Background
Diabetes mellitus (DM) is a common health problem with serious medical and economic consequences. Between 2010 and 2030, there will be a 69% increase in numbers of adults with diabetes in developing countries and a 20% increase in developed countries [
1]. The Arab world (North Africa, Middle East, and Gulf area) will have second highest increase in percentage of people with DM in 2030 compared to other parts of the world [
1]. Few studies about prevalence of DM were carried out in Palestine and showed a higher rate of this disease in an urban community than in a rural community [
2‐
4]. However, no reliable data exist for treatment, complications, economic effect, and outcomes of treatment of DM in Palestine [
5].
Depression is another prevalent condition. Approximately 340 million people worldwide suffer from depression at any given time [
6]. A study designed to examine the prevalence of mood disorders in 14 countries found that the 12-month prevalence of mood disorders was lowest in Nigeria and highest in the United States [
7]. It was estimated that depressive disorders were the fourth leading cause of disease burden in women and seventh leading cause in men [
8,
9]. Major depression was found to be the second leading cause of disability-adjusted life years (DALYs) lost in women and the tenth leading cause of DALYs in men [
10]. In Palestine, psychological distress is high, quality of life is very low, and daily life of Palestinians is constantly under threat which make Palestinians more vulnerable to stress and depression [
5].
Relationship between DM and depression has been investigated by many researchers. Prevalence of depression among individuals with DM appears to vary by type of DM, race/ethnicity, and among developed and developing nations [
11,
12]. Therefore, screening for depression among diabetic patients is important in different races and ethnicities. Actually, studies indicated that about 49% of the diabetic patients having severe depression were misrecognized on the primary healthcare clinics [
13,
14]. Unfortunately, Palestinians with chronic diseases are usually less likely to have regular sources of medical care for screening or preventive services [
15].
The aim of this study was to estimate the prevalence of depression among Palestinian type 2 diabetic patients attending primary healthcare settings and to describe socio-demographic and clinical characteristics associated with diabetic patients having depression. Given the scarcity of research about depression among Palestinians in general and among diabetic patients in particular, such study is needed. The knowledge gained from this study will assist healthcare practitioners to better understand depression in diabetes mellitus and design treatments that address the psychological and the metabolic needs of affected individuals to improve overall health outcomes.
Results
During the study period, approximately 1400 patients visited the center and did the HbA1c test per required by the physician. Many of those patients leave the clinic immediately after taking the blood sample which made them un-available for interview. A net total of 301 diabetic patients were available and agreed to do the interview. Seven patients were excluded because of their inability to communicate or understand the questions or have sensory impairment like being deaf. A net of 294 type 2 diabetic patients met the inclusion criteria and were interviewed and their HbA1c test was obtained.
The majority (216; 73.5%) of participants were younger than 65 years old. Age was negatively skewed with a median (Q1 – Q3) age of 60 (52 – 66.25) years. The majority (164; 55.8%) of participants was female. The largest portion (n = 213, 72.5%) of participants reported being either illiterate or had a limited school education while those with college education were minority (n = 40; 13.6%). The marital status as reported by participants was married (n = 243; 82.7%), single/divorced/widowed (n = 51, 17.3%). A small proportion (n = 62, 21.1%) of the participants reported that they have a current job. The majority (n = 223, 75.9%) of participants were non-smokers at the time of the study. Participants reported the number of years since diagnosis with type 2 diabetes which showed a wide distribution ranging from 5 – 35 years, positive skewness with a median (Q1 – Q3) of 10 (5 – 16) years. The Body Mass Index (BMI) of the participants showed positive skewness with a median (Q1 – Q3) of 30.6 (27.2 – 35.1). The BMI for males was 28.8 (26.6 – 33.8) while that for females was 32 (28.3 – 37.4). The difference in BMI between males and females was significant (P < 0.01) with a Z score of – 3.44. More than half (166; 56.5 %) of the participants had a BMI ≥ 30. Both systolic and diastolic blood pressures (SBP, DBP) showed positive skewness. The median (Q1 – Q3) values for SBP and DBP were 132 (120 – 143) and 80 (75 – 88) mmHg respectively.
The number of diabetes medications reported by participants ranged from 0-3 types with a median (Q1 – Q3) of 2 (1 – 2) medications. One hundred and forty seven (50%) reported insulin use either as a monotherapy or in combination with other medications for diabetes management. Two hundred and sixty participants acknowledged having additional illnesses along with diabetes. The number of additional illnesses ranged from 1 – 5 with a median (Q1 – Q3) of 2 (1 - 3). The most common chronic illness reported by the participants was high cholesterol (n = 219, 74.5%), followed by hypertension (n = 178, 60.5%), cardiac problems (n = 84, 28.6%) and renal problems (37, 12.6%). In the reported medical history, none of the participants recalled being informed that they had depression. Additionally, none of the participants reported taking medications for depression.
The results for HbA1C ranged from 5.2 to 13%, a median (Q1 – Q3) of 8.2% (7.2 – 9.1). The median and mean scores of HbA1C exceed the recommendation of < 7.0% established by the ADA [
29]. The results were un-equally distributed with 3 extreme scores representing HbA1C between 12 and 13%. The HbA1C results were further divided into two categories: controlled (< 7%) and uncontrolled (≥ 7.0%). The percentage of participants who had a controlled glucose was 17.7% (n = 52). However, 82.3% (n = 242) were in the uncontrolled level which places individuals at a greater risk for developing diabetic complication.
The depression symptoms score distribution was positively skewed with the majority of participants [n = 174, 59.2%] scored less than 16 while 120 (40.2%) patients scored ≥ 16 in the depression scale. Medication adherence scores were negatively skewed. The median value for medication adherence scores was 5.8 (4.8 – 6.8). The majority (n = 167, 56.8%) of the participants had low adherence scores (MMAS-8 < 6). Table
1 shows univariate analysis of demographic and clinical factors with depression scores. The analysis showed that female diabetic patients, low level of education, having no current job, having higher number of additional illness, low medication adherence score and having high BMI were significantly associated with depression score of ≥ 16. Multivariate analysis (Table
2) showed that: 1) diabetic patients with college education were less likely to have depression [O.R = 0.24; (0.09 – 0.66)] than those with lesser level of education, 2) diabetic patients with no current job were more likely to have depression [O.R = 2.78 (1.23 – 6.27)] than those who are currently working, 3) diabetic patients with multiple additional illnesses were more likely to have depression [O.R = 1.81 (1.05 – 3.11)] than those with ≤ 1 additional illness, and finally 4) diabetic patients with high medication adherence score were less likely to have depression [O.R = 0.31 (0.18 – 0.53)] than those with low (< 6) medication adherence score.
Table 1
Univariate analysis of factors associated with depression
Age (years)
| | | | | |
≤ 65 | 216 (73.5%) | 85 (39.4%) | 131 (60.6%) | 0.8 (0.5 – 1.3) | 0.40 |
> 65 | 78 (26.5%) | 35 (44.9%) | 43 (55.1%) | Reference |
Gender
| | | | | |
Male | 130 (44.2%) | 43 (33.1%) | 87 (66.9%) | Reference | 0.02 |
Female | 164 (55.8%) | 77 (47.0%) | 87 (53.0%) | 1.8 (1.1 – 2.9) | |
Education
| | | | | |
Illiterate | 87 (29.6%) | 47 (39.2%) | 40 (23.0%) | Reference | |
Elementary | 126 (42.9%) | 53 (44.2%) | 73 (42.0%) | 0.6 (0.4 – 1.1) | 0.08 |
High school | 41 (13.9%) | 13 (10.80%) | 28 (16.1%) | 0.4 (0.2 – 0.9) | 0.02 |
College | 40 (13.6%) | 7 (5.8%) | 33 (19.0%) | 0.2 (0.1 – 0.5) | < 0.01 |
Marital status
| | | | | |
Married | 243 (82.7%) | 101 (41.6%) | 142 (58.4%) | 1.2 (0.6 – 2.2) | 0.58 |
Others | 51 (17.3%) | 19 (37.3%) | 32 (62.7%) | Reference |
Currently smoking
| | | | | |
Yes | 71 (24.1%) | 32 (45.1%) | 39 (54.9%) | 1.3 (0.7 – 2.2) | 0.40 |
No | 166 (56.5%) | 79 (47.6%) | 84 (52.4%) | Reference |
Working
| | | | | |
Yes | 62 (21.1%) | 12 (19.4%) | 50 (80.6%) | Reference | < 0.01 |
No | 232 (78.9%) | 108 (46.6%) | 124 (53.4%) | 3.6 (1.8 – 7.7) |
Duration of illness
| 10 (5 – 16) | 12 (6 – 16.75) | 10 (5 – 15) | 1.0 (0.9 – 1.05) | 0.32 |
Number of additional illness
| | | | | |
≤ 1 | 114 (38.8%) | 36 (31.6%) | 78 (68.4%) | Reference | 0.01 |
≥ 2 | 180 (61.2%) | 84 (46.9%) | 96 (53.3%) | 1.9 (1.2 – 3.1) |
Number of anti-diabetic medication
| | | | | |
≤ 1 | 112 (38.1%) | 44 (39.3%) | 68 (60.7%) | Reference | 0.68 |
≥ 2 | 182 (61.9%) | 76 (41.8%) | 106 (58.2%) | 1.1 (0.7 – 1.8) |
Insulin use
| | | | | |
Yes | 147 (50%) | 64 (43.5%) | 83 (56.5%) | 1.25 (0.8 – 2.0) | 0.34 |
No | 147 (50%) | 56 (38.1%) | 91 (61.9%) | Reference |
HbA1C
| | | | | |
< 7 | 52 (17.7%) | 23 (44.2%) | 29 (55.8%) | Reference | 0.58 |
≥ 7 | 242 (82.3%) | 97 (40.1%) | 145 (59.9%) | 0.8 (0.5 – 1.5) |
Medication adherence score
| | | | | |
< 6 | 167 (56.8%) | 86 (51.5%) | 81 (48.5%) | Reference | < 0.01 |
≥ 6 | 127 (43.2%) | 34 (26.8%) | 91 (73.2%) | 0.3 (0.2 – 0.6) | |
BMI
| | | | |
< 25 | 30 (10.2%) | 7 (23.3%) | 23 (76.7%) | Reference | |
25 – 29 | 98 (33.3%) | 34 (34.7%) | 64 (65.3%) | 1.8 (0.7 – 4.5) | 0.25 |
≥ 30 | 166 (56.5%) | 79 (47.6%) | 84 (52.4%) | 3.0 (1.2 – 7.3) | 0.02 |
Table 2
Multivariate analysis of factors associated with depression
Gender
| Female | | | | | Reference |
Male | -0.04 | 0.31 | 0.02 | 0.898 | 0.96 (0.52-1.77) |
Education
| Illiterate | | Reference |
Elementary | -0.33 | 0.31 | 1.15 | 0.284 | 0.72 (0.39- 0.32) |
High school | -0.85 | 0.45 | 3.61 | 0.057 | 0.43 (0.18- 1.03) |
College | -1.42 | 0.52 | 7.57 | 0.006 | 0.24 (0.09-0.66) |
Occupation
| Yes | | | | | Reference |
No | 1.02 | 0.42 | 6.05 | 0.014 | 2.78 (1.23-6.27) |
Additional illness
| ≤ 1 | | | | | Reference |
≥ 2 | 0.59 | 0.28 | 4.55 | 0.033 | 1.81 (1.05-3.11) |
BMI
| < 25 | | | | | Reference |
25 - < 30 | 0.50 | 0.52 | 0.92 | 0.336 | 1.65 (0.60-4.58) |
≥ 30 | 0.86 | 0.50 | 3.00 | 0.083 | 2.36 (0.90-6.23) |
Medication adherence score
| < 6 | | | | | Reference |
≥ 6 | -1.17 | 0.27 | 14.17 | 0.001 | 0.31 (0.18- 0.53) |
Discussion
This study investigated the prevalence of depression among adult Palestinian type II diabetic patients and identified demographic and disease-related risk factors for depression. Our study showed that 40% of the screened patients are potential cases of depression. It should be emphasized that high BDI II scores presented in this study are not diagnostic of depression. Rather, high scores are indicative of the presence of depressive symptoms and further clinical consultation to establish definitive diagnosis of depression is needed. In our study, most patients who are potential cases of depression were females, had multiple additional illnesses, currently jobless, had low educational level, had low medication adherence, and had an abnormally high BMI. However, multivariate analysis showed that the only significant predictors of depression were low education, having no current job, having multiple additional illnesses, and having low medication adherence. Many research groups have investigated the relationship between diabetes mellitus and depression. Some research groups indicated that depression is highly prevalent among diabetics and the risk of depression might be increased in the presence of other co-morbid conditions [
11,
32‐
38]. Compared to other published studies, our results reported higher prevalence rate of depression among diabetics. A meta-analysis of 42 published studies found that the prevalence of major depression among diabetic patients was 11% and the prevalence of clinically serious depression was 31% [
39]. A study in Jordan found that the prevalence rate of undiagnosed depression among Jordanian diabetic patients was 19.7% [
17]. A meta-analysis study held in the United States found that the prevalence rate of depression among adult diabetic patients ranging from 3.8% to 27.3% [
39]. Other studies reported a prevalence rates of 5.4% [
40], 8% [
41], 32.4% [
42] and 41.3% [
43]. One possible reason for the differences in the prevalence of depression among diabetic patients reported by different studies is the use of different scales used to screen for depressive symptoms. Some studies used the PHQ-8, others used Ham-D or BDI II scales.
The increased vulnerability to depression in individuals with type 2 diabetes is not yet clearly understood. However, depression involves physiological changes of the neuroendocrine system. The underlying cause of depression is thought to be related to changes in the neurotransmitters in the brain such as serotonin (5-HT), dopamine (DA), and norepinephrine (NE) which are monoamine neurotransmitters which affect mood and behavior. It is believed that during psychological stress counter regulatory hormones such as catecholamine a neurotransmitter, glucocorticoids, growth hormones, and glucagons are activated [
44]. The activation of the counter regulatory hormones interferes in the action of insulin which is not able to lower glucose but instead elevates blood glucose. The increase in glucose level creates a greater challenge in maintaining metabolic control. Poor glycemic control and functional impairment due to increasing diabetes complications may cause or worsen depression and lessen the response to antidepressant treatment [
45]. In spite of the known devastating effect of depression on diabetes, it was found that only 31% of patients with diabetes and depression received adequate antidepressant treatment and only 6% received 4-5 sessions of psychotherapy in a 12 month period [
46]. Studies of the economics of treatment of depression among diabetic patients have yielded positive results. The health care expenses sustained by individuals with diabetes and depression are higher than those with diabetes alone [
47,
48].
It may be argued that the high prevalence rate of depression found in our study is due to many patients having uncontrolled diabetes (82%). However, we did not find significant relationship between glycemic control and potential for depression. Studies about the relationship between depression and poor glycemic control gave mixed results. A study had shown a negative relationship between depression and poor glycemic control and diabetes complications. Worse glycemic control was observed in depressed adults with diabetes [
49]. A meta-analysis study [
50] found that patients with type 1 and type 2 diabetes and depression persistently had higher HbA1c levels [
50]. Wagner et al. also found higher HbA1c and more diabetes complications in African Americans with higher depressive symptoms after controlling for confounders between depression and HbA1c levels [
51]. In a prospective representative study of patients with type 2 diabetes, depression predicted problems with medication adherence, and unsatisfactory glycemic control [
52]. Diabetic complications and mortality were also found to be greater among depressed patients [
14,
53,
54]. The microvascular and macrovascular complications of diabetes are augmented by the presence of depression in diabetes thus contributing to the increased mortality rate in this population [
55]. In contrast, other investigators found no relationship between depression and diabetic complications [
42,
56‐
58]. Cross-sectional studies found a significant positive correlation between depressive symptoms and HbA1c in patients with Type 1 diabetes but no significant correlation in patients with Type 2 diabetes [
59‐
61], indicating that depression affects glycemic control in patients with Type 1 but not Type 2 diabetes. In support of this, a study found that patients with Type 1 but not Type 2 diabetes who had depression showed significantly worse glycemic control than their counterparts who had no depression [
62,
63]. A study showed that changes in depressive symptoms were not associated with changes in HbA1c or fasting glucose levels over a 1-year period in either patients with Type 1 or Type 2 diabetes [
63].
Our study showed that approximately two thirds of those who scored ≥ 16 points were females. The higher prevalence of depression among females is in consistence with other studies [
17,
35,
55,
64,
65]. In a meta-analysis, Anderson et al found that diabetes doubles the risk of depression and it is especially more among females [
39]. However, not all studies reported this gender differences in depression [
58]. Our results indicated that about 90% of the total sample was either overweight or obese (had a BMI > 25). The BDI score was significantly higher among obese patients. A study found that depression was more common among diabetic women especially if they were overweight and that body weight was a predictor of depression more than diabetes itself [
66]. Similarly, recent studies have found that higher BMI was a predictor of depression in type 2diabetes [
64,
67]. High educational level decreases the odds of being classified as depressed patients which is similar to findings published in other studies [
13,
17]. Medication adherence showed a significant relationship with depression. Adherent patients are associated with lower odds of being depressed. It has been reported by other researchers that depressed diabetic patients do not pay much effort on daily management activities [
68,
69] and this result is consistent with many studies that reported that depressed diabetic patients are likely to have physical limitations and a poor quality of life [
70‐
73], bearing in mind that self-care behaviors in diabetes include adherence to dietary restrictions and medications, adequate physical exercise and blood glucose monitoring [
68,
69]. Management guidelines for diabetes mellitus emphasize the importance of medication adherence, physical activity, diet and self-monitoring of blood glucose [
74]. Gonzalez et al. proposed that the presence of depressive symptoms are good predictors of poor adherence to self-care particularly in adherence to medications and diet and exercise regimens [
75]. A systematic review of treatment adherence among individuals with diabetes and depression indicated that there was a significant relationship between depression and treatment non-adherence [
75]. Finally, our study indicated that having no work is significantly associated with depression score ≥ 16 points. This is expected since having no work is by itself a depressing factor.
In contrast to some other studies, our study showed no significant association between depression score and age among diabetic patients [
76]. Furthermore, our study showed no significant association between depression and the use of insulin which is contrary to other published studies [
11]. A study reported that delayed initiation of insulin in type 2 makes a significant number of diabetic patients vulnerable to diabetic complications and its adverse outcomes, including depression [
77]. Our study found that the duration of DM was not significantly associated with depression which is consistent with the findings reported by other studies [
78,
79]. However, because the incidence of diabetic complication increases with increased illness duration, one could expect greater depression risk in those who have been ill for a longer time. Some research groups reported a significantly higher rate of depression in individual with long standing history of diabetes than in the newly diagnosed diabetic [
80].
Our study is the first to be conducted in Palestine to determine the prevalence of undiagnosed depression among diabetic patients. However, our study has few limitations. First, this study is cross-sectional where causal relationship between diabetes and depression cannot be established. Variables identified as significantly associated with depression may precede depression, but in some cases, these variables could also occur as a result of depression; thus, high scores of BDI II among diabetic patients must be interpreted with caution. Further studies with longitudinal prospective design and the presence of age and gender matched group is needed to shed more light on the potential relationship between depression and diabetes mellitus. It is important to state that depression screening measures provide an estimation of the severity of depressive symptoms and assess the severity within a specific period of time but they do not diagnose depression. However, those who score high in depression measure scales need to interviewed and assessed for a confirmation of depression. The second limitation of this study was its partial reliance on self-report for its measures, including depression and medication adherence; therefore, a clinical interview to assess depression maybe a superior measure because of its higher level of specificity. Accordingly, a longitudinal clinical study of a community-based sample using clinical symptoms for diagnosis of depression is needed to assess the relationship between various variables including diabetes control. Third limitation of this study was HbA1c measurement. The presence of many clinically silent hemoglobin variants might cause deviation in the HbA1c results leading to falsely high or low values [
81,
82]. Therefore, the accuracy of the results and validity of the HbA1c interpretation cannot be assured. Finally, gender differences in risk of depression among diabetic patients need further investigation using large sample size is needed since univariate analysis showed such gender differences with higher risk among females. Community-based study among diabetic women will be of great value given that women are more prone to depression than men.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors were involved in drafting the article and all authors approved the final version to be submitted for publication. All authors have added an intellectual significant value to the manuscript. H.A was involved in subject recruitment and interview, data collection, data coding and entry, literature review, and manuscript editing. This was done in partial fulfillment of a master degree in Community Mental Health Nursing at An-Najah National University. S.A and S.Z were involved in concept, study design, manuscript writing and editing, and academic co-supervision for H.A according to An-Najah university regulations. W.S was involved in the study conception and design, literature review, manuscript writing, manuscript submission, manuscript revision, head of the research group, and academic supervision for H.A according to An-Najah university regulations. This manuscript is part of a project for master degree in the graduate program of Community Mental Health Nursing. The project was initially and originally conceptualized and designed by the Clinical Pharmacology/ Toxicology Research Group at An-Najah National University (W.S, S.Z, and S.A). The project was then assigned to H.A as a thesis project and was academically supervised by W.S and S.A in adherence to An-Najah University regulations with regard to academic supervision for graduate students.