Stroke-free status and depression scores among Saudi dialysis patients

Abstract

Background: To assess the prevalence of stroke symptoms and depression among Saudi dialysis patients and related factors. Methods: This is a cross-sectional multicentre study of Saudi dialysis patients. Demographics, clinical and laboratory’s data were collected. Freedom from stroke symptoms was assessed using the Questionnaire for Verifying Stroke-Free Status and depression using the Geriatric Depression Scale. Results: Five-hundred and forty-nine patients (77.3% response rate); 94.6% were receiving hemodialysis and 5.4% peritoneal dialysis were included in the study. Freedom from stroke was reported in 76.6% of patients, was higher in females (p = 0.07), and was not affected by the presence of diabetes mellitus coronary artery disease or peripheral vascular disease. However, it was significantly lower in hypertensive patients (p = 0.035) and was not affected by age, dialysis duration, Kt/V, albumin or hemoglobin levels. It was, however, more prevalent in the non-depressed patients compared to depressed patients (p = 0.036). Mild and major depression scores were noted in 45.2% and 6.1%, respectively; the depression score being significantly higher in HD than in PD patients (6.3 ± 3.4 vs. 5.0 ± 3.2 p = 0.049) but was not related to sex, employment, vascular access type, age, dialysis duration, Kt/V or serum albumin or hemoglobin levels. Conclusion: A quarter of the patients were not stroke symptom-free and these were more likely the depressed patients, in females, but significantly less in hypertensive patients. Almost half of the patients were depressed with 6.1% having major depression. The depression score was significantly higher in HD than in PD patients.

• Depression
• dialysis
• Saudi Arabia
• stroke

Introduction

The incidence of stroke is many folds higher among patients receiving maintenance dialysis compared to the general population after adjustment for age, sex and comorbidities.1 The incidence of hospitalized ischemic strokes among HD patients is 2.4 times more common and hemorrhagic stroke 5.7 times more common than in sex- and age-adjusted general population.2 Patients on peritoneal dialysis (PD) have a lower risk of hemorrhagic stroke but similar risk for ischemic stroke compared to hemodialysis (HD) patients.2 The independent risk factors for both types of strokes are older age, male sex, diabetes and hypertension.2

Even without prior history of stroke or transient ischemic attacks (TIA), 36.5% of the dialysis patients had one or more stroke symptoms. After adjustment for age, sex, race, education, language, diabetes, and cardiovascular disease, it was found that those with stroke symptoms had lower scores on tests of attention, psychomotor speed, and executive function and higher likelihood of cognitive impairment.3

Even in the absence of a history of stroke or TIA, MRI detected silent infarcts in 48.8% of patients on dialysis compared to only 9.6% of healthy controls.4 Such silent strokes had been reported to be associated with doubling of the risk of future clinically evident strokes or/ and dementia.5

Ethnic variations in stroke prevalence have been reported. After adjustment for age, sex and DM, hemorrhagic strokes we found to be commoner among African Americans, Hispanics than in white Americans (ARR 1.43, 1.78 and 1.43, respectively).6

In this study, we used the Questionnaire for Verifying Stroke-Free Status (QVSFS) to assess for the presence of stroke symptoms or freedom thereof. This tool was previously reported to have high sensitivity (1.0) and specificity (0.86) for stroke, TIA, or both. More relevantly, its negative predictive value (NPV), was found to be exceptionally high (1.0). The positive predictive value (PPV), however, was much lower (0.36)7 This high degree of NPV associated with the QVSFS can be observed even in populations with a high prevalence of prior strokes or TIAs.8

Depression is a common finding among patients on hemodialysis. In a large DOPPS study carried out in the United States five European countries, the prevalence of depression was found to be 20%. The adjusted relative risks of mortality among depressed versus non-depressed was 1.23 for physician-diagnosed depression and even higher for the response to two simple questions: (a) “so down in the dumps” question, and (b) “downhearted and blue” question, ARR 1.48 and 1.35 respectively.9

Our study was designed to assess the prevalence of stroke symptoms and depression among Saudi patients and their relationship to each other and to clinical, demographic covariates as well as performance outcome measures among Saudi hemodialysis patients. We also compared the findings between HD and PD patients.

Patients and methods

This was a cross-sectional multicentre study on dialysis patients from three major dialysis centers in Saudi Arabia from three different cities (Riyadh, Jeddah and Dammam). All the dialysis patients over the age of 21 years who had been on maintenance dialysis for at least 90 d were approached to participate in the study.

All the patients who fulfilled the inclusion criteria of the study were included in the study. Those included represented 77.3% of the total number of patients in the three centers studied.

The whole project took one month to complete (from mid June 2014 to mid-July 2014).

Demographic data (type of dialysis, gender, age, level of education and employment status), presence of comorbidities [diabetes, hypertension, coronary artery disease (CAD) and peripheral vascular disease (PVD)], clinical data (pre-dialysis weight, systolic and diastolic blood pressure) and laboratory data (Kt/V, hemoglobin and serum albumin) were all recorded.

Physical activity was estimated using self-perceived activity score (score 1 to 10; 1 indicating immobility and 10 indicating fully mobility).

To evaluate stroke-free status, we used QVSFS which is a validated 8-item structured interview designed to identify stroke-free individuals. We elected to use the Geriatric Depression Scale (GDS) because of its ability to distinguish symptoms of depression and dementia and its high sensitivity and specificity (92% and 89%) when evaluated against diagnostic criteria. Additionally, it has been widely used in different medical conditions and settings and different ethnicities. Although designed primarily for the elderly, the GDS appears valid in younger patients.10 Moreover, a recent study compared the presence of depression among elderly and non-elderly Japanese HD patients using both visual analogue scale (VAS) and the short GDS scale. No difference was found between the VAS scores for elderly and non-elderly patients and for mood correlated with GDS-15 in HD patients.11

The GDS scale also has high internal consistency and reliability12 and proved to be consistently better in differentiating depressed from non-depressed patients than other depression measures particularly in the elderly.13

Depression among the dialysis patients was assessed using the GDS, which is a 15-item questionnaire. Scores of 0–4 are considered normal, 5–10 as showing mild depression and >10 as marked depression. The GDS has a 92% sensitivity and a 89% specificity when evaluated against diagnostic criteria. The validity and reliability of the tool have been supported through both clinical practice and research.

The translations of the questionnaires from English to Arabic were carried out by Arabic speaking nephrologists and translated back again into English independently by professional translators for consistency and reliability. The translation was then tried initially on 10 patients to check how easy it is to understand.

The local IRB gave its approval for the study the oral consent was obtained from all the participants.

Statistical methods

Data was analyzed using SPSS package (version 21). Descriptive statistics were generated. Significance of differences in the various laboratory, demographic and laboratory variables between those with and without depression and those with or without stoke symptoms as well as those in PD vs. those in HD were calculated using Pearson Chi sq. for non-parametric variables and comparisons of proportions and by 2-tail independent t test for parametric variables. These methods were used rather than the linear or mixed-methods modeling as the data were cross-correctional and not longitudinal in nature. There were no missing data to deal with.

Results

Five-hundred and forty-nine patients (this represents 77.3% of the total number of patients in the three centers studied and 6.8% of all the dialysis patients treated in the Saudi Ministry of health facilities in the entire Kingdom of Saudi Arabia) were included in the study.

Validation of QVSFS in the Saudi dialysis population

QVSFS has not been previously validated in the Saudi population. Previous validation studies from other population used MRI findings as reference parameter. We are unable to do the same validation tool in our patients as the effect from terminal renal failure and MRI in this setting would be contraindicated to avoid the development of gadolinium-induced nephrogenic systemic fibrosis (NSF).14

Instead of this, we used another method of QVSFS validation previously used15 which is based on clinical evaluation instead of an imaging evaluation.

Twelve of our patients with positive QVSFS findings and 12 patients with negative QVSFS findings from the three centers were interviewed and examined by senior experienced physicians who were blinded to the QVSFS results. The objective for them to determine stroke-free status, defined as no history or examination evidence of previous stroke and/or TIA. Out of these patients, 70.1% were males and the mean age was 47.5 ± 11.8 years. The negative predictive value of the QVSFS in Saudi dialysis patients was 1.0 and positive predictive value of 0.8.

Demographic data, comorbidities and performance outcomes

The mean duration on dialysis was 64.5 ± 70.8 months (Table 1). Of all the patients, 94.6% were receiving hemodialysis and 5.4% peritoneal dialysis (this ratio reflects the current national ratio). The mean age was 50.5 ± 14.8 years, 65.5% were male; 30% were illiterate and only 13.5% went to college. Of all the respondents, 62.3 % were unemployed.

Table 1. Demographic data.

Age (years)	50.5 ± 14.8
Sex (males)	65.5%
Duration on dialysis (months)	64.5 ± 70.8
Education: illiterate/primaryintermediate/secondary/college	30%/34.8%/20.9%/13.5%
Mean Activity Scores	1–5 = 13.2%/10 = 86.8%
Employed, student, retired. unemployed	18%, 1.3%, 18. 2%, 62.3%
Dialysis type (1) hemodialysis (2) peritoneal dialysis	(1) 94.6% (2) 5.4%
History of diabetes mellitus	36.4%
History of hypertension	87.1%
History of coronary artery disease	17.6%
History of peripheral vascular disease	15.3%
Vascular access (Catheter/AVF/AVG)	21.1%, 73.9% 4%
Pre-dialysis SBP	143.5 ± 25.1 mmHg
Pre-dialysis DBP	95.1 ± 14.3 mmHg
Pre-dialysis weight	68.4 ± 20 kg
Kt/V	1.7
Albumin	3.5 ± 2 g/L
Hemoglobin	11 ± 6 mg%

The prevalences of co-morbid conditions were as follows: diabetes mellitus (36.4%), hypertension (87.1%), CAD (17.6%) and PVD (15.3%).

The mean pre-dialysis systolic blood pressure, mean pre-dialysis diastolic blood pressure and mean pre-dialysis weight were 143.5 ± 25.1 mmHg, 95.1 mmHg and 68.4 ± 20 kg, respectively.

The mean Kt/V, serum albumin and hemoglobin were 1.7, 3.5 ± 2 g/L and 11 ± 1.6 g%, respectively.

The self-perceived activity score was 1 to 5 in 13.2% and 6 to 10 in 86.8% of the patients.

Patients on PD were more likely to have PVD (p = 0.06), CAD (p = 0.04) or diabetes mellitus (p = 0.000). This could be explained on the basis of the selection for the type of dialysis. However, the prevalence of hypertension was not different between the two groups (p = 0.37).

There was no difference in the employment history between those on PD compared to those on HD (p = 0.42) but there were more patients with higher education in the PD group (p = 0.04). The prevalence of females among PD patients was higher than that of male patients (p = 0.018).

No differences were observed in between the HD and PD groups in terms of age (p = 0.8) or hemoglobin level (p = 0.6). The serum albumin level was significantly lower in the PD group (3.0 ± 0.4) compared to the HD group (3.53 ± 2.1) (p = 0.000).

Stroke-free status

Using the QVSFS tool we found that as many as 27.5% of the patients had at least one criteria for positive stroke history criteria and 72.5% were stroke symptom-free (Table 2).

Table 2. The Questionnaire for Verifying Stroke-Free Status.

		Yes (%)
1	Were you ever told by a physician that you had a stroke?	5.2
2	Were you ever told by a physician that you had a TIA, ministroke, or TIA?	5
3	Is there any evidence of the patient’s file that he/she had a stroke	0.2
4	Have you ever had sudden painless weakness on one side of your body?	11.3
5	Have you ever had sudden numbness or a dead feeling on one side of your body?	10.7
6	Have you ever had sudden painless loss of vision in one or both eyes?	8.9
7	Have you ever suddenly lost the ability to understand what people were saying?	5.7
8	Have you ever suddenly lost the ability to express yourself verbally or in writing?	5.9

The frequency of freedom from stroke was higher in females than males (p = 0.07). The presence or absence of DM, CAD or PVD in patients did not affect the degree of stroke freedom in them. However, it was significantly lower in patient with hypertension (p = 0.035).

Stroke freedom was not affected by age (median 50 years), dialysis duration (median 36 months), Kt/V (median 1.2), albumin level (median 3.3) or hemoglobin (median 10.7).

However, freedom from stroke was significantly more prevalent in the non-depressed patients compared to those who were depressed (p = 0.036) (Table 3).

Table 3. Freeform from stroke and depression.

	Depressed	Non-depressed	p
Possible stroke	74/269(27.5%)	51/250 (20.4%)	0.036
Stroke free	195/269 (72.5%)	199/250 (79.6%)

Depression and activity scores

In our study, we found that 51.9% of the patients suffered from mild depression and 6.1% with major depression which was significantly higher than in HD patients compared to the PD patients (Table 4).

Table 4. The Geriatric Depression Scale (GDS).

	Yes (%)
Are you basically satisfied with your life	8.9
Have you dropped many of your activities and interests?	65.5
Do you feel that your life is empty?	37.6
Do you often get bored?	52.2
Are you in good spirits most of the time?	27
Are you afraid that something bad is going to happen to you?	28
Do you feel happy most of the time?	43.2
Do you often feel helpless?	38.2
Do you prefer to stay at home, rather than going out and doing new things?	54.4
Do you feel you have more problems with memory than most?	24.7
Do you think it is wonderful to be alive now?	11.9
Do you feel worthless the way you are now?	30.6
Do you feel full of energy?	65.4
Do you feel that your situation is hopeless?	29.4
Do you think that most people are better off than you are?	35.9

The depression score was significantly higher in HD patients (6.3 ± 3.4) compared to the PD patients (5.0 ± 3.2; p = 0.049).

We found no significant relationship between depression score and sex, employment, diabetes mellitus or vascular access type.

Depression score was also not affected by age (median 50 years), dialysis duration (median 36 months), SBP (median 143), DBP (median 80), Kt/V (median 1.2), serum albumin level (median 3.3) or hemoglobin (median 10.7).

Of the HD patients, 87.6% had an activity scores ranging between 6 and 10 compared to only 69% in the PD group (p = 0.000).

Discussion

Our finding that the patients on PD were more likely to have PVD, CAD or DM could be explained on the basis of the criteria or the type of dialysis. The lack of difference in the employment history between the PD and HD patients may be related to the fact that there were significantly more females in the PD group in a society in which significantly less number of women go to work.

No differences were observed in between the HD and PD groups in terms of age or hemoglobin level. The serum albumin level was, however, significantly lower in the PD group compared to the HD group.

Depression is reportedly a common finding among patients on hemodialysis occurring in 28.2% to 45.2%.9^,16 It is associated with increased risk in mortality of 35% to 48%.9^,17^,18 Depression in dialysis patients is also significantly associated with non-compliance to treatment, frequency of hospitalization and reduced quality of life.16^,19

In our study, we found that 51.9% of the patients suffered from mild depression and 6.1% with major depression which was significantly higher than in HD patients compared to the PD patients.

The difference in the prevalence of depression between our study and other studies may just reflect the differing tools used to assess depression. We elected to use the GDS scale because of its ability to distinguish symptoms of depression and dementia and its high sensitivity and specificity (92% and 89%) when evaluated against diagnostic criteria. Additionally it has been widely used in different medical conditions and settings and different ethnicities. Although designed primarily for the elderly, the GDS appears valid in younger patients.10 Moreover, a recent study compared the presence of depression among elderly and non-elderly Japanese HD patients using both visual analogue scale (VAS) and the short GDS scale. No difference was found between in VAS scores between elderly and non-elderly patients and that VAS scores for mood correlated with GDS-15 in HD patients.11

One would expect that patients with different cultures, experience, express and perceive depressive symptoms differently necessitating caution in interpreting GDS-15 findings in the Saudi culture. Nevertheless, a cross-sectional survey carried out in a large number of primary healthcare patients in Saudi Arabia using GDS scale20 showed that this tool is appropriate in that population with depressive symptoms seen in 39% of the subjects studied of whom 8.4% were in the severe depressive symptoms score group. Depression correlated with poor education, unemployment (p = 0.001), divorced old age and female sex.20 These findings are similar to ours, especially if one factors in the study by Al-Shammari and Al-Subaie20 which reported that perception of poor health and dependence on others for daily activities were associated with more depressive symptoms (p = 0.001). These factors apply to our chronic dialysis patients who have many comorbid conditions and are often dependent on others.

In previous studies from outside Saudi Arabia, depressive symptoms were found to be more common in women and were associated with unemployment, diabetes, hypoalbuminemia and low level of education.21^,22 By contrast, in our study the depression score was not influenced by gender, employment, diabetes mellitus or vascular access type, age, Kt/V, serum albumin or hemoglobin levels.

The existence of depression was associated with IL-6, but lesser associations were also found with albumin, hemoglobin and creatinine levels.23^,24

Although PD and HD have increased prevalence of depression, the HD patients reported more suicidal thoughts and sleep problems compared to PD patients.7^,25

We found the depression score to be significantly higher in HD patients compared to the PD patients.

The incidence of clinically evident strokes among dialysis patients is known to be very high with incidence of hospitalized ischemic strokes among HD patients reported to be 2.4 times more common and hemorrhagic stroke to be 5.7 times more common than in sex- and age-adjusted general population.2 PD patients had a lower risk of hemorrhagic stroke (but similar risk for ischemic stroke compared to HD patients.2

Ethnicity appears to play a part on the incidence of hemorrhagic stokes. In the USA, after adjustment for age, sex and DM, hemorrhagic stroke was found to be commoner among African Americans, Hispanics and individuals of other races than in white.2

Using the QVSFS tool we found that as many as 23.4% of the patients had at least one criterion for positive stroke history and 76.6% were symptom-free stroke with no difference in the prevalence seen between PD and HD patients or due to age, dialysis duration, Kt/V, albumin or hemoglobin levels.

The lower prevalence of stroke symptom status in our patients compared to that reported from the West may simply reflect the relatively lower age of the men in our patients and the lower prevalence of DM among them but may well be ethnically/genetically related. In a review paper on strokes in Saudi Arabia the authors stated that the incidence and prevalence of strokes were low when compared to those reported from Western countries but this could be largely explained by age differences.26

The frequency of freedom from stroke symptoms, however, was higher in females than males. Similar findings were reported by others.2 The presence or absence of DM, CAD or PVD in patients did not affect the degree of stroke freedom in them. However, it was significantly lower in patient with hypertension which is similar to previous reports.2 We found, however, that freedom from stroke was significantly more prevalent in the non-depressed patients compared to those who were depressed.

Among HD patients without prior stroke or TIA 36.5% were reported to have one or more stroke symptoms.3 Patients with stroke symptoms had lower scores on tests of attention, psychomotor speed, executive function and higher likelihood of cognitive impairment than in a comparable control group.3 MRI detected silent infarcts are seen in 48.8% among patients on (as compared to only 9.6% of health controls).4 Silent strokes has been reported to double the risk for future clinical strokes or dementia.5

Conclusions

The prevalence of diabetes mellitus, coronary artery disease and peripheral vascular disease were lower than those reported from the West probably because the mean age of our cohort was lower. However, the prevalence of hypertension was similar.

As many as 23.4% of the patients indicated at least one criteria for positive stroke criteria and 76.6% were stroke-free. Freedom from stroke was significantly more prevalent in the non-depressed patients compared to those who were depressed.

Over half of the patients were depressed with 6.1% of all the patients having major depression.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.