2.5.4. GPx Activity

A reaction mixture was prepared (phosphoric buffer, glutathione reductase, GSH, NADPH + H+, and (haemo)lysate with transforming reagent incubated for 5 min at room temperature beforehand; transforming reagent used only for RBC) and incubated for 10 min at 37 ◦C. After the incubation, the reaction was initiated by adding tert-butyl hydroxide (or hydrogen peroxide), and the decrease in the absorption at λ 340 nm was measured. The amount of enzyme that oxidised 1 µmol of GSH (0.5 µmol NADPH + H) in one minute was defined as a unit of enzyme activity.

### 2.5.5. GST Activity

A reaction mixture was combined (phosphoric buffer, GSH, CDNB (1-chloro-2,4-dinitrobenzene), and (haemo)lysate), and the increase in absorbance at λ 340 nm was measured. Glutathione transferase activity was determined using molar absorption coefficient of the synthesised conjugate (e = 9600 M−<sup>1</sup> cm−<sup>1</sup> ).

### 2.5.6. R-GSSG Activity

The (haemo)lysate was diluted, mixed with 1 mL of a diluted RI working reagent (900µL EDTA and 100µL RI; RI: NADPH<sup>+</sup> + H<sup>+</sup> diluted in 0.01 M NaOH, Sigma Aldrich, St. Louis, MO, USA) and incubated (5 min, 30 ◦C). Then, RII reagent was added (GSSG (glutathione disulphide) diluted in EDTA) and extinction was measured at λ 340 nm over 3 min at 30 ◦C.

### 2.5.7. Statistical Analysis

The quantitative parameters measured in both eyes were averaged before further analysis. The distributions of all the analysed variables related to components of antioxidant system were right-skewed and significantly different from normal distribution (*p*<0.05, Shapiro-Wilk test). Therefore, the nonparametric Kruskal-Wallis and Mann-Whitney tests were used to compare quantitative values between groups, whereas Spearman's rank correlation coefficient (Rs) was calculated to measure the strength of associations between these values. We used Fisher's exact test to compare qualitative variables between groups. A multivariate analysis of AMD as an independent variable was performed using a general linear model (GLM) adjusted for age, sex and smoking status (pack-years) by inclusion of the confounding factors as independent variables, with logarithmic transformation applied to the

dependent variables to normalize their distributions. Standardised regression coefficients (β) were calculated to measure the strength of associations between independent and dependent variables. The interpretation of β and Rs coefficients is similar: values +1 and −1 indicate perfect positive and negative association, respectively, while 0 indicates complete lack of association. Quantitative variables were presented as mean ± standard deviation. *p* < 0.05 was considered statistically significant without correction for multiple testing other than included in the applied test itself (number of compared groups in Kruskal-Wallis test, number of variables in GLM). Statistica 13 software (Dell Inc., Round Rock, TX, USA) was used for statistical analyses.

#### **3. Results**

#### *3.1. Characteristics of the Study Subjects*

We enrolled 330 patients with AMD and 121 healthy controls in this study. The clinical characteristics of the studied groups are presented in Table 1. The AMD and control groups did not differ in age and well-known atherosclerotic risk factors, including serum lipid and glucose levels. Statistically significant differences in physical activity were not observed between the groups. On the other hand, the number of past smokers and the number of pack-years of smoking were considerably higher in the AMD group than in controls (*p* < 0.001). Importantly, a strong positive correlation between the number of pack-years of smoking and disease severity was identified (Rs = +0.23, *p* < 0.001), corroborating the well-documented relation between smoking and AMD. The positive correlation between the patient's age and clinical classification of AMD (Rs = +0.19, *p* < 0.001) indicates that age is one of the main factors affecting disease severity.


**Table 1.** Characteristics of the study groups. In bold, *p*-value < 0.05, which was considered statistically significant.


**Table 1.** *Cont.*

\* Mann—Whitney test/Fisher's exact test.

On the other hand, the HDL level correlated negatively with disease severity (Rs = −0.15, *p* = 0.007), suggesting its protective effect on AMD progression. Another negative correlation was observed between the education level and disease severity (Rs = −0.14, *p* = 0.01). Thus, a higher educational attainment might be associated with better health awareness and subsequently reduce AMD progression.

#### *3.2. Analysis of Lifestyle Habits*

Since diet is considered one of the potentially modifiable risk factors for AMD, we aimed to compare the dietary habits of patients with AMD and healthy controls (Table 2) and to analyse whether dietary habits differed between the early, intermediate and late AMD groups (Table 3).

Patients with AMD consumed much less fatty fish than the control group (*p* = 0.008). Similar findings were observed for egg consumption (*p* = 0.04). In contrast, the AMD group recorded higher consumption of fruits and fruit juices than the controls (*p* = 0.01). We did not observe significant differences in the consumption of green vegetables, omega-3 rich oils, and simple and complex carbohydrates between groups (*p* = 0.39, *p* = 0.66, *p* = 0.18, and *p* = 0.26, respectively). Accordingly, no differences in alcohol intake were observed between groups. Interestingly, we observed a significantly higher consumption of fatty fish in patients with intermediate AMD than in patients with the late form of the disease (*p* = 0.01). Significant differences in egg, green vegetable, fruit and fruit juice, omega-3-rich oil, simple and complex carbohydrate consumption and alcohol intake were not observed between the early, intermediate and late AMD groups.

We aimed to determine whether physical activity and the consumption of specific food groups or alcohol were associated with disease severity to better assess the roles of lifestyle and dietary habits in the progression of AMD. In the AMD group, we observed strong positive correlations between VA and the total physical activity MET score (Rs = +0.17, *p* = 0.003), but these correlations were not observed in controls. Positive correlations between VA and physical activity intensity were clearly detectable in the nAMD group, including the intense physical activity MET score (Rs = +0.17; *p* = 0.04), average physical activity MET score (Rs = +0.21; *p* = 0.01) and total physical activity MET score (Rs = +0.22, *p* = 0.006). Thus, patients with AMD who were more physically active displayed better visual function. Accordingly, the total physical activity MET score negatively correlated with the severity of AMD (Rs = −0.14, *p* = 0.01). Similarly, time (in min) spent sitting in the last 7 days was associated with more advanced stages of AMD (Rs = +0.20, *p* = 0.0005). Based on these findings, sedentary behaviour facilitates the progression of AMD.


**Table 2.** Dietary habits of the AMD group and control group. *p*-values < 0.05, which were considered statistically significant, are shown in bold.

\* Mann–Whitney test.

Regarding dietary factors, VA correlated positively with green vegetable consumption (Rs = +0.24, *p* = 0.004) and omega-3-rich oil intake (Rs = +0.17, *p* = 0.03) in the nAMD group. Therefore, the consumption of these food products might preserve visual function in patients with nAMD, although no differences in the consumption of those products were observed between the AMD and control groups. Accordingly, fatty fish consumption correlated positively with the retinal volume in the AMD group (Rs = +0.23, *p* = 0.003). We observed negative correlations between the drusen size and consumption of green vegetables (Rs = −0.19, *p* = 0.02), fruit and fruit juice (Rs = −0.28, *p* = 0.0004), omega-3-rich oils (Rs = −0.24, *p* = 0.002) and complex carbohydrates (Rs = −0.21, *p* = 0.01) in the AMD

group. These correlations were not observed in the control group. Thus, the consumption of these products might exert beneficial effects on the disease course and reduce AMD progression.


**Table 3.** Dietary habits of the early, intermediate and late AMD groups. The bold font indicates *p*-values < 0.05, which were considered statistically significant.

<sup>1</sup> Kruskal-Wallis test, <sup>2</sup> Mann-Whitney test.

### *3.3. Components of the Antioxidant System*

Excess oxidative stress coupled with overwhelmed antioxidant defence systems are thought to be the important contributors to the complex pathophysiology of AMD. We chose 6 factors to analyse the efficiency of the antioxidant system in patients with AMD and controls: the activities of five enzymes (SOD, CAT, GPx, R-GSSG and GSH transferase) and concentrations of reduced glutathione (GSH) in red blood cells (RBCs) and platelets (PLT). The AMD group presented higher values for 7 of the 12 tested factors compared with the control group (Table 4, Figure 1): GPx, R-GSSG, GSH transferase in RBCs and SOD, catalase, GPx and R-GSSG in PLT. A significant downregulation in catalase activity levels was observed in RBCs from patients with AMD (0.33 ± 0.21 U/mg Hb) compared to controls (0.53 ± 0.24; *p* < 0.0001). A multivariate analysis of patients and controls adjusted for age, sex, and smoking status (pack-years) revealed that AMD was an independent variable associated with a lower RBC catalase (β = −0.37, *p* < 0.001) and higher PLT catalase (β = +0.25, *p* < 0.001), RBC GPx (β = +0.26, *p* < 0.001), PLT GPx (β = +0.16, *p* = 0.001), RBC R-GSSG (β = +0.13, *p* = 0.009), PLT R-GSSG (β = +0.12, *p* = 0.02) and RBC GSH transferase (β = +0.23, *p* < 0.001) activity levels.

shown in bold.


**Table 4.** Comparison of levels of components of the antioxidant system in patients with AMD and controls. The bold font indicates *p*-values < 0.05, which were considered statistically significant.

\* Mann-Whitney test.

*Antioxidants* **2020**, *9*, x FOR PEER REVIEW 2 of 23

**Figure 1.** Comparison of the antioxidant system component levels in patients with AMD and controls. *p*-values < 0.05, which were considered statistically significant, are **Figure 1.** Comparison of the antioxidant system component levels in patients with AMD and controls. *p*-values < 0.05, which were considered statistically significant, are shown in bold.

An analysis stratified by AMD severity revealed that the early AMD group presented higher GSH (RBC) concentration and lower R-GSSG (PLT) activity than the late AMD group (*p* = 0.03 and *p* = 0.04, respectively) (Table 5). However, we should be rather careful in interpreting the results of the Mann–Whitney test, since the Kruskal-Wallis test showed no statistically significant differences between the early, intermediate and late AMD groups.




Next, we investigated the associations between ophthalmic parameters and concentrations of the analysed antioxidants to more specifically evaluate the functions of antioxidants in the development of AMD. In the AMD group, RBC catalase activity and GSH concentrations negatively correlated with the disease severity (Rs = −0.11, *p* = 0.04; R = −0.11, *p* = 0.05, respectively). This relationship corresponds to lower RBC catalase activity in the AMD group than in controls. Similarly, we observed weak positive correlations between the clinical classification of AMD and RBC GPx (Rs = +0.10, *p* = 0.07), PLT catalase (Rs = +0.10, *p* = 0.08) and R-GSSG PLT (Rs = +0.10, *p* = 0.08) activities that corresponded to higher activities of these enzymes in RBCs and PLT from patients with AMD. Accordingly, the drusen size in patients with AMD correlated positively with SOD, GPx and GSH transferase activities in RBCs (Rs = 0.31, *p* < 0.001; Rs = 0.16, *p* = 0.003; and Rs = 0.18, *p* < 0.001, respectively) and negatively correlated with both RBC activity and PLT GSH concentration (Rs = −0.27, *p* < 0.001; and Rs = −0.24, *p* < 0.001, respectively). This association was similar to the higher RBC GPx and GSH transferase activity in the AMD group than in controls. Based on these results, the antioxidant system might be a major contributor to the clinical course of AMD.

#### *3.4. Correlations between the Antioxidant System and Lifestyle Factors*

Different nutritional factors are proposed to modulate the antioxidant potential of various cells [38,39]. We evaluated the possible associations between the activity levels of components of the antioxidant system with physical activity and diet to assess whether the activities of the investigated antioxidant enzymes in RBCs and PLT were associated with lifestyle factors and whether these correlations were specific for patients with AMD. Overall, GPx, R-GSSG and GSH transferase activities in RBCs from the AMD group correlated negatively with egg consumption (Rs = −0.22, *p* < 0.001; Rs = −0.17, *p* = 0.003; and Rs = −0.13, *p* = 0.03, respectively), whereas the RBC catalase activity was positively correlated with the amount of egg consumption (Rs = +0.11, *p* = 0.05). Similar correlations were not observed in controls. These relationships clearly correspond to higher GPx, R-GSSG and GSH transferase activities and lower catalase activity in the AMD group, as well as lower egg consumption, than in controls. Similarly, in the AMD group, positive correlations between RBC catalase activity and fatty fish consumption (Rs = +0.18, *p* = 0.001) paralleled the lower consumption of this dietary product and lower catalase activity in patients with AMD than in controls. Accordingly, we observed positive correlations between fruit and fruit juice consumption with two PLT enzymes: SOD and GPx (Rs = +0.12, *p* = 0.03 and Rs = +0.14, *p* = 0.02, respectively). This finding confirms the aforementioned observation, since the activities of both of these enzymes and fruit and fruit juice consumption were higher in the AMD group than in the control group. The identified relationships indicate the possible effect of dietary habits on antioxidant activity in patients with AMD.

In terms of physical activity, R-GSSG activity levels in RBCs correlated negatively with the total physical activity MET score (Rs = −0.17, *p* = 0.04) in patients with nAMD, but this correlation was not observed in controls.

#### *3.5. Genotypes and Components of the Antioxidant System*

We also explored the relationship between antioxidant activity and genetic risk factors for AMD to further elucidate the interactions between various risk factors for AMD and their contributions to disease pathogenesis. For this purpose, we investigated associations between the six selected antioxidants and polymorphisms in genes previously associated with AMD: *CFH* Y402H, *ARMS* A69S and a single nucleotide variant that our team recently linked to a higher AMD risk, *PRPH2* c.582-67T > A (rs3818086) (paper in press). However, when the correction for multiple testing was applied, we did not identify any statistically significant relationships between components of the antioxidant system and the genotypes of these genes.

#### **4. Discussion**

AMD remains the major cause of visual impairment among the elderly population, significantly reducing the quality of life of affected individuals [40]. Several genetic and environmental factors, including the *CFH* Y402H polymorphism, age and cigarette smoking, have been identified as contributing to the complex landscape of AMD, although the exact pathogenesis of the disease remains unclear [41,42]. At the molecular level, various risk factors for AMD share a common denominator, oxidative stress, which is thought to be the main component of AMD pathology [27,43]. In fact, manipulations of dietary and lifestyle habits, which are thought to contribute to the tight balance of the endogenous antioxidant system, might be beneficial in preventing and/or slowing the progression of AMD [11]. Thus, in the present study, we aimed to investigate the role of antioxidant components in AMD and to assess whether dietary and lifestyle factors modulate the levels of those endogenous antioxidants and clinical parameters of disease severity. We also assessed possible relationships between antioxidant activity and genetic risk factors for AMD.

First, we assessed the systemic levels of components of the antioxidant system in peripheral blood and found that the activity of the majority of tested substances were significantly increased in patients with AMD (GPx, R-GSSG, and GSH transferase levels in RBCs and SOD, CAT, GPx, and R-GSSG levels in PLT), whereas only the CAT activity in RBC was evidently reduced in patients with AMD compared with controls. Our observation of increased GPx activity is in contrast to the results reported by Mrowicka et al. [44] and Plestina-Borjan et al. [31], where significantly lower GPx (RBC) activity was observed in AMD patients in comparison with controls. On the other hand, in the large POLA study of a cohort of 2584 participants, the increased levels of plasma GPx concentration, which catalyses H2O<sup>2</sup> degradation by GSH, correlated with a nine-fold increase in the prevalence of late AMD [45]. GPx not only protects RPE cells in models of oxidative damage-induced retinal degeneration but is also required for the maturation of photoreceptor cells [46]. As proposed by Tokarz et al., the increased activity of GPx in patients with AMD reflects the activity of RPE cells, which attempt to dispose of overwhelming amount of H2O<sup>2</sup> formed during the disease course [47]. R-GSSG interacts with GPx to regulate the GSH concentration, as it converts glutathione disulfide (GSSG) to GSH [48]. In contrast to our results, the R-GSSG concentration and activity have been reported to be rather low in patients with AMD and was associated with a decrease in GSH levels [49,50], the product of enzymatic reaction catalysed by R-GSSG. In our study, we did not observe a significant association between R-GSSG activity and GSH concentration in any of the tested groups; thus, the increase in R-GSSG activity in patients with AMD might be an analogous indicator of increased antioxidant activity, similarly to GPx or increased GSH transferase activity. Interestingly, conflicting results on the involvement of *GSTM1* and *GSTM5* (glutathione s-transferase mu 1/5) polymorphisms and gene expression in AMD pathology exist [51,52], once again suggesting that the necessary balance of the antioxidant system is achieved through the proper activity of several enzymes, and not a single enzyme. The tight cooperation of endogenous antioxidants is reflected in SOD activity, which functions together with GPx and CAT to convert H2O<sup>2</sup> to nontoxic products and by that protect the photoreceptors and RPE from oxidative damage [31,49]. SOD activity decreases in the RPE periphery with ageing and at the same time its immunoreactivity increases [53]. However, in the aforementioned POLA study, a high level of erythrocyte SOD activity was not associated with AMD [45]. This finding is consistent with our results, although we also observed a higher activity of SOD in PLT from the AMD group. On the contrary, Venza et al. reported lower SOD (both in plasma and RBC) activity in AMD patients compared to controls [54]. Interestingly, in vitro studies have shown a reduction in SOD activity in response to oxidative stress when ARPE-19 cells were treated with acrolein, a powerful initiator of oxidative stress and mitochondrial dysfunction [55], whereas the upregulation of *SOD1*/*2* expression resulted in oxidative damage in RPE cells [56]. In fact, excess SOD (in relation to the activities of GPx and CAT) may cause damage [57], which further suggests the need for a tight balance of the antioxidant system. Similar to SOD, CAT function decreases in the macular and peripheral RPE with ageing [58], but in contrast to SOD, CAT immunoreactivity is reduced in RPE cells in the eyes of patients with and

without AMD [47,59]. Our finding of reduced CAT activity in patients with AMD is consistent with previous reports [44]. As proposed by Tate et al., treatment with ROS-generating compounds induces CAT expression in RPE cells, which protects against H2O2, even in the adjacent RPE cells without upregulated CAT expression. Overall, our findings of increased activity of several antioxidants in the AMD group suggest an enhanced response to oxidative damage that might contribute to AMD pathogenesis by disrupting the tight balance of the antioxidant system [32].

Dietary antioxidants aid the endogenous antioxidant defence system in protecting against oxidative damage and enhanced ROS production and consequently, prevent or slow related disorders, including AMD [60]. Several major clinical trials, in particular the Age-Related Eye Disease Study (AREDS) and AREDS2, have shown that nutrients with antioxidant properties, namely, lutein, zeaxanthin, polyunsaturated omega-3 fatty acids (PUFAs), zinc, vitamins C and E, delay the progression of advanced AMD in persons with intermediate AMD [13,61]. In our study, patients with AMD consumed much lower levels of fatty fish and eggs than controls, whereas greater consumption of green vegetables and omega-3-rich oils was correlated with favourable clinical outcomes (better visual acuity and a smaller drusen size in patients with AMD). These observations are consistent with previous studies, as patients with AMD are generally encouraged to increase their intake of green vegetables, eggs and fish [9,14,33]. Both green vegetables and eggs are rich sources of lutein and zeaxanthin, potent anti-inflammatory and antioxidant factors that exert beneficial effects on slowing the progression of AMD [15]. The antioxidant potential of these macular carotenoid pigments combined with their ability to filtrate blue-light may serve not only to protect the ocular tissue from oxidative damage, but also to improve visual acuity [62]. Our study further supports this notion, as we found a positive correlation between VA and consumption of green leafy vegetables, which are a well-known source of these potent macular carotenoid pigments [63]. According to Gopinath et al., eggs also contain large quantities of selenium, which directly protects cells from oxidative damage [64]. Additionally, eggs, similar to fatty fish, are a good source of omega-3 PUFAs, which may minimize retinal inflammation, oxidation, and degeneration [10,64,65]. Indeed, the Blue Mountains Eye Study (BMES) of a large Australian cohort has shown that a greater consumption of fish and omega-3s may slow AMD progression [66]. Interestingly, we did not observe any correlations between alcohol consumption and the clinical parameters in our patients, although at the molecular level, the toxicity of alcohol is associated with lipid peroxidation and oxidative stress and potentially represents another causative factor for AMD [67]. This finding is consistent with a study by Knudtson et al. [68], although some recent reports suggest a modest association between alcohol consumption and an increased AMD risk [69].

The preventive strategies incorporating a modification of the diet represent an attractive approach to slow AMD progression, but some debate exists in the literature regarding whether physical exercise is also recommended to protect against AMD [70]. Our study provides a clear indication that a sedentary lifestyle worsens the AMD course, as more physically active patients presented better visual acuity. Several previous studies have associated physical activity with a lower risk of AMD [71,72], but reports have also described the lack of significant relationships between exercise and AMD risk [73]. As shown in a recent study by Gopinath et al., the most physically active patients aged at least 75 years are 79% less likely to develop late AMD in 15 years [70]. However, when other confounding factors were considered, no significant association was observed in this group. Overall, the systemic benefits of physical activity (e.g., protective effects on obesity, diabetes, inflammation, etc.) make it a vital part of a healthy lifestyle [70,74], which should be recommended to patients with AMD, along with refraining from cigarette smoking and consuming a diet rich in vegetables, fish and eggs.

It is worth noting; however, that the correlations found in the present study do not necessarily indicate causation. Thus, further studies on larger cohorts are needed to provide more straightforward evidence of various dietary and lifestyle factors affecting AMD course.
