Background
End-stage renal disease (ESRD) patients undergoing hemodialysis (HD) cope with several problems and difficulties related to the disease and its treatment. As a consequence, depressive feelings are common among them [
1,
2]. We care for HD patients from a low-income region in Brazil. Our patients need all kinds of help to cope with the treatment, like housing for patients living far from the renal unit, provision of food and government financial assistance. So, we are very interested in all aspects of socioeconomic status and their relationship with patient outcomes. Previously, using a validated Brazilian instrument, we were not able to detect differences in prevalence of depression according to economic class [
3].
In light of our daily practice of caring for socially disadvantaged patients, the data on the influence of socioeconomic status on depression led to our interest in finding socioeconomic predictors of depression among HD patients [
4,
5]. Searching for an appropriate instrument, we found the Social Adaptability Index (SAI) [
6]. This instrument is a composite index comprising five kinds of social aspects: education level, marital status, employment status, income and substance abuse. As a sum of these five social items, SAI was able to predict death among HD patients and graft loss among kidney recipients [
7‐
9].
Therefore, we aimed to compare the SAI between depressed and non-depressed HD patients and to test the SAI and its substance abuse item separately as predictors of depression.
Results
The sample characteristics are shown in Table
1. The results concerning each component of SAI are shown in Table
2. Among 145 patients, there were 36 (24.8%) with depression. The comparisons between depressed and non-depressed patients are shown in Table
3. The mean SAI among depressed and non-depressed patients was, respectively, 6.1 ± 1.6 vs. 6.2 ± 1.9 (p=0.901) (Table
3). The percentage of patients experiencing substance abuse among depressed and non-depressed patients was, respectively, 13.9 vs. 13.8% (p=1.000) (Table
3). Gender, age, time on dialysis, total SAI, SAI without the substance abuse item, only the substance abuse score and substance abuse as a categorical variable were not predictors of depression (Table
4).
Table 1
Sample characteristics
Gender
| |
Male | 88 (60.7) |
Female | 57 (39.3) |
Age
| 48.1 ± 15.6 |
Etiology of kidney disease
| |
Glomerulonephritis | 47 (43.1) |
Hypertensive nephrosclerosis | 32 (29.4) |
Diabetic | 12 (11.0) |
Obstructive | 6 (5.5) |
Polycistc | 5 (4.6) |
Lups | 1 (0.9) |
Undetermined | 6 (5.5) |
Time on dialysis (months) | 53.4 ± 51.4 |
Social Adaptation Index
| 6.21 ± 1.8 |
Social Adaptation Index
| 5.97 ± 2.1 |
without the substance abuse item
| |
Substance abuse
| |
Yes | 20 (13.8) |
No | 125 (86.2) |
Depression
| |
Yes | 36 (24.8) |
No | 109 (75.2) |
Laboratory
| |
Hemoglobin | 10.0 ± 1.7 |
Creatinine | 8.7 ± 2.4 |
Albumin | 4.3 ± 0.4 |
Kt/V | 1.8 ± 0.4 |
Table 2
Distribution of sample according to each variable of the Social Adaptability Index
Employment status
| | |
Unemployed | 42 | 29.0 |
Retired | 97 | 66.9 |
Working part time | 4 | 2.7 |
Working full time | 2 | 1.4 |
Education level
| | |
Did not complete high school | 104 | 71.7 |
High school graduate | 26 | 18.0 |
College graduate | 12 | 8.3 |
Postgraduate study | 3 | 2 |
Marital status
| | |
Not married | 39 | 27.0 |
Divorced | 11 | 7.8 |
Married without children | 6 | 4.1 |
Married with children | 89 | 61.1 |
Substance abuse
| | |
Abusing drugs, alcohol and tobacco | 5 | 3.5 |
Abusing substances in 2 of 3 categories | 4 | 2.7 |
Abusing substances in 1 of 3 categories | 11 | 7.6 |
None | 125 | 86.2 |
Income
| | |
≤$20K/year per household | 112 | 77.2 |
$20-50K/year per household | 25 | 17.2 |
≥50K/year per household | 8 | 5.6 |
Table 3
Comparisons between depressed and non-depressed patients
Gender
| | | |
Male | 17 (47.2) | 71 (65.1) | 0.076 |
Female | 19 (52.8) | 38 (34.9) | |
Age
| 47.0 ±14.6 | 48.5 ± 16.0 | 0.627 |
Diabetes
| | | |
Yes | 7 (19.4) | 12 (11.0) | 0.253 |
No | 29 (80.6) | 97 (89.0) | |
Time on dialysis (months) | 54.3 ± 48.3 | 55.1 ± 50.5 | 0.384 |
Social Adaptation Index
| 6.1 ± 1.6 | 6.2 ± 1.9 | 0.901 |
Social Adaptation Index without the substance abuse item
| 5.8 ± 1.9 | 6.0 ± 2.2 | 0.793 |
Substance abuse
| | | |
Yes | 5 (13.9) | 15 (13.8) | 1.000 |
No | 31 (86.1) | 94 (86.2) | |
Depression score
| 30.9 ± 6.4 | 11.2 ± 6.5 | <0.001 |
Laboratory
| | | |
Hemoglobin | 9.6 ± 1.6 | 10.1 ± 1.7 | 0.162 |
Creatinine | 8.2 ± 2.3 | 8.9 ± 2.4 | 0.123 |
Albumin | 4.3 ± 0.4 | 4.4 ± 0.4 | 0.528 |
KtV | 1.8 ± 0.5 | 1.8 ± 0.4 | 0.437 |
Table 4
Logistic regression to detect predictors for depression
Gender
1
| 2.135 | 0.944-4.831 | 0.069 |
Age
| 1.003 | 0.975-1.031 | 0.856 |
Time on dialysis
| 1.001 | 0.994-1.008 | 0.865 |
Social Adaptability Index
| 0.971 | 0.765-1.231 | 0.805 |
Social Adaptability Index without the substance abuse item
| 0.977 | 0.822-1.165 | 0.791 |
Only the substance abuse score
| 0.963 | 0.540-1.724 | 0.898 |
Substance abuse
2
| 1.091 | 0.316-3.764 | 0.891 |
Discussion
Our finding of 24.8% depressed patients is in accordance to the overall literature, which reports depression of 20 to 30% among HD patients [
1]. We found the same prevalence of depression among men and women, in line with another study and also with our previous research [
3,
14]. We think the lack of gender difference is due to the powerful stressors associated with HD, which neutralize the gender differences occurring in the general population [
15]. Social factors, like male perception of loss of the status of main household breadwinner due to illness, and biological factors, like low level of testosterone, a protective hormone against depression, can also be explanations [
16,
17].
We aware of divergences about the optimal cut-off of CES-D to classify depression. We chose a cut-off of ≥ 24, which is within the range for the screening of mild depression, described in the literature as being between 16 and 26 [
18].
We were surprised that even though the patients come from a low-income area, the mean SAI of our sample is close to that found among American patients [
6,
7]. This can be due to the fact that socioeconomic disadvantages are risk factors for ESRD worldwide [
19].
We were disappointed that SAI was not different between depressed and non-depressed patients, and did not predict depression. Our initial hypothesis of SAI being a predictor of depression was based on data about the relationship of socioeconomic aspects with depression [
4,
5]. On the other hand, we have to recognize that the links of low SAI score and bad clinical objective outcomes, like death and graft loss, are more easily explained by health care barriers, lower literacy level (negatively influencing treatment adherence), and less social support to help solve problems [
9]. Our hypothesis was that all these factors could also be associated with depressive feelings. But when dealing with subjective outcomes such as depression, rather than with objective outcomes like mortality and morbidity, plausibility usually does not ensure existence. The main reason for unexpected results concerning subjective outcomes is the modulation of socioeconomic aspects by personality and ways of coping. Among the mechanisms proposed to explain how socioeconomic status influences clinical outcomes, social support is the best studied concerning depression. There is doubtless an overlap between social adaptability and social support, especially regarding aspects like presence of cohabitants, employment status and education level. On the other hand, some emotional aspects of social support, for instance reliance on friends and family or opportunities for emotional expression, can be independent of social aspects comprised by the SAI. Even when social support is evaluated by a specific tool, like the 24-item Social Provisions Scale, depression among ESRD patients depends more on patients’ personality than on the perceived social support [
20]. In this last referenced study, better social support among patients high in the personality trait of “agreeableness” was associated with a decrease in depressive symptoms, whereas social support had little effect on depression change for individuals ranked as low in “agreeableness”.
In clinical practice, there are no widely used and well-validated instruments to assess socioeconomic status able to work as predictor tools in ESRD. So, an instrument like the SAI, validated as a predictor of objective clinical outcomes among HD and transplanted patients, seemed to us a very practical way of using a short instrument also to predict subjective outcomes, such as depression, surely a main outcome among HD patients. Medical monitoring of patients on HD is a hard task and renal units need to work with practical and, if possible, few instruments. Unfortunately, based on our preliminary data, SAI does not predict depression. Granted, our sample is small and from an underdeveloped area, which means there are more young patients and fewer diabetics, with glomerulonephritis being the main cause of ESRD. So, our results may not hold for other more typical samples. In our country, SAI was never tested as a predictor tool for any outcome. We believe our results can stimulate further research using SAI in larger samples with different profiles from ours.
The use of the SAI offered us an opportunity to evaluate an important question in HD patients: substance abuse. In the nephrology area, substance abuse must be highlighted as a risk for ESRD [
21,
22]. Substance abuse as a prior condition to the beginning of dialysis can be one explanation for the relatively high prevalence of substance abuse among patients undergoing HD. We found 13.8% of the patients admitted to being or having been substance abusers, a number very close to the 19% found by others [
23]. Moreover, substance abuse also needs to be studied as an emotion-oriented coping method against stressors and difficulties of ESRD and HD. In our context of a renal unit located in a very poor region, we cannot forget that in the neighborhoods where most of our patients live the use of recreational drugs is very common, as is their sale.
There were limitations of this study, mainly due to the cross-sectional design. Trends of causality related to depression, social adaptability and substance abuse cannot be clarified. Also, no data about duration of substance abuse (previous or after the beginning the dialysis) were collected. A sample from a single renal unit is always a barrier for generalizations, and as previously commented, our sample is quite different from those from developed areas. However, despite these limitations, we emphasize that it is the first time the SAI has been studied in Brazil, which contributes both to stimulate new research in our country using the same instrument and opens the possibility for comparing results across different countries.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
PRS was responsible for the conception, design, analysis and interpretation of data. FPNA edited and made critical revision of the manuscript. Both authors read and approved the final manuscript.