Background

Age-related macular degeneration (AMD) is one of the leading causes of blindness in the aging population [1,2,3,4]. Late-onset AMD can be classified into dry and wet forms. Although intravitreal injection of anti-vascular endothelial growth factor agents is the first-line therapy for AMD, not all patients benefit from this treatment [5]. The underlying mechanism is multifactorial and remains unclear [6]. Recent evidence has suggested a potential influence of ambient air pollutants on AMD risk [7, 8]. Thus, further investigation of this correlation would be clinically meaningful.

Many compounds comprise ambient air pollution, including nitrogen dioxide (NO2), carbon monoxide (CO), sulphur dioxide (SO2), ozone (O3), and particulate matter less than 2.5 and 10 μm in diameter (PM2.5 and PM10, respectively) [9]. Air pollution is a major contributor to the global disease burden and associated with health hazards [10,11,12]. Prior epidemiological studies have demonstrated that ambient air pollution is a potential risk factor for AMD [13, 14]. For example, a national cross-sectional study in China reported a significant positive association between PM2.5 and AMD. For PM2.5, compared with the lowest quartile, the odds ratios (ORs) and 95% confidence intervals (CIs) across increasing quartiles were 0.828 (0.674, 1.018), 1.105 (0.799, 1.528), and 2.602 (1.516, 4.468) [14]. However, conflicting findings were observed on the association between AMD and NO2 or O3 [7, 13]. Further investigation is needed to clarify the correlation between ambient air pollution and AMD. Therefore, we conducted this meta-analysis to report a more robust and reliable outcome.

Methods

Search strategy

This meta-analysis was conducted following the Meta-Analysis of Observational Studies in Epidemiology guidelines. The PubMed, EMBASE, and Web of Science databases were searched for literature using the key words ‘ambient air pollution, particulate matter, ozone, sulphur dioxide, nitrogen dioxide, or carbon monoxide’ and ‘AMD or age-related macular degeneration’ from 01 January 2000 to 30 January 2024. Furthermore, references from original studies or relevant reviews were manually searched to identify other relevant studies. All included studies were epidemiological investigations; the language was restricted to English. Two investigators independently retrieved and reviewed the full texts and abstracts of all related literature. Conflicts were resolved through a full-text review and discussed by two independent reviewers until a consensus was reached.

Inclusion and exclusion criteria

The following inclusion criteria were applied: the studies (1) referred to the association between ambient air pollution and AMD; (2) contained calculable information, such as ORs, hazard ratios (HRs), the respective 95% CIs, and P-values; and (3) were English-only peer-reviewed studies using cross-sectional, prospective, or retrospective cohorts and case–control study designs. The exclusion criteria were as follows: (1) duplicate subjects; (2) abstracts, case reports, comments, reviews, and experimental study designs in laboratory settings; and (3) studies without necessary data.

Data extraction

Two independent investigators (Jiali Wu and Yuzhu Zhang) extracted data from the included studies. The following summary data were included: last author, year of publication, air pollutant(s), statistical model, main results, study cohort, and diagnostic criteria for AMD. If articles reported the OR and P-value instead of the 95% CI, they were manually calculated.

Bias assessment

Two independent authors assessed the study risk using the Newcastle–Ottawa Scale (NOS). Cohort studies that scored ≥ 7, 4–6, and < 4 were considered to have a low, intermediate, and high risk, respectively, whereas cross-sectional studies that scored ≥ 7, 6, and ≤ 5 were considered to have a low, intermediate, and high risk, respectively.

Statistical analysis

Heterogeneity among studies was assessed using χ2-based Q-tests and inconsistency scores (I2). Heterogeneity was high, moderate, low, or none for I2 values ≥ 75%, 50–74%, 25–49%, and < 25%, respectively. Subsequently, random- and fixed-effect models were used based on the heterogeneity test results. The pooled ORs and 95% CIs were calculated to assess the risk of AMD due to ambient air pollution exposure. Regarding pooled outcome analyses, P < 0.05 was considered significant. All analyses were performed using RevMan 5.3. software (Review Manager, Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark).

Results

Study selection and characteristics

Figure 1 shows a flow diagram of the literature search. The initial search identified 358 articles from databases. After screening titles and abstracts combined with necessary full-text review, eight studies with 15,029,888 individuals were eligible for further analysis. Among these, four studies analysed PM2.5 [8, 13,14,15], and three examined NO2 [7, 13, 16] or O3 [16,17,18]. Two of the eight studies were longitudinal cohort studies [7, 8], while the rest were cross-sectional studies [13,14,15,16,17,18]. In the longitudinal cohort study that explored the association between the AMD risk and CO or NO2, 1,442 individuals among 39,819 AMD-free residents developed AMD during the study period of 11 years [7]. Moreover, 4,284,128 participants enrolled in the longitudinal cohort study evaluating the relationship between PM2.5 and AMD, and 12,095 AMD cases were identified during the 11-year follow-up [8].

Fig. 1
figure 1

Flow diagram of the included studies

Full size image

The assessment and definition of AMD varied among studies. Among six studies applying standardised criteria for AMD diagnosis, grading was performed by at least two independent ophthalmologists to ensure the accuracy of diagnosis in three studies [8, 16, 17], whereas the other three did not describe the specific grading methodology [7, 14, 18]. Different from the standardised criteria, cases were diagnosed based on medical record review or self-reporting in two studies [13, 15]. Two AMD stages (early and late) were analysed separately in one study [16]. One study assessed AMD in patients with and without visual impairment [15]. One study explored the relationship of air pollution exposure with exudative and non-exudative AMD [18].

Table 1 summarizes the results of the individual studies in this meta-analysis. Liang et al. presented a single-pollutant model, proposing an increased risk of AMD among those exposed to the highest exposure quartile of CO (HR = 1.84; 95% CI: 1.57–2.15) and NO2 (HR = 1.91; 95% CI: 1.64–2.23). The risk did not increase in the second or third quartiles, indicating that moderate exposure did not lead to AMD [7]. Second, Freeman et al. reported that those exposed to higher PM2.5 levels were more likely to develop visually impaired AMD, as indicated by a single-pollutant regression model (OR = 1.52; 95% CI: 1.10–2.09). However, in a multipollutant model, higher exposure to PM2.5 merely showed a borderline association with visual impairment in AMD (OR = 1.41; 95% CI: 0.96–2.08, P = 0.08) [15]. In addition, single-pollutant model findings by Patel et al. demonstrated increased odds of AMD among participants exposed to higher levels of PM2.5 (OR = 1.08; 95% CI: 1.01–1.16). However, they did not find an association between exposure to PM10 (OR = 0.94; 95% CI: 0.86–1.02) or NO2 (OR = 0.99; 95% CI: 0.91–1.08) and AMD [13]. Furthermore, Choi et al. modelled air pollution within administrative division units, which suggested that NO2 (OR = 1.24; 95% CI: 1.05–1.46) and CO (OR = 1.22; 95% CI: 1.09–1.38) were risk factors for AMD, but O3 was associated with a decreased prevalence of AMD (OR = 0.8; 95% CI: 0.70–0.92). When air pollution was modelled as local/town units, the associations were slightly diminished. Besides, the study showed that higher levels of CO exposure led to higher prevalence of AMD [16]. However, the findings by Manookin et al., which were diametrically different from those by Choi et al., revealed that O3 was the only air pollution that statistically significant associated with any AMD (OR = 1.011; 95% CI: 1.003–1.019). It was demonstrated that NO2, SO2, CO and PM2.5 did not increase the AMD risk, whereas the exact effect size of them was not reported [18]. Both Hwang et al. and Yan et al. found a significant positive association between AMD and higher PM2.5 levels (HR = 1.19; 95% CI: 1.13–1.25 and OR = 2.602; 95% CI: 1.516–4.468, respectively) [8, 14]. Recently, multipollutant model findings by Sun et al. have revealed that the AMD risk was on a monotonic increasing trend with higher O3 concentration; the harmful effect increased rapidly after reaching a turning point of 110 μg/m3 (OR = 1.15; 95% CI: 1.13–1.16 and OR = 1.66; 95% CI: 1.63–1.69, respectively) increase [17]. Based on the results above, most studies revealed that higher concentrations of air pollution increased the AMD risk. However, no studies examined the correlation between the AMD risk and the exposure length.

Table 1 Characteristics of the studies included in the systematic review
Full size table

Quantitative synthesis

We used the data of single-pollutant models for effect estimates. Two articles calculated and divided air pollution concentrations into four quartiles; we used the data of the highest quartile in the meta-analysis [7, 14]. According to quantitative methods in the review of epidemiologic articles, we ignored the distinctions among the HRs and ORs and calculated the pooled OR [19].

Figure 2 illustrates the association between PM2.5 and AMD; the pooled OR was 1.16 (95% CI: 1.11–1.21, I2 = 82%, P = 0.0009). Figure 3 shows a forest plot of the outcomes of three studies on the relationship between NO2 and AMD; the pooled OR was 1.17 (95% CI: 1.09–1.25, I2 = 96%, P < 0.00001). Figure 4 illustrates the relationship between O3 and AMD; the pooled OR was 1.06 (95% CI: 1.05–1.07, I2 = 100%, P < 0.00001). All results demonstrated a positive relationship between ambient air pollution and AMD, as well as indicated high heterogeneity.

Fig. 2
figure 2

Forest plot of the association between PM2.5 and AMD risk. The pooled OR is 1.16 (95% CI: 1.11–1.21, I2 = 82%, P = 0.0009), demonstrating a positive relationship between PM2.5 and AMD. Abbreviations: PM2.5, particulate matter less than 2.5 μm in diameter; AMD, age-related macular degeneration; OR, odds ratio; CI, confidence interval; I2, inconsistency score

Full size image
Fig. 3
figure 3

Forest plot of the association between NO2 and AMD risk. The pooled OR is 1.17 (95% CI: 1.09–1.25, I2 = 96%, P < 0.00001), indicating a positive relationship between NO2 and AMD. Abbreviations: NO2, nitrogen dioxide; AMD, age-related macular degeneration; OR, odds ratio; CI, confidence interval; I2, inconsistency score

Full size image
Fig. 4
figure 4

Forest plot of the association between O3 and AMD risk. The pooled OR is 1.06 (95% CI: 1.05–1.07, I2 = 100%, P < 0.00001), indicating a positive relationship between O3 and AMD. Abbreviations: O3, ozone; AMD, age-related macular degeneration; OR, odds ratio; CI, confidence interval; I2, inconsistency score

Full size image

Evidence evaluation

Due to the small number of studies eligible for analysis, a statistical evaluation of publication bias was not feasible [20]. NOS was used to assess the quality of the included studies (Additional file 1); one study was classified as having moderate quality [15], whereas the others were classified as having high quality [7, 8, 13, 14, 16,17,18].

Discussion

AMD is a progressive retinal disease associated with photoreceptor atrophy and degeneration of the retinal pigment epithelium with a high prevalence and limited therapeutic benefits [4, 5, 21, 22]. The mechanisms underlying AMD are multifactorial. Smoking is a detrimental factor for AMD [23, 24]. Thus, air pollutants, to which the outer eye segment is directly exposed, might also be potential risk factors for eye diseases. Previous studies have shown that air pollution has detrimental effects on ocular surface and increases the risk of dry eye disease and allergic conjunctivitis [25,26,27]. Recently, researchers have suggested that air pollution may also affect the inner eye segment. The relationships between air pollution and AMD, glaucoma, cataract, and diabetic retinopathy have also been investigated [28,29,30,31,32,33]. However, the results remain inconclusive. Therefore, we reviewed the existing studies and performed a meta-analysis to clarify this relationship. Notably, various studies have relatively consistently demonstrated a correlation between AMD and PM2.5. However, the results of individual studies exploring the exact association between AMD and NO2 or O3 are contradictory. Our meta-analysis indicated that three air pollutants increased the risk of AMD. Merely one study explored the relationship between SO2 and AMD. Additionally, two studies on the association between CO and AMD showed consistent findings that CO increased the AMD risk. For the reasons above, we did not analyse the association between the AMD risk and SO2 or CO.

Several potential mechanisms could explain these associations. As chemical components of air pollution, CO, NO2, O3, SO2, and PM2.5 share a common biological pathway known to induce oxidative stress and inflammation, which are recognised as AMD risk factors [34, 35]. Moreover, animal studies have demonstrated that PM2.5 can impair microvascular function [36]. In the eye, choroidal microcirculation deterioration plays a critical role in AMD [37]. Chua et al. also proposed that exposure to PM2.5 is associated with adverse retinal structural features, which may lead to AMD [13, 38]. In addition, PM2.5 can cause neurodegenerative diseases, including reduced cognitive function [39, 40], accelerated cognitive decline [41], Parkinson’s disease, and Alzheimer’s dementia [42]. Given that AMD is a neurodegenerative disease, these studies further justify the plausibility of a correlation between AMD and air pollution.

It should be noted that meterological variables have a significant effect on changes in air pollution. Yan et al. reported that combined exposure to PM2.5 and atmospheric pressure remarkably increased the risk of AMD, while temperature and humidity acted a weakly antagonistic effect on AMD [14]. Higher temperature is known to cause lower relative humidity. Moreover, both air temperature and atmospheric pressure affect the distribution and concentration of PM2.5 [14, 43]. At present, limited studies elucidate the joint effects of meterological factors and ambient air pollution on AMD. Therefore, further studies are needed to clarify this correlation.

Our meta-analysis has certain limitations. First, only a few studies were eligible for inclusion owing to the novelty of this topic. However, most of the included studies had large sample sizes and all were considered credible. Second, the standard definition of AMD was inconsistent among the studies. In addition, the measurement variability of the exposure assessment and the quantitative differences of the exposure extent could bias our findings. Moreover, we used the data of the highest quartile of two articles in the meta analysis and did not standardize the effect size across the studies. The small amount of studies for every air contaminant made subgroup analysis by the exposure level difficult. Finally, other pollutants can affect single-pollutant models. Multipollutant models are less likely to be affected by confounding factors but may be susceptible to other biases.

Conclusions

The current evidence suggests that ambient air pollutants, such as PM2.5, NO2, and O3, detrimentally affect AMD. Extensive studies are urgently required to investigate additional air pollution and their influence on AMD or other ocular diseases. Further strategies for reducing ambient air pollution exposure are essential for public health, which may ultimately mitigate AMD.