio was calculated. Associations of allergy with AMD/late AMD were assessed by logistic regression analysis; C3d:C3 ratio was compared between groups.

Results.: The discovery cohort from Cologne included 864 AMD patients and 1014 controls; 495 patients had late AMD. Positive history of allergy showed strong protective effects on the phenotype AMD (OR 0.52; P = 3.42 × 10⁻⁹) and late AMD (OR 0.32; P = 2.57 × 10⁻¹³). Subclassification in allergy-provoking agents showed significant protective effects in all groups. After adjustment for age, sex, smoking, and corticosteroid use, protective effects for AMD (OR 0.75; P = 0.018) and late AMD (OR 0.49; P = 2.87 × 10⁻⁵) were confirmed. Although the C3d:C3 ratio was higher in AMD/late AMD patients (both P < 0.001), there was no association with allergy in AMD (P = 0.22). The protective effect of allergy on AMD was confirmed in the replication cohort from Nijmegen (P = 0.002 for AMD; P = 0.0001 for late AMD).

Conclusions.: Allergy has a protective effect on the development of AMD independent of the provoking allergen, which cannot be explained by complement activation. Further investigations are necessary to elucidate the molecular mechanisms underlying the protective effect of allergy on AMD.

Introduction

Age-related macular degeneration (AMD) is the leading cause of blindness in the elderly population in the industrialized world. Both genetic and environmental factors are involved in the pathogenesis of the disease. Genetic studies have identified various risk factors in genes of the complement system, including complement factor H (CFH), complement component 3 (C3), complement component 2, and complement factor I . ^1–3 It is thought that these polymorphisms lead to a dysregulation of the alternative pathway of the complement system. ⁴ Although AMD manifests locally in the eye, complement is also systemically activated in serum of AMD patients. ^5,6 While the role of the innate immune system in AMD is well established, there is now increasing evidence of a contribution of adaptive immunity on the development of AMD as well: autoimmunity seems to be involved in AMD pathogenesis and various autoantibodies against retinal antigens, but also various antigens not derived from the retina have been detected in AMD patients. ⁷  

Allergy is an immune-mediated inflammatory response to normally harmless environmental substances known as allergens, resulting in diseases such as asthma, allergic rhinitis, atopic dermatitis, and food allergy. Traditionally, allergic diseases are referred to immediate or type 1 hypersensitivity reactions. ⁸ Allergens stimulate mast cells to release vasoactive substances that lead to an inflammatory response. There is also evidence for complement activation in allergic disorders and complement components, such as C3, have been shown to be elevated in allergic asthma. ⁹  

Because of the involvement of various components of the immune system in the pathogenesis of both AMD and allergy, we studied their relationship in two large Caucasian case-control cohorts. 

Materials and Methods

In our case-control study, data of 3585 participants were analyzed. The study was performed in accordance with the tenets of the Declaration of Helsinki and the Medical Research Involving Human Subjects Act and was approved by the local ethics committee of the University Hospitals in Cologne and Nijmegen. Written informed consent was obtained from all participants. 

Discovery Cohort

The discovery cohort consisted of 1878 Caucasian participants from Cologne, Germany, and was part of the European Genetic Database (EUGENDA, www.eugenda.org). 

Age-related macular degeneration staging was performed by grading of retinal images including stereo fundus photographs (FP), fluorescein angiograms (FA) and spectral domain optical coherence tomograms (SDOCT) according to the standard protocol of the Cologne Image Reading Center by certified graders. Age-related macular degeneration was classified by the presence of pigmentary changes together with at least 10 small drusen (<63 μm) or presence of intermediate (63–124 μm) (early AMD) or the presence of large drusen (≥125 μm diameter) or ≥15 intermediate drusen (intermediate AMD) in the Early Treatment Diabetic Retinopathy Study (ETDRS) grid on FP. The subgroup of late AMD was defined as either AMD with subfoveal geographic atrophy and/or choroidal neovascularisation (CNV) in at least one eye. Geographic atrophy was defined as sharply demarcated round or oval area of depigmentation of the RPE of 175 μm or larger diameter with increased visibility of choroidal vessels within the central circle of the ETDRS grid on FP without signs of CNV. Late AMD with CNV was defined as choroidal neovascular lesion within the ETDRS grid secondary to AMD on FP, FA or SDOCT, when there was evidence for fluid, blood, or fibrovascular tissue on FP, active classic or occult CNV, or signs of previous CNV, such as staining scar on FA and/or retinal or subRPE fluid and/or tissue secondary to AMD. 

Demographic data and nongenetic parameters, including history of smoking, use of corticosteroids, and history of allergy, were obtained by standardized interviewer-assisted questionnaires. Participants were asked: “Do you have any allergy? If yes, please specify the causing agent.” For analysis, history of allergy was categorized in the allergy-causing categories: pollen, drugs, food, house dust mites, and other allergens. 

Complement component C3 and the activation fragment C3d were measured in serum samples as described previously, ¹⁰ and the C3d:C3 ratio was calculated to quantify systemic complement activation in the discovery cohort. 

Replication Cohort

The analysis was replicated in 1707 participants from Nijmegen, The Netherlands, who were also included in EUGENDA. 

Retinal images were graded as described in the discovery cohort. Data from the same questionnaires also included age, sex, history of allergy, history of smoking, and use of corticosteroids. 

Statistical Analysis

All calculations were performed using SPSS software version 21.0 (IBM Software and Systems, Armonk, NY). Associations between allergy and AMD were assessed by logistic regression analysis. The possible confounding factors for adjustment age, sex, smoking (ever/never smoker), and use of corticosteroids were added. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated in unadjusted and adjusted models for AMD and late AMD. Subgroup analyses were not carried out if groups consisted of fewer than 1% of cases. Associations between AMD phenotype, history of allergy, and the C3d:C3 ratio were analyzed in the discovery cohort using Mann-Whitney U test. P values of 0.05 or less were considered statistically significant. 

Results

Demographics

Of 1878 individuals in the discovery cohort from Cologne, 864 (46.0%) were AMD patients and 1014 (54.0%) were controls. Among AMD patients, 495 patients were categorized as late AMD (57.3% of AMD cases, 26.4% of whole group). Early AMD was present in 202 cases (10.8%), intermediate AMD in 167 cases (8.9%), geographic atrophy (GA) in 53 cases (2.8%), AMD with CNV in 414 cases (22.0%), and mixed type with CNV and GA in one eye, respectively, in 28 patients (1.5%). 

Of 1707 cases in the replication cohort from Nijmegen, 938 (55.0%) were AMD patients and 769 (45.0%) were controls. In the AMD group, 664 patients showed late AMD (70.7% of AMD cases, 38.9% of whole group). Early AMD was present in 147 cases (8.6%), intermediate AMD in 127 cases (7.4%), GA in 47 cases (2.8%), AMD with CNV in 600 cases (35.1%), and mixed type with CNV and GA in one eye respectively in 17 patients (1.0%). 

Clinical characteristics for both cohorts are summarized in Table 1. 

Table 1

 

Clinical Characteristics of the Study Population

	Controls	AMD	Late AMD	Total
Cologne discovery cohort	n = 1014	n = 864	n = 495	n = 1878
Staging, n; %	1014; 54.0	864; 46.0	495; 26.4	1878; 100
Age, y, mean ± SD	70.2 ± 7.7	76.4 ± 8.8	77.6 ± 8.5	73.0 ± 8.8
Female sex, n; %	590; 58.2	528; 61.1	300; 60.6	1118; 59.5
Ever Smoked, n; %	500; 50.0	396; 50.0	227; 52.1	896; 50.0*
Corticosteroid use, n; %	19; 1.9	27; 3.1	14; 2.8	46; 2.4
History of allergy,† n; %	302; 29.8	155; 17.9	59; 11.9	457; 24.3
Pollen allergy	108; 10.7	48; 5.6	18; 3.6	156; 8.3
Drug allergy	70; 6.9	39; 4.5	16; 3.2	109; 5.8
Food allergy	29; 2.9	12; 1.4	2; 0.4	41; 2.2
House dust mite allergy	20; 2.0	9; 1.0	2; 0.4	29; 1.5
Other allergens	67; 6.6	36; 4.2	14; 2.8	103; 5.5
C3d/C3 ratio, n = 1255, median (first–third quartile)	4.34 (3.24–5.76)	4.93 (3.84–6.57)	4.93 (3.91–6.58)	4.65 (3.51–6.23)
Nijmegen replication cohort	n = 769	n = 938	n = 664	n = 1707
Staging, n; %	769; 45.0	938; 55.0	664; 38.9	1707; 100
Age, y, mean ± SD	68.8 ± 7.3	75.4 ± 7.7	76.7 ± 7.6	72.4 ± 8.2
Female sex, n; %	423; 55.0	556; 59.3	388; 58.4	979; 57.4
Ever Smoked, n; %	471; 63.4	565; 64.4	402; 65.5	1036; 63.9*
Corticosteroid use, n; %	35; 4.6	66; 7.0	51; 7.7	101; 5.9
History of allergy, n; %	161; 20.9	143; 15.2	88; 13.3	304; 17.8

*  Smoking status available for n = 1792 in the discovery cohort and n = 1621 in the replication cohort.

†  Subgroup analysis available for n = 438.

Association of Allergy With AMD and Late AMD

Using logistic regression analysis, a positive history of allergy showed a strong protective effect on the development of AMD in the discovery cohort (OR 0.52; 95% CI 0.41–0.64; P = 3.42 × 10⁻⁹). An even stronger protective effect was observed for late AMD (OR 0.32; 95% CI 0.24–0.43; P = 2.57 × 10⁻¹³). There was also a strong effect for the GA group, if analyzed separately (OR 0.09; 95% CI 0.02–0.38; P = 0.001). The protective effect was confirmed in the replication cohort (Table 2). We then analyzed associations between allergy-provoking agents and the development of AMD or late AMD. We found significant associations for all subgroups (pollen, drugs, food, house dust mite, others) although some subgroups were excluded from analysis due to the small number of cases (Table 3). 

Table 2

 

Associations Between Allergy and AMD in the Discovery and Replication Cohorts

	AMD		Late AMD
	OR, 95% CI	P Value	OR, 95% CI	P Value
Unadjusted model
Discovery cohort	0.52, 0.41–0.64	3.42 × 10−9	0.32, 0.24–0.43	2.57 × 10−13
Replication cohort	0.68, 0.53–0.87	0.002	0.58, 0.43–0.77	0.0001
Adjusted model*
Discovery cohort	0.75, 0.59–0.95	0.02	0.49, 0.35–0.69	2.87 × 10−5
Replication cohort	0.76, 0.57–0.99	0.05	0.64, 0.46–0.88	0.007

*  Adjusted for age, sex, use of corticosteroids, and smoking.

Table 3

 

Subgroup Analysis of Allergens in the Discovery Cohort

	AMD		Late AMD
	OR, 95% CI	P Value	OR, 95% CI	P Value
Unadjusted model
Pollen allergy	0.44, 0.31–0.63	7.49 × 10−6	0.27, 0.16–045	5.99 × 10−7
Drug allergy	0.56, 0.37–0.83	0.005	0.37, 0.21–0.65	0.0005
Food allergy	0.41, 0.21–0.82	0.01	*	*
House dust mite allergy	0.45, 0.20–0.99	0.05	*	*
Other allergens	0.54, 0.35–0.82	0.004	0.34, 0.19–061	0.0003
Adjusted model†
Pollen allergy	0.73, 0.50–1.06	0.10	0.46, 0.27–0.80	0.006
Drug allergy	0.75, 0.49–1.17	0.20	0.52, 0.28–0.95	0.03
Food allergy	0.54, 0.26–1.11	0.09	*	*
House dust mite allergy	0.84, 0.37–1.87	0.66	*	*
Other allergens	0.73, 0.47–1.14	0.17	0.55, 0.30–1.02	0.06

*  Not performed due to small group size.

†  Adjusted for age, sex, use of corticosteroids, and smoking.

Association of Allergy With AMD and Late AMD in an Adjusted Model

An additional analysis in an adjusted model, including age, sex, smoking, and use of corticosteroids, affirmed the protective effect of history of allergy on the development of AMD in the discovery cohort (OR 0.75; 95% CI 0.59–0.95; P = 0.018). Analysis of the subgroups of allergens in an adjusted model did not return significant associations (Table 3). 

Applied for late AMD, the adjusted model also confirmed the protective effect of allergy on the development of late AMD with higher significance than for the total AMD group (OR 0.49; 95% CI 0.35–0.69; P = 2.87 × 10⁻⁵). In subgroups of different allergens, a significant association was observed for pollen and drugs (Table 2). The association between allergy and AMD/late AMD in the adjusted model was confirmed in the replication cohort (Table 2). 

Association of C3d:C3 Ratio With AMD and Allergy

The C3d:C3 ratio, a measure for the level of complement activation, was available for 1255 individuals. Median values (first–third quartiles) were 4.65 (3.51–6.23) for the whole group, 4.34 (3.24–5.76) for the control group, 4.93 (3.84–6.57) for the AMD group and 4.93 (3.91–6.58) for the late AMD group (Table 1). C3d:C3 ratio was significantly higher in AMD patients (P < 0.001) as well as in patients with late AMD (P < 0.001) compared with controls (Mann-Whitney U test). The group with positive history of allergy showed a median C3d:C3 ratio of 4.44 (first–third quartiles 3.41–5.75), whereas the group without allergy showed a mean ratio of 4.72 (first–third quartiles 3.55–6.32), which was a statistically significant difference (P = 0.023). However, if taking the phenotype into account, there was a significant difference in C3d:C3 ratios only between the groups with and without AMD, but there was no difference between the groups with and without allergy (Figure), indicating that the difference of C3d:C3 levels depends on the phenotype AMD and not on allergy. 

Figure

 

Systemic complement activation as determined by the ratio of C3d and C3 in the different subgroups of AMD and allergy. Associations between AMD phenotype, history of allergy, and the C3d:C3 ratio were analyzed in the discovery cohort using Mann-Whitney U test.

Discussion

In the current study we describe a protective effect of allergy on the development of AMD with an OR of 0.49 for late AMD after adjustment for confounders. The protective effect was detected for all subtypes of allergy irrespective of the triggering allergen. 

Literature regarding allergy and AMD is limited. History of asthma, which may be induced by allergens, and its relation to AMD showed conflicting results. Klein et al. ¹¹ found no association between asthma and AMD, whereas Sun et al. ¹² found an increased risk for CNV in patients with asthma. This finding was also demonstrated in an animal model, although a meta-analysis of several studies could not confirm this association. 

The role of the complement system in AMD pathogenesis is well established. ¹³ As AMD patients show increased systemic complement activation, the C3d:C3 ratio can be measured as a biomarker in serum. ¹⁰ Recent studies revealed an interaction between allergy and the complement system. ¹⁴ Complement components C3 and C5, also known as anaphylatoxins, play a crucial role in the development of allergy and asthma. ¹⁴ The cleavage products of C3, including C3a, C3d, and C5a, form a set of soluble mediators that bind distinct cell-surface receptors expressed on a variety of target cells. They target a broad spectrum of immune and nonimmune cells and regulate vasodilation. ¹⁵  

In our cohort, AMD patients also had a higher median C3d:C3 ratio than control individuals. However, allergy itself was not associated with increased complement activation. These results indicate that increased complement activation may not be the connecting marker between AMD and allergy. 

Although AMD research has focused on complement activation, there are multiple additional factors contributing to the disease regarding the immune system and inflammation. It is possible that allergy mediates some of these factors. 

For example, drusen in AMD are able to activate the multiprotein complex, encoded by the nucleotide-binding and oligomerization domain–like receptor family, pyrin domain containing 3 (NLRP3) gene, also known as the inflammasome, which induces the secretion of IL-1β and IL-18. These cytokines are linked to apoptotic and cytotoxic effects. ^16–18 Many allergic reactions also have an inflammatory component and the NLRP3 inflammasome influences the development of allergic airway inflammation through the regulation of IL-1β and IL-18. ^19,20 While in dry AMD, NLRP3-mediated secretion of IL-18 was demonstrated to induce retinal pigment epithelium cell death, ²¹ Doyle et al. ¹⁶ identified IL-18 as an inhibitor of CNV in a laser-induced model and as a regulator of vascular endothelial growth factor synthesis. In addition, the cytokine IL-10 limits immune responses and prevents host damage. IL-10, as well as IL-4, IL-5, and IL-13, are implicated in allergy. ²² Bauman et al. ²³ found reduced IL-10 levels in patients with allergic rhinitis, but Mo et al. ²⁴ did not see a difference of IL-10 levels between AMD patients and controls. Taken together, these studies show that inflammatory factors or pathways are simultaneously able to be destructive and protective for AMD progression. Therefore, the counterintuitive finding that allergy has a protective effect on the development of AMD also might be explained by one or multiple of those mechanisms, which have to be investigated in allergic and nonallergic AMD patients and control subjects to identify the pathway for the protective effect of history of allergy on AMD. Another explanation could be that allergy itself is not protective against AMD but that the susceptibility for allergy and the protection against AMD share a common cause. 

A limitation of our study is that the history of allergy was self-reported and was not confirmed by allergy tests. However, our study was well powered, consisting of two large, Caucasian populations with more than 3500 participants and a balanced rate of healthy controls and AMD patients, including many late AMD cases. Furthermore, both cohorts are independent but originate from a relatively similar environment in Western Europe. Additionally, multimodal imaging for AMD staging in EUGENDA avoids misclassification of AMD phenotypes. 

In summary, we identified a positive history of allergy as a protective factor for AMD independent of the underlying allergen. Systemic complement activation did not explain the effect. Further investigations of the conflicting roles of the immune system as a protector and a promoter for AMD are crucial to identify relevant factors and pathways to explain the connection between allergy and AMD, eventually providing new opportunities for the development of therapeutic agents. 

Acknowledgments

Supported by the Retinovit Foundation, Germany. The sponsor or funding organization had no role in the design or conduct of this research. The authors alone are responsible for the content and writing of the paper. 

Disclosure: T. Ristau, None; L. Ersoy, None; Y. Lechanteur, None; A.I. den Hollander, None; M.R. Daha, None; M. Hahn, None; C.B. Hoyng, None; S. Fauser, None