According to the literature, 7-15% of the population suffers from the chronic kidney disease (CKD), characterized by decreased glomerular filtration rate (GFR), proteinuria, and structural kidney damage [1, 2]. The main etiology and pathogenesis factors of CKD are diabetes mellitus (DM), atherosclerosis, arterial hypertension (AH), toxic and immunocomplex damage [3]. According to various estimates, proteinuria, hypertension, impaired glucose tolerance and genetic factors contribute to the progression of CKD to end-stage 5 (CKD C5) [3] in 1—2% of patients with CKD or 0.02% of the total population [4, 5].
The presence of CKD C5 is associated with increased mortality and morbidity compared to the general population [6, 7]. These risk factors of CKD C5 explain the widespread prevalence of cardiovascular diseases, diabetes mellitus and other metabolic disorders in patients with CKD C5 [6]. They are characterized by high incidence of complications and they can lead to damage to the nervous system. Uremia also negatively affects the nervous system like other progressive metabolic disorders. In general, damage to the nervous system in CKD manifests itself in cognitive impairment, peripheral neuropathy with damage to autonomic fibers, anxiety and depressive disorders [8, 9]. Moreover, in CKD C5 the renal replacement therapy can exacerbate the cardiovascular disease and directly affect the nervous system. In particular, chronic renal hemodialysis that is the most common method of renal replacement therapy in CKD C5 (CKD C5d) [10] is associated with dialysis dementia, cerebellar ataxia, progression of cerebrovascular disease, Wernicke's encephalopathy, intracranial hypertension and neuropathy [11].
Sleep and wakefulness disorders are observed in 80–95% of patients with CKD C5d [12, 13] where 35% of patients have a combination of at least two types of sleep disorders [14]. In the structure of sleep and wakefulness disorders in patients with end-stage renal failure the poor sleep quality (53—74%), insomnia (50-75%), restless legs syndrome (RLS; 13—55%), periodic lower limb movement syndrome, breathing sleep disorders (BSD; 24—49%), excessive daytime sleepiness (EDS; 12—21%) and fatigue often accompanying sleep disturbances (83%) [12—24]. In addition, sleep and wakefulness disorders in patients with CKD C5d can be caused by metabolic and endocrine disorders. In patients with CKD C5d the association between the sleep quality and the level of hemoglobin [21], BSD with creatinine blood level, thyroid-stimulating and parathyroid hormones [13], insomnia with anemia [25-28], C-reactive protein and calcium levels [14, 29], RLS with parathyroid hormone levels, calcium and phosphorus has been described [26, 30, 31]. The habit of sleeping during hemodialysis at daytime may be a risk factor for the circadian disorders development [18] while the symptoms of depression may contribute to the deterioration of overall sleep quality in patients with CKD C5d [28].
CKD C5d is characterized by a high comorbidity with mineral and bone disorders caused by minerals metabolism changes, parathyroid hormone and vitamin D due to kidney damage and renal replacement therapy [32]. Mineral and bone disorders in CKD C5d include uremic osteodystrophy, osteo-fibrous osteitis and amyloid bone disease [33, 34]. They can contribute to the common clinical symptoms in patients with CKD C5d like pathological bone fractures (in 5.6% of patients with CKD C5d compared with 0.07-0.22% of the general population), arthralgia (in 37% of patients with CKD C5d) and erosive arthritis (up to 20% of patients with CKD C5d) [35–38]. The CKD is aggravating factor in the bone and joint disease leading to more severe pain and weak response to medical and surgical therapy [39]. In turn, pain, anxiety and depression in musculoskeletal diseases can negatively affect sleep and general well-being [40, 41]. However, the literature contains insufficient data on sleep disorders in patients with CKD C5d and bone diseases.
Objective
Comparative assessment of subjective sleep and wakefulness disorders in patients with osteoarthritis associated with CKD C5d and patients with osteoarthritis without CKD, as well as association of subjective sleep characteristics with anxiety and depression, pain, general health and laboratory parameters in these cohorts.
Materials and methods
This pilot cross-sectional case-control study included two groups of patients with stage III hip joint osteoarthritis treated preparatory to total arthroplasty at First St. Petersburg State Medical University named after acad. I.P. Pavlov. Inclusion criteria: patients aged 18-85 who signed written consent to participate in the study. Exclusion criteria: decompensated mental and somatic disorders that make impossible to complete the questionnaires of research protocol.
The first group included patients with C5d CKD receiving hemodialysis for> 6 months while the second group consisted of patients without CKD (GFR> 60 ml/ min/ m2) individually matched by gender and age without other signs of kidney damage according to laboratory data research [42]).
General characteristics of the population
The study involved 38 patients (22 women and 16 men; mean age 53.5 (46.0—63.9) years) with stage III hip joint osteoarthritis. There were two groups depending on the presence of CKD C5d: 19 — with CKD C5d (11 women and 8 men; average age 53.3 (46.3—61.0) years) and 19 — without CKD C5d (11 women and 8 men; average age 52.4 (46.1—64, 4 years).
The average time of renal replacement therapy in CKD C5d was 8.0 (4.5—15.5) years. The etiological factors of CKD in this population included morphologically unverified nephritic syndrome — 10 (52.6%), diabetic nephropathy — 2 (10.5%), nephropathy in rheumatoid arthritis — 2 (10.5%), membranous-proliferative glomerulonephritis — 1 (5.3%), secondary amyloidosis — 1 (5.3%), urolithiasis resulted in nephrosclerosis — 1 (5.3%), secondary pyelonephritis resulted in nephrosclerosis — 1 (5.3%), hypertensive nephrosclerosis — 1 (5.3%).
The structure of comorbidities in the surveyed cohort is shown in Fig. 1. Patients with CKD C5d compared with control group had a lower incidence of obesity — 21.1 and 57.9%, respectively (p = 0.046) and a lower body mass index (BMI) — 26.6 (21.3 —29.2 kg/m2 and 31.6 (25.7-36.1) kg/m2, respectively (p = 0.016). Only patients with CKD C5d had viral hepatitis B and/or C — 26.3% (p = 0.046) in contrast to patients without CKD. There were no significant differences in the incidence of hypertension (HT), ischemic heart disease (IHD), heart failure (HF), cardiovascular events (CVE) in anamnesis, DM 2 type, chronic obstructive pulmonary disease (COPD), chronic gastritis and smoking.
Fig. 1. Comorbidities in patients with EDRD-HD (n=19) and without CKD (n=19).
Digits above the bars represents the number of patients. Only significant differences are shown (p<0.05, exact Fisher-test).
Clinical examination data
All patients, as part of a routine clinical preoperative examination, consulted a traumatologist and therapist, performed a general clinical examination including anthropometry (height, body weight, calculation, BMI), clinical and biochemical blood tests (erythrocytes, hemoglobin, platelets, leukocytes, erythrocyte sedimentation rate, total protein, C-reactive protein, fibrinogen, amylase, calcium, sodium, potassium, glucose, total cholesterol, urea, creatinine, total bilirubin, alanine aminotransferase, aspartate aminotransferase). Additionally, vitamin D, parathyroid hormone and phosphorus examined only in patients with CKD C5d. Moreover, based on current medical records the comorbidities (AH, IHD, CHF, acute cerebrovascular accident or myocardial infarction), type 2 diabetes, COPD, chronic gastritis, obesity, viral hepatitis B and/or C, smoking has been registered.
Subjective sleep characteristics were assessed using validated questionnaires. The Pittsburgh Sleep Quality Index (PSQI) questionnaire was used to assess sleep quality (poor sleep quality was determined with a score of> 5 [43]). The Epworth Sleepiness Scale (ESS) applied for EDS (EDS was determined at score > 10 [44]). Fatigue severity scale (fatigue was determined at score ≥4 [45]), an insomnia severity scale (insomnia was determined at score ≥4 [46]) also used; Diagnostic criteria of the International Restless Leg Syndrome Study Group Diagnostic Criteria 2012 [47] for RLS and the Berlin questionnaire for assessing the risk of obstructive sleep apnea (OSA) with high and low risk [48].
It was also assessed the subjective health perception for the previous month using the visual analogue scale of general health, VAS (horizontal scale 0—100 points) and assessed the intensity of pain associated with osteoarthritis in the previous week using the pain VAS (horizontal scale 0—10 points) [49, 50].
In addition, the anxiety-depressive disorders was assessed according to Hospital Anxiety and Depression Scale (HADS). Separately it was done by its subscales: anxiety was assessed according to HADS (normal value less than 8 points) and depression according to HADS and Beck Depression Inventory, BDI (depression was determined by 8 or more points on the HADS Depression Scale or 14 or more points on BDI) [51].
Statistics methods
Statistical processing was performed using R Statistical Software 4.0.1 (https://cran.r-project.org/) and additional packages ggplot2, ggsignif, lme4, jtools, tableone, and psych. Missing values were not imputed.
A logarithmic transformation of some variables was performed (the number of points on the PSQI scale, on the HADS scale, on the BDI) after analyzing the type of distribution. The values obtained from the logarithmic transformation were used as dependent variables in the corresponding regression analysis.
Differences between patient groups were studied using nonparametric Wilcoxon tests for continuous variables and Fisher's exact test for categorical variables.
The association of CKD C5d as an independent variable with studied parameters as dependent variables (sleep characteristics, health perception, anxiety and depression, pain and laboratory tests) taking into account age and gender were studied. It was used the multiple linear regression for continuous variables (the rest of the studied variables) and logistic regression (presence of RLS and high risk OSA) for categorical variables. Regression analysis with simultaneous use of RLS and patients’ group depending on CKD presence as independent variables was not performed due to only 3 patients with RLS in the group without CKD (see below).
The associations between the studied parameters (sleep characteristics, perception of health, anxiety and depression, pain and laboratory parameters) were studied using linear and logistic multiple regression. Data are presented as odds ratio (OR) and 95% confidence interval (CI). The method of simple and partial correlations and the nonparametric Wilcoxon test was also used.
Categorical variables are presented as absolute and relative (%) values. Continuous variables were presented as median and interquartile range. The differences were regarded as significant when p<0.05.
Results
Comparison of subjective sleep characteristics, health perception, anxiety and depression, pain and laboratory parameters in patients with CKD C5d and control group
The subjective characteristics of sleep, general health, anxiety and depression, pain and laboratory tests in patients with CKD C5d in comparison control group have been studied. The absolute and relative numerical values and differences significance between the groups are shown in the Table.
Subjective sleep characteristics, perception of general health, anxiety and depression, pain and blood markers in patients with chronic kidney disease 5 and control group. The continuous data is presented as mean and interquartile range
|
Characteristic |
CKD C5d (n=19) |
Control group (n=19) |
p |
|
PSQI, score |
12.00 (8.50—14.50) |
8.00 (6.00—11.00) |
0,046* |
|
Poor sleep quality, % |
19 (100.0) |
19 (100.0) |
|
|
ESS, score |
16.00 (13.00—17.00) |
13.00 (12.00—15.50) |
0,154 |
|
EDS, % |
17 (89.5) |
16 (84.2) |
1 |
|
RLS, % |
11 (58.9) |
3 (16.8) |
0,017* |
|
Fatigue severity scale, score |
4.33 (3.83—5.78) |
3.78 (3.11—4.61) |
0,179 |
|
Fatigue, % |
13 (68.4) |
9 (47.4) |
0,324 |
|
OSA high risk, % |
3 (15.8) |
9 (47.4) |
0,078 |
|
Insomnia severity scale, score |
5.00 (1.00—9.00) |
5.00 (2.00—9.50) |
0,769 |
|
Insomnia, % |
9 (47.4) |
6 (31.6) |
0,507 |
|
VAS, General health, score |
50.00 (40.00—75.00) |
80.00 (70.00—80.00) |
0,014* |
|
VAS, Pain, score |
6.00 (4.00—8.00) |
6.00 (4.00—8.00) |
0,599 |
|
Anxiety scale, HADS, score |
0.00 (0.00—3.00) |
3.00 (2.00—5.50) |
0,025* |
|
Anxiety by HADS |
1 (5.3) |
2 (10.5) |
1 |
|
Depression scale, HADS, score |
2.00 (2.00—5.00) |
4.00 (2.50—6.00) |
0,100 |
|
Depression by HADS |
2 (10.5) |
5 (26.3) |
0,403 |
|
BDI, score |
0.00 (0.00—6.50) |
7.00 (0.50—15.50) |
0,052 |
|
Depression by BDI |
3 (15.8) |
7 (36.8) |
0,269 |
|
RBC, 1012/l |
4.00 (3.70—4.25) |
4.60 (4.50—5.05) |
<0,001* |
|
Hemoglobin, g/l |
121.00 (106.00—127.00) |
140.00 (129.00—151.00) |
<0,001* |
|
Platelets, 109/l |
224.00 (188.00—252.00) |
285.00 (228.50—316.00) |
0,042* |
|
WBC, 109/l |
6.00 (5.00—8.00) |
6.91 (5.38—8.10) |
0,321 |
|
Immature neutrophils, 109/l |
0.01 (0.01—0.04) |
0.02 (0.01—0.03) |
0,733 |
|
Neutrophils, 109/l |
3.61 (2.48—4.93) |
3.97 (2.98—4.45) |
0,683 |
|
Lymphocytes, 109/l |
1.50 (1.35—1.80) |
2.10 (1.75—2.90) |
0,003* |
|
Monocytes, 109/l |
0.61 (0.41—0.76) |
0.54 (0.49—0.67) |
0,815 |
|
Basophils, 109/l |
0.03 (0.02—0.06) |
0.04 (0.02—0.05) |
0,846 |
|
Eosinophils, 109/l |
0.17 (0.11—0.32) |
0.23 (0.18—0.28) |
0,589 |
|
Immature granulocytes, % |
0.25 (0.20—0.50) |
0.20 (0.20—0.30) |
0,444 |
|
Neutrophils, % |
59.50 (51.60—62.75) |
52.90 (50.20—59.60) |
0,261 |
|
Lymphocytes, % |
27.80 (20.30—34.70) |
32.80 (28.75—36.95) |
0,093 |
|
Monocytes, % |
8.60 (7.55—12.50) |
8.10 (7.30—9.85) |
0,214 |
|
Basophils, % |
0.60 (0.35—0.85) |
0.40 (0.35—0.75) |
0,556 |
|
Eosinophils, % |
3.10 (2.25—3.95) |
3.00 (2.10—4.65) |
0,930 |
|
Erythrocytes sedimentation rate, mm |
22.00 (13.50—35.00) |
8.00 (5.00—24.00) |
0,035* |
|
Total protein, g/l |
68.50 (65.00—74.75) |
68.00 (67.00—73.00) |
0,542 |
|
C-reactive protein, mg/l |
3.30 (2.75—24.70) |
6.00 (4.50—10.67) |
0,311 |
|
Fibrinogen, g/l |
3.60 (3.20—4.45) |
3.25 (2.58—3.77) |
0,075 |
|
Amylase, UI/l |
81.00 (46.00—119.00) |
49.50 (40.00—61.50) |
0,031* |
|
Calcium, mmole/l |
2.19 (2.13—2.40) |
2.38 (2.28—2.42) |
0,055 |
|
Natrium, mmole/l |
139.00 (137.15—140.40) |
141.00 (140.35—142.00) |
0,003* |
|
Potassium, mmole/l |
5.50 (5.20—6.00) |
4.30 (4.00—4.70) |
<0,001 |
|
Glucose, mmole/l |
4.50 (4.05—4.85) |
5.20 (4.91—6.08) |
0,004* |
|
Total cholesterol, mmole/l |
4.81 (4.32—5.60) |
4.80 (3.80—5.80) |
0,950 |
|
Urea, mmole/l |
17.80 (14.50—19.25) |
4.60 (4.35—5.75) |
<0,001* |
|
Creatinine, mmole/l |
0.77 (0.59—0.84) |
0.07 (0.06—0.08) |
<0,001* |
|
Total bilirubin, μmol /l |
9.50 (7.40—10.30) |
12.50 (8.30—18.35) |
0,031* |
|
Alanine aminotransferase, UI/l |
16.00 (11.00—22.85) |
16.00 (13.00—26.50) |
0,569 |
|
Aspartate aminotransferase, UI/l |
17.00 (13.50—24.00) |
20.00 (17.00—27.50) |
0,578 |
Note. * — p<0.05. VAS — visual analogue scale, OSA — obstructive sleep apnea, HADS — Hospital anxiety and depression scale, BDI — Beck Depression Inventory, RBC — red blood cells, WBC — white blood cells, UI — international units. The categorical data presented as count and percentage to total
All studied patients had a poor sleep quality; however, patients with CKD C5d had a worse sleep quality according to PSQI compared to control group. EDS prevailed in the structure of subjective sleep disorders that was equally common in CKD C5d patients and control group. A higher incidence of RLS was found in CKD C5d patients compared with control group. Patients with CKD C5d had a lower risk of OSA compared to control group but the differences were not significant. In addition, there were no significant differences between CKD C5d patients and control group in the severity of insomnia and fatigue.
Patients with CKD C5d expressed worse subjective health perception compared to control group.
There were no significant differences in pain intensity associated with osteoarthritis in CKD C5d patients compared to control group.
CKD C5d patients had lower level of anxiety compared to control group. Although CKD C5d patients had also lower level of depression on BDI compared to control group, however these differences were not significant.
Laboratory tests differences were found in patients with osteoarthritis and CKD C5d compared to patients with osteoarthritis without CKD that indicates anemia, a more pronounced inflammatory reaction, metabolic and electrolyte disturbances in patients with CKD C5d.
Sleep quality, anxiety, depression, pain, general health and laboratory tests in CKD C5d patients
According to multivariate regression adjusted to gender and age, an association between CKD C5d as an independent variable with RLS (OR 2.06 [0.58—3.81] 95% CI, p=0.010), lower OSA risk (OR –1.761 [–3.607—(–0.199)] 95% CI, p=0.038) and general health VAS scores (b= –16.58 points, p=0.001) as dependent variables was established (Fig. 2). In addition to the well-known associations of CKD C5d with urea (b= 11.93 mmol/l, p<0.001) and creatinine (b= 0.65 mmol/l, p <0.001), CKD C5d also correlated with following laboratory tests (Fig. 2): RBC (b= –0.76 · 1012/l, p<0.001), hemoglobin (b= –22.16 g/l, p<0.001), WBC (b= –0.63 · 109/l, p<0.001), amylase (b= 32.45 UI/l, p=0.005), potassium (b= 1.20 mmol/l, p<0.001), sodium (b= –2.50 mmol/l, p<0.001), total bilirubin (b= –5.95 μmol/l, p=0.01).
Fig. 2. Multiple regression coefficient plot depicting the associations between ESRD-HD as independent variable with the investigated parameters as dependent variables adjusted for sex and age.
Only models with significant associations between chronic kidney disease and other parameters are shown. Continuous variables (VAS for general health, RBC count, hemoglobin, WBC count, amylase, potassium, sodium, total bilirubin, and urea) are show as z-values for the relative estimation of the association strength.
Association of sleep quality, anxiety, depression, pain, general health, and laboratory tests
A regression analysis revealed an association between sleep quality (dependent variable) and RLS (independent variable) adjusted to age and gender (b= 0.36, p= 0.015) that was aligned with a lower sleep quality in RLS patients (12, 43 (10.15—15.25)) compared with patients without RLS (11.00 (6.00—11.00), p=0.040), Fig. 3, a. However, no significant differences were found between sleep quality ((11.91 (8.50—14.50)) in CKD C5d patients with RLS and C5d CKD patients without RLS (11.75 (8.75—13.75), p= 0.868).
Fig. 3 (a—k). The associations between sleep parameters, anxiety, depression, pain, perception of general health and blood markers.
a — RLS patients have lower sleep quality compared to the patients without RLS; b — aleep quality is associated with creatinine level; c — EDS is higher in RLS patients, compared to patients without RLS; d — FSS is associated with eosinophil count; e — insomnia is associated with anxiety; f — insomnia is associated with depression (BDI); g — creatinine level is higher in RLS patients compared to patients without RLS; h — urea level is higher in RLS patients compared to patients without RLS; i — potassium level is higher in RLS patients\n compared to patients without RLS; j — VAS score for general health is associated with sleep quality; k — anxiety is associated with RBC count.
ESRD-HD — end-stage renal disease (hemodialysis), RLS — restless leg syndrome, PSQI — Pittsburg Sleep Quality Index, ESS — Epworth Sleepiness Scale, FSS — Fatigue Severity Scale, ISI — Insomnia Severity Index, OSA — obstructive sleep apnea, RBC — red blood cells, VAS — visual analogue scale, HADS — Hospital anxiety and depression scale, BDI — Beck Depression Inventory.
In addition, the regression analysis revealed association between sleep quality (dependent variable) and creatinine level (independent variable): b= 0.36, p=0.015 adjusted to age and gender, and b= 1.07, p=0.012 adjusted to age, gender and CKD S5d. This association was also observed in a simple correlation analysis (ρ=0.48, p=0.003) and partial correlations taking into account CKD (ρ=0.34, p=0.039; Fig. 3, a).
According to regression analysis, an association between the severity of EDS (dependent variable) and RLS (independent variable) adjusted to age and gender (b= 3.56, p=0.023) was established that aligned to higher EDS in patients with RLS (16.93 (15.00—19.25)) compared with patients without RLS (13.42 (11.00—16.00), p=0.005; Fig. 3, c)). Significant differences were found between IDS in CKD C5d patients with RLS — 17.27 (15.50—18.50) points and CKD C5d patients without RLS — 13.12 (10.75—16.25) points (p=0.042).
According to the regression analysis, an association between fatigue severity as dependent variable with absolute one (b=3.03 points, p=0.017 adjusted to age and gender and b=3.13 points, p=0.012 adjusted to age, gender and CKD C5d) and the relative (b=0.18 points, p=0.016 adjusted to age and gender and b=0.17 points, p=0.017 adjusted to age, gender and CKD C5d) the number of eosinophils as independent variable was established. The association between fatigue severity and absolute number of eosinophils was significant in partial correlations when CKD is presented (ρ=0.34, p=0.035; Fig. 3, d).
According to the regression analysis, an association between insomnia severity (dependent variable) and HADS anxiety subscale points (b=0.21 points, p<0.001 adjusted to age and gender and b=0.24 points, p<0.001 adjusted to age, gender and CKD C5e) and the number of BDI points (b=0.05 points, p=0.013 adjusted to age and gender and b=0.06 points, p=0.005 adjusted to age, gender and C5d CKD) was established. This association was also observed in simple correlation analysis (ρ=0.46, b=0.003) and partial correlations when CKD is presented (ρ=0.47, p=0.003; Fig. 3, f).
According to the regression analysis, the association between RLS (dependent variable) and the urea level (OR 0.14 [0.04—0.27] 95% CI, p=0.015), creatinine (OR 2.74 [0.76—5, 08] 95% CI, p=0.011) and potassium (OR 1.23 [0.28—2.29] 95% CI, p = 0.019) adjusted to age and gender was established. These associations were aligned with higher creatinine level (0.62 (0.23—0.84) mmol/l in patients with RLS and 0.28 (0.07—0.55) mmol/l in patients without RLS, p=0.002), urea (15.29 (8.57—19.67) mmol/l in patients with RLS and 8.84 (4.40—14.25) mmol/l in patients without RLS, p=0.007) and potassium (5.47 (5.00—5.90) mmol/l in patients with RLS and 4.74 (4.20—5.20) mmol/l in patients without RLS, p=0.009) in patients with RLS compared to patients without RLS (Fig. 3, g-i). However, when analyzing the differences between these indicators in C5d CKD patients with RLS and without RLS there no significant differences were found between creatinine level (0.77 (0.65—0.90) mmol/l in C5d CKD patients with RLS and 0.68 (0.55—0.79) mmol/l in CKD C5d patients without RLS, p=0.321), urea (17.78 (14.50—20.20) mmol/l and 16.41 (17.75—18.75) mmol/l, respectively, p=0.563) and potassium (5.57 (5.20—5.90) mmol/l and 5.56 (5.20—6.15) mmol/l, respectively, p=0.836) (Fig. 3, g—i).
Correlation and regression analyzes did not reveal an association of OSA risk with indicators of sleep, anxiety, depression, pain, general health and laboratory tests.
According to the regression analysis, an association between the subjective health perception (dependent variable) and sleep quality according to PSQI (independent variable) (b= –2.06 points, p<0.001 adjusted to age and gender and b= –1.66 points, p=0.002 adjusted to age, gender and CKD C5d) was established. This association was also observed in simple correlation analysis (ρ= –0.64, p<0.001) and partial correlations when CKD is presented (ρ= –0.59, p<0.001; Fig. 3, j).
According to the regression analysis adjusted to age and gender, an association between HADS anxiety subscale points (b= 1.81 points, p=0.023), the HADS depression subscale (b= 0.58 points, p=0.022), BDI (b= 1.44 points, p<0.001) as dependent variables with the number of erythrocytes was established. These relationships were partially confirmed by a simple correlation analysis (ρ= 0.39, p=0.014; ρ= 0.30, p=0.066; ρ= 0.35, p=0.027, respectively). However, this association was not significant in correlation and regression analysis when CKD C5d is presented. The association between HADS anxiety subscale points (ρ= 0.47, p=0.043), the HADS depression subscale (ρ= 0.75, p<0.01) and BDI (ρ= 0.73, p<0.01) was observed by Spearman correlation coefficient in patients with CKD C5d only (Fig. 3, k—m).
Fig. 3 (l—m). The associations between sleep parameters, anxiety, depression, pain, perception of general health and blood markers.
l — depression (HADS) is associated with RBC count; m — depression (BDI) is associated with RBC count.
ESRD-HD — end-stage renal disease (hemodialysis), RLS — restless leg syndrome, PSQI — Pittsburg Sleep Quality Index, ESS — Epworth Sleepiness Scale, FSS — Fatigue Severity Scale, ISI — Insomnia Severity Index, OSA — obstructive sleep apnea, RBC — red blood cells, VAS — visual analogue scale, HADS — Hospital anxiety and depression scale, BDI — Beck Depression Inventory.
According to correlation and regression analysis, no association between pain intensity associated with osteoarthritis and the studied indicators of sleep, anxiety, depression, pain, general health and laboratory tests was found.
Discussion
This pilot cross-sectional case-control study was the first to examine the subjective characteristics of sleep, general health, anxiety and depression, pain and laboratory tests in osteoarthritis patients with CKD C5d and non-CKD.
Poor sleep quality was found in 100% of patients both with CKD C5d and non-CKD. The higher incidence of poor sleep quality compared with other studies showing it in 53—74% of CKD C5d patients [20—22] may be explained by concomitant somatic pathology, in particular osteoarthritis.
The association between sleep quality according to PSQI and the subjective health perception has been established that is aligned with current literature data [52, 53]. A worse sleep quality and a subjective health perception were revealed in osteoarthritis patients associated with CKD C5d compared with non-CKD osteoarthritis patients that is aligned with the results of previous similar studies involving CKD C5d and non-CKD patients [21].
In the studied cohort, some particular sleep disorders were noted in CKD and non-CKD patients.
First, osteoarthritis patients with CKD C5d had a higher incidence of RLS (59%) compared with osteoarthritis patients without C5d CKD (17%) that is aligned with similar studies that estimate the RLS prevalence at 13—55% in CKD patients compared with 10—15% in the general population [15, 16, 23, 24]. The association of RLS with sleep quality and IDS has been established. However, when analyzing the differences in sleep quality and EDS in C5d CKD patients with and without RLS the significant differences were found only for EDS confirming well-known association of RLS and EDS [54].
Secondly, in osteoarthritis patients with CKD C5d there was a tendency towards a decrease in OSA risk compared with osteoarthritis patients without CKD. That may be explained by lower BMI in CKD C5d patients compared to non-CKD patients, since BMI is one of the most important risk factors for OSA, as well as the lower accuracy of Berlin questionnaire to assess OSA in CKD patients [55].
Current literature data indicate a higher level of anxiety (12—52%) and depression (20—30%) in CKD C5d patients compared with the general population (1—12 and 27%) [56—58]. However, in this study the osteoarthritis patients with CKD C5d had a lower level of anxiety (5%) and a tendency to a lower level of depression (11—16%) compared with osteoarthritis patients without CKD (11 and 26—37%). That may be explained by decreased sensitivity of the questionnaire because of uremia [59] or personal characteristics of those osteoarthritis patients with CKD C5d who decided on surgical treatment of osteoarthritis. However, the well-known association of insomnia with anxiety and depression severity adjusted to age, gender and CKD C5d in patients with osteoarthritis was not recognized [60].
There were no expected significant differences in pain intensity associated with osteoarthritis in patients with osteoarthritis and CKD C5d and patients without osteoarthritis and CKD [61]. In addition, there were no association between pain intensity associated with osteoarthritis and sleep quality, anxiety, depression, pain, general health and laboratory tests that contradict existing ideas [62]. The lack of significant differences may be due to routine pain-relief medications and the small sample size.
This study also examined the association between sleep disorders, anxiety and depression, general well-being and laboratory tests.
A positive association was found between PSQI scores only with creatinine levels if it was adjusted to age, gender, and CKD C5d. This observation is inconsistent with the results of E. Iliescu et al. [22] denying a significant association between subjective sleep quality and creatinine levels in CKD C5d patients. However, our observation supports the association between poor sleep quality with proteinuria and decreased GFR in the general population [63, 64].
According to the literature, anemia and peripheral polyneuropathy are possible factors predisposing the RLS development in patients with CKD C5d [23]. In this study, the association of RLS with iron level of was not revealed. However, positive associations of RLS with the level of urea, creatinine and potassium were found if adjusted to age and gender. These associations were not significant in connection with CKD presence. Therefore, as other similar studies say [15, 16, 24] it may confirm the minor role of these factors in the pathogenesis of RLS in CKD C5d.
This study also found a positive association between fatigue and the absolute and relative number of eosinophils that was significant if adjusted to age, gender, and CKD C5d. This observation is partially consistent with the results of J. Steel et al. [65] who described the association between the number of eosinophils and the severity of fatigue, pain, and depression in patients with hepatobiliary cancer. Evidence on the association of eosinophil activation markers (eosinophilic cationic proteins) with chronic fatigue syndrome [66] and the number of eosinophils with sleep quality in patients [67] confirm this observation and may indicate activation of the inflammatory response in CKD C5d patients.
In addition, a direct association between severity of anxiety and depression and red blood cells count was found; it was observed in osteoarthritis patients with CKD C5d and not found in osteoarthritis patients without CKD. There is contradictory information in the literature on the association between anemia and anxiety and depression symptoms [68, 69]. However, there is evidence of direct association between anxiety and depression and volume distribution of erythrocytes [70] that is also increased in CKD C5d patients [71]. Therefore, it can be assumed there is a contribution of changed erythrocytes’ function to the development of anxiety and depression in CKD C5d patients possibly due to impaired nutrition and metabolism of the nervous tissue that requires clarification in prospective studies.
The advantages of this work include a comprehensive multifactorial assessment of subjective sleep and wakefulness disorders including an assessment of sleep quality, EDS, fatigue, insomnia, RLS as a risk of OSA in age-matched osteoarthritis patients with and without CKD. The disadvantages of this work are limited patients’ sample and lack of objective methods for sleep evaluation (polysomnography, actigraphy).
Conclusion
Thus, the obtained data indicate a high incidence of sleep disorders in osteoarthritis patients that requires special focus in case of concomitant CKD C5d due to the potential risk of complications associated with poor sleep quality and high RLS frequency. Based on the studied associations of sleep disorders with pain, anxiety, depression and laboratory tests it can be assumed the pathophysiological mechanisms of sleep disorders are common for patients with osteoarthritis regardless the CKD presence while changes in erythrocyte metabolism play a role in the pathogenesis of anxiety and depression in patients with osteoarthritis and CKD. The observed associations between sleep quality and health perception underlines the importance to detect the sleep disorders and improve the sleep quality in patients with osteoarthritis.
The authors declare no conflict of interest.