Therefore, to identify depressive symptoms and QOL among Korean AMD patients and to also identify associated factors that predispose AMD patients to develop depression, we used the data acquired from the fifth Korea National Health and Nutritional Examination Survey V-2 (KNHANES V-2) performed to sample households that represent the entire nation of Korea in 2011.
Discussion
Using data acquired from the 2011 KNHANES V-2 nationally representative survey, we determined that the prevalence of depressive symptoms in Korean AMD patients was 19.8%. We observed that the incidence of depressive symptoms in AMD patients was higher than that of non-AMD aging controls (19.8% vs. 14.4%,
p = 0.013). This finding is consistent with several previous epidemiologic studies that have suggested that the prevalence of depression is higher in patients with AMD than in non-institutionalized older persons, whose rate of depression ranges between 8% and 16% [
23]. The prevalence of depression in the geriatric population in Korea is generally within the range of 4.2% to 13.3% [
24,
25]. However, the incidence of depression in the Korean population in the present study was slightly lower than that identified by previous studies in other populations (32.5%-33.0% in the United States [
7,
9], 21.3% in Canada [
8], and 26.4% in India [
6]). This finding may be partially explained by differences in the study designs and the population samples. Comparing the prevalence of depressive symptoms between different studies is often difficult because the study populations, depressive symptom measurement techniques, study methodologies, and definitions of depression are different. The present study was open to all patients with AMD in the community, and early AMD was included; other reports have limited their sample to patients recently diagnosed in outpatient retina clinics [
8], those with a reduced baseline visual acuity [
9], or patients who have suffered acute vision loss [
7].
We also identified female gender, being in the ADL “dependent” group, and having “some problems” on the “anxiety/depression” dimension of the EQ-5D as independent associated factors for depressive symptoms in AMD patients.
Generally, the interrelationship between gender and depressive symptoms is well known. The prevalence of depression in Korea is significantly higher in women than in men [
26]. However, previous studies of depression in AMD patients from Western populations have reported that gender is not significantly associated with depression in AMD patients [
7,
8,
9]. Racial/ethnic differences between East Asian and Western populations may explain these disparate results.
In the present study, dependency in ADL was associated with a four-fold increase in the odds of depressive symptoms in AMD. This finding may be partially explained by the fact that a loss of vision can significantly disrupt daily life or social activity. The results of the EuroQol support this hypothesis, which showed that AMD patients were more likely to have problems with usual activities and to experience pain/discomfort than non-AMD controls. These results suggest that simply asking AMD patients about difficulty with ADL during ophthalmologic evaluations can be a valuable screening technique to identify patients at risk of depressive symptoms.
Our previous study indicated that increasing age was the most important factor associated with depression in patients with late AMD treated with intravitreal ranibizumab [
10]. However, in the present population-based study, increasing age was not significantly associated with depressive symptoms in Korean AMD patients. Therefore, this finding was inconsistent with previous results. This discrepancy may be explained by differences in the study designs. The present study included early AMD as well as late AMD, whereas the previous study limited the sample to late AMD patients who had been treated with intravitreal ranibizumab. According to a previous epidemiologic study in Korea, the mean age of early AMD patients also tends to be younger than that of late AMD patients, which might have also affected the results [
4].
In the current study, we expected that ophthalmic factors, including visual acuity and AMD grade, would show a strong association with depressive symptoms in AMD patients. However, the results revealed non-significant relationships between these ophthalmic factors and depressive symptoms. Previous studies have reported conflicting results with regard to visual acuity. In the literature, two studies found weak or non-significant correlations between visual acuity and depression [
7,
9], while two other reports concluded that visual acuity is an important determinant of depression [
6,
8]. This discrepancy may be due to differences in the sample selection process. Grade of AMD (early or late) was also not significantly associated with depressive symptoms in the present study. According to previous studies that compared Asian and Western populations, the prevalence of late AMD in Asians was comparable with that reported for Western patients, but early AMD was less common among Asians [
3,
4]. This racial/ethnic difference needs to be considered when interpreting these results.
To determine if the associations between female gender and dependency in ADL with depressive symptoms were specific to patients with AMD, we additionally performed a multivariate logistic regression analysis of non-AMD patients and found similar associations. Female gender and dependency in ADL were also associated with depressive symptoms in non-AMD patients (aOR, 1.785; 95% CI, 1.298 to 2.456; aOR, 2.250; 95% CI, 1.619 to 3.126, respectively).
Although these factors were not specific only in AMD patients, the results of our study can be valuable. When ophthalmologists encounter AMD patients with these associated factors, there should be more careful consideration of any depressive symptoms that may be present. Additionally, it is important to recognize that all AMD patients, regardless of the stage and degree of visual acuity, are at risk for difficulty with ADL and therefore an increased incidence of depressive symptoms. Therefore, all AMD patients should be screened for both ADL and depressive symptoms.
In the current study, health-related QOL showed no difference between AMD patients and non-AMD controls. Although general health-related QOL assessments provide a good hallmark for comparison of QOL across diseases, instruments developed specifically to measure vision-specific QOL may provide improved accuracy for assessing the impact of vision-threatening disease [
11,
12,
13,
14]. Future population-based studies using vision-specific QOL instruments are needed to investigate QOL in AMD patients.
When the current study began, we expected that a poor QOL (lower EQ-5D and EQ-VAS scores) would show a strong association with depressive symptoms in AMD patients. A simple comparison in AMD patients revealed significant differences for all EuroQol indices between people both with and without depressive symptoms. However, multivariate regression showed that only the “anxiety/depression” dimension of the EQ-5D was significantly associated with depressive symptoms. The indices representing comprehensive health-related QOL and the EQ-VAS and EQ-5D did not show a significant association with depressive symptoms. However, when the QOL as assessed by the “anxiety/depression” dimension was poor, the possibility of having depressive symptoms increased. This result could be because the question used to identify the depressive symptom group and the one used for the EQ-5D (“anxiety/depression”) were similar. This aspect might also serve to explore a possible association between AMD and depression and to increase the statistical power of our study.
Depression and AMD must be handled comprehensively, not separately, as one can affect another. These conditions should be treated as a context within the larger framework of comorbid depression with chronic illness.
Comorbid depression can lead to increased morbidity of AMD as it compounds the disability resulting from vision loss and may even impact a patient's ability to seek out appropriate care. Several studies have shown that depression is linked to increased disease-related morbidity and mortality. This has been particularly well studied with chronic illness; depression has been shown to be clearly associated with a poorer prognosis and more rapid progression of chronic illnesses [
27,
28,
29]. In ischemic heart disease and diabetes, physiologic changes may occur that are affected by depression and can increase morbidity in addition to decreased treatment adherence and mobility and increased substance abuse [
27,
28].
In addition, although these patients are at greater risk than the general population, having a chronic illness may actually make it less likely that providers will screen for depression given the demands of chronic illness management and coordination of care.
Furthermore, treating depression might change the outcomes of AMD. It was previously reported that a mental health intervention for patients with AMD halved the incidence of depression compared to those receiving standard treatment for AMD [
30].
Our study had several limitations. First, AMD was diagnosed using fundus photography alone in the present study; no other ophthalmic examination method, including fluorescein angiography or optical coherence tomography, was carried out because of the nature of this large population-based survey. In the future, we will be able to assess ophthalmic risk factors in more detail and identify their specific effects on depressive symptoms by incorporating data from other ophthalmic examinations into our study. Second, a single question was provided to measure the degree of depressive symptoms instead of established depression questionnaire tools, so the accuracy of our responses could be limited. Further, this one question provided little information regarding the degree of severity of the depression or its relationship with the severity of AMD and level of dependency. However, the question “do you often feel sad or depressed?” showed a sensitivity of 86%, a specificity of 78%, and positive predictive and negative predictive values of 82% in previous studies, so the error is likely small [
31]. Third, the screening for ADL was carried out with just a yes or no answer, and the question appears relatively non-specific. The question does not indicate the degree of dependency on ADL because there is likely variability in what individuals consider “disturbances” in everyday life. Fourth, since this was a cross-sectional study, a causal relationship could not be defined. Fifth, the possible influence of treatment with intravitreal anti-vascular endothelial growth factor injection for AMD and anti-depressants for depression was not considered. These factors could be used as covariates in a future multivariate analysis.
Despite these limitations, this study used representative nationwide, population-based data to determine the prevalence of depressive symptoms in Korean AMD patients. The current study was the first population-based study to evaluate depression and QOL in Korean AMD patients. Because the social, cultural, and racial/ethnic characteristics of Korean AMD patients may be different from those in Western countries, our evaluation of depression and QOL in Korean AMD patients is valuable. Our study indicated that female gender and dependency in ADL were the most important associated factors for depressive symptoms in Korean AMD patients. Ophthalmologists may consider using these findings to determine which patients should undergo further evaluation and treatment for depression, such as neuropsychiatric support.