Participants
The examinees were mostly women (71.7%; n = 81) living in the city vs. village (65.8%; n = 50). The median age was 76 in city residents (min–max: 17–101) and 72 (min–max: 18–94) in village residents (Table 1).
Table 1. Socioeconomic situation in the respondent group*
Variable
|
Place of residence
|
n
|
M
|
SD
|
Q.25%
|
Me
|
Q.75%
|
min
|
max
|
Wilcoxon
Test
p
|
Age
(in years)
|
City
|
113
|
71.50
|
17.21
|
62.00
|
76.00
|
84.00
|
17.00
|
101.00
|
0.202
|
Village
|
75
|
69.25
|
16.56
|
62.50
|
72.00
|
81.50
|
18.00
|
94.00
|
Variable
|
Categories
|
City
|
Village
|
Fisher Test - p
|
n
|
%
|
n
|
%
|
Gender
|
Women
|
81
|
71.7
|
50
|
65.8
|
0.423
|
Men
|
32
|
28.3
|
26
|
34.2
|
Total
|
113
|
100
|
76
|
100
|
Legend: n-group quantity, % - percentage; M – mean; SD – standard deviation; Q.25% – first quartile; Me – median; Q.75% – third quartile; Min. – minimum; Max. – maximum; p – calculated level of significance for standard test Shapiro-Wilk; p – calculated level of significance for standard test Fisher. * The figures in column n do not sum up to 193 due to gaps in the questionnaires completed by the carers.
The evaluation of anxiety and depression
The analysis of anxiety and depression occurrence in the group of patients, calculated into standard ten, is presented in Table 2. The pathological disorder such as the anxiety syndrome was found in 61.1% (n = 72) patients living in cities and in 68.9% (n = 51) of rural residents. Low intensity (mild) of anxious behaviours was observed in 32.1% (n = 35) City vs. 28.4% (n = 21) Village. A lack of abnormalities was confirmed in 1.8% (n = 2) City vs. 2.7% (n = 2) Village. The analysis of HADS-M Depression variable found depression in 68.8% (n = 75) of city residents vs. 68% (51) of village patients. Mild depressive behaviours were found in 26.6% (n = 29) City vs. 28% (n=21) Village. Lack of depression was confirmed in 4.6% (n = 5) City vs. 4% (n = 3) Village. Lack of significant differences between city and village residents was observed at the level of significance equal to 0.05.
Table 2. The level of anxiety and depression by place of residence
Variable
|
Place of residence
|
n
|
M
|
SD
|
Q.25%
|
M3
|
Q.75%
|
min
|
max
|
Wilcoxon
Test
p
|
HADS-M
Anxiety
|
City
|
109
|
109
|
11.51
|
2.29
|
10.00
|
12.00
|
13.00
|
6.00
|
0.491
|
Village
|
74
|
74
|
11.73
|
2.26
|
10.00
|
12.00
|
13.00
|
6.00
|
HADS-M
Depression
|
City
|
109
|
11.51
|
2.25
|
10.00
|
12.00
|
13.00
|
5.00
|
17.00
|
0.64
|
Village
|
75
|
11.39
|
1.76
|
10.00
|
11.00
|
12.00
|
7.00
|
15.00
|
Variable
|
Categories
|
City
|
Village
|
Fisher Test - p
|
n
|
%
|
n
|
%
|
HADS-M
Anxiety
|
Lack of abnormality (0–7)
|
2
|
1.8
|
2
|
2.7
|
0.78
|
Border conditions (8–10)
|
35
|
32.1
|
21
|
28.4
|
Abnormality confirmed (11–21)
|
72
|
66.1
|
51
|
68.9
|
Total
|
109
|
100
|
74
|
100
|
HADS-M Depression
|
Lack of abnormality (0–7)
|
5
|
4.6
|
3
|
4
|
0.962
|
Border conditions (8–10)
|
29
|
26.6
|
21
|
28
|
Abnormality confirmed (11–21)
|
75
|
68.8
|
51
|
68
|
Total
|
109
|
100
|
75
|
100
|
Legend: n-group quantity, % - percentage; M – mean; SD – standard deviation; Q.25% – first quartile;
Me – median; Q.75% – third quartile; Min. – minimum; Max. – maximum; p – calculated level of significance for standard test Shapiro-Wilk; p – calculated level of significance for standard test Fisher. * The figures in column n do not sum up to 193 due to gaps in the questionnaires completed by the carers.
Correlations of the number of visits and interventions in the prior 12 months with the HADS-M scale
Statistically significant differences between the correlation coefficients of city and village residents were found relating to anxiety and the number of visits in cardiological clinics (p = 0.005) and the number of interventions of a family nurse (p = 0.03). In village patients the number of visits to cardiological units (r = 0.32, p = 0.005) and the number of nurse interventions (r = 0.25, p = 0.033) correlates positively with the HADS-M Anxiety scale. This correlation was not found in city patients. The correlation between the number of visits to cardiological units and HADS-M Anxiety was found to be insignificant (r = −0.1, p = 0.306) as well as between the number of a nurse interventions and HADS-M Anxiety (r = −0.06, p = 0.516) (Table 3).
Table 3. The correlation between number of visits, appointments and interventions in the prior 12 months with HADS-M scale
|
|
|
GP appointments
|
Visits in Cardiological Clinic
|
Family Nurse Interventions
|
Variable
|
|
|
r
|
r=0 p
|
*
|
n
|
r=r p
|
*
|
r
|
r=0 p
|
*
|
N
|
r=r p
|
*
|
r
|
r=0 p
|
*
|
n
|
r=r
|
*
|
HADS-M
|
Anxiety
|
City
|
-0.11
|
0.244
|
|
109
|
0.439
|
|
-0.1
|
0.306
|
|
109
|
0.005
|
*
|
-0.06
|
0.516
|
|
109
|
0.03
|
*
|
Village
|
-0.23
|
0.051
|
|
74
|
|
|
0.32
|
0.005
|
*
|
74
|
|
|
0.25
|
0.033
|
*
|
74
|
|
|
Depression
|
City
|
-0.21
|
0.031
|
*
|
109
|
0.694
|
|
-0.03
|
0.739
|
|
109
|
0.37
|
|
-0.05
|
0.628
|
|
109
|
0.35
|
|
Village
|
-0.15
|
0.204
|
|
75
|
|
|
0.1
|
0.374
|
|
75
|
|
|
0.09
|
0.427
|
|
75
|
|
|
Legend: (r) – Spearman’s rank correlation coefficient (0 means r ≤ 0.01), (r = 0 p) – calculated significance level for test verifying null hypothesis that correlation coefficient r equals 0; * appears in column (*) if p ≤ 0.05, then null hypothesis is rejected (0 means p <0.001) (r = r p) – calculated significance level for test verifying null hypothesis that two correlation coefficients are equal, * appears in column (*) if p ≤ 0.05, then null hypothesis is rejected (0 means p <0.001). * The figures in column n do not sum up to 161 due to gaps in the questionnaires completed by the carers.
Results of logistic regression
The logistic regression analysis in the group of CVD patients living in urban areas (Table 4a) led to the selection of models, which allowed for the odds ratio calculation (Table 4b).
Table 4a. The results of logistic regression analysis in group of city residents. Explained variable: HADS-M Anxiety (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory variables
|
bi
|
SEi
|
zi
|
pi = Pr(>|zi|)
|
Models with 4 explanatory variables
|
Model 1 (n=100)
|
Chi2 = 25.57, df = 4, p<0.001, pseudo R2 = 0.2
|
Free term
|
-
|
-
|
-
|
-
|
1
|
X18
|
WHOQOL-BREF physical domain
|
-0.231
|
0.116
|
-1.991
|
0.046
|
2
|
X19
|
WHOQOL-BREF psychological domain
|
0.399
|
0.147
|
2.723
|
0.006
|
3
|
X26
|
HBI - health practices
|
-0.583
|
0.225
|
-2.597
|
0.009
|
4
|
X28
|
HADS-M depression
|
1.406
|
0.481
|
2.924
|
0.003
|
Legend: Chi-squared – statistical hypothesis test of chi2 model adjustment; df – number of degrees of freedom;
p – calculated level of test significance (if p ≤ 0.05, model introduces relevant information as it differs significantly from free term model); pseudo R2 – value which evaluates explanatory variable anticipation according to the model (0 ≤ pseudo R2 < 1, the bigger the value the better the anticipation); bi – coefficient estimation in regression model; SEi – standard error estimation for bi coefficient; zi – value of test statistics in standard distribution; (pi = Pr (>|zi|) – calculated probability value pi for double-sided critical area equal to z (if pi ≤ 0,05, null hypothesis is rejected that bi coefficient =0 which means that i-variable is relevant in the model); n-group quantity
Table 4b presents the results of the odds ratio in a logistic regression model for the risk of anxiety occurrence in chronic CVD patients living in cities. It was found that in patients who differed at the level of Qol in physical domain of WHOQOL-BREF questionnaire by 1 degree, those with a lower score have a 1.25 times higher chance of anxiety occurrence than those with a higher result. In patients who differed in this score by 14.85, those with a lower one have a 30.93 times higher chance of developing anxiety.
In patients who differed in the level of Qol in the psychological domain of WHOQOL-BREF questionnaire by 1 degree, those with a higher score have a 1.49 times higher chance of anxiety occurrence than those with a lower one. In patients who differed in this score by 14.66, those with a higher result have a 48.75 times higher chance of anxiety abnormalities.
It was found that in patients who differed in the level of Qol in the intensity of health practices of WHOQOL-BREF questionnaire by 1 degree, those with a lower score have a 1.79 times higher chance of anxiety occurrence than those with a higher one. In patients who differed in this score by 2.83, those with a lower one have a 5.21 times higher chance of developing anxiety.
It was also discovered that for patients who differed in the assessment of depression in HADS-M scale by 1 degree, those with a higher score have a 4.07 times higher chance of anxiety occurrence than those with a lower one.
Table 4b. The odds ratio in the model of logistic regression in group of city residents. Explained variable: HADS-M Anxiety (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory variables
|
Per unit
|
On the Range
|
OR
|
95% CI
|
1/OR
|
OR
|
95% CI
|
1/OR
|
range
|
Model 1
|
|
|
|
|
|
|
|
|
|
|
|
X18
|
WHOQOL-BREF physical domain
|
0.79
|
0.62
|
–
|
0.98
|
1.25
|
0.03
|
0.00
|
–
|
0.84
|
30.93
|
14.85
|
X19
|
WHOQOL-BREF psychological domain
|
1.49
|
1.13
|
–
|
2.01
|
0.67
|
48.75
|
6.17
|
–
|
29800
|
0.00
|
14.66
|
X26
|
HBI - Health Practices
|
0.55
|
0.35
|
–
|
0.85
|
1.79
|
0.19
|
0.05
|
–
|
0.64
|
5.21
|
2.83
|
X28
|
HADS-M Depression
|
4.07
|
1.62
|
–
|
10.83
|
0.24
|
4.07
|
1.62
|
–
|
10.80
|
0.24
|
1.00
|
Legend: OR - odds ratio, CI - 95% confidence interval for OR
The logistic regression analysis in the group of CVD patients living in rural areas (Table 5a) led to the selection of models, which allowed for odds ratio calculations (Table 5b).
Table 5a. The results of logistic regression analysis in the village group. Explained variable: HADS-M Anxiety (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory variables
|
bi
|
SEi
|
zi
|
pi = Pr(>|zi|)
|
Models with 6 explanatory variables
|
Model 1 (n=72)
|
Chi2 = 40.05, df = 6, p<0.001, pseudo R2 = 0.4
|
Free term
|
-
|
-
|
-
|
-
|
1
|
X9
|
Age (in years)
|
-0.040
|
0.020
|
-2.014
|
0.044
|
2
|
X15
|
Camberwell
|
-5.683
|
2.741
|
-2.073
|
0.038
|
3
|
X18
|
WHOQOL-BREF physical domain
|
0.446
|
0.185
|
2.405
|
0.016
|
4
|
X20
|
WHOQOL-BREF social relations domain
|
-0.427
|
0.214
|
-1.997
|
0.046
|
5
|
X21
|
WHOQOL-BREF environment domain
|
0.516
|
0.260
|
1.987
|
0.047
|
6
|
X28
|
HADS-M depression
|
1.768
|
0.698
|
2.532
|
0.011
|
Models with 4 explanatory variables
|
bi
|
SEi
|
zi
|
pi = Pr(>|zi|)
|
Model 2 (n=73)
|
Chi2 = 24.39, df =4 , p<0.001, pseudo R2 = 0.24
|
Free term
|
-
|
-
|
-
|
-
|
1
|
X4
|
Number of home visits by a nurse during the last 12 months
|
-0.102
|
0.050
|
-2.059
|
0.040
|
2
|
X15
|
Camberwell
|
-4.397
|
2.098
|
-2.096
|
0.036
|
3
|
X19
|
WHOQOL-BREF psychological domain
|
0.281
|
0.124
|
2.276
|
0.023
|
4
|
X28
|
HADS-M depression
|
1.201
|
0.559
|
2.148
|
0.032
|
Legend: Chi-squared – statistical hypothesis test of chi2 model adjustment; df – number of degrees of freedom;
p – calculated level of test significance (if p ≤ 0.05, model introduces relevant information as it differs significantly from free term model); pseudo R2 – value which evaluates explanatory variable anticipation according to the model (0 ≤ pseudo R2 < 1, the bigger the value the better anticipation); bi – coefficient estimation in regression model; SEi – standard error estimation for bi coefficient; zi – value of test statistics in standard distribution; (pi = Pr (>|zi|) – calculated probability value pi for double-sided critical area equal to z (if pi ≤ 0,05, null hypothesis is rejected that bi coefficient =0 which means that i-variable is relevant in the model); n-group quantity.
Table 5b presents the results of the odds ratio in a logistic regression model for the risk of anxiety occurrence in chronic CVD patients living in villages.
It was found that for patients who differed in age by 1 year, younger patients have a 1.04 times higher chance of anxiety occurrence than older patients, while in patients who differed by 76 years the chance was 21.2 times higher.
In patients who differed in the Camberwell evaluation of needs by 0.83, those with a lower score have a 113 times higher chance of anxiety occurrence than those with a higher assessment. In patients who differed in the assessment by 1, those with a lower assessment have a 293.86 times higher chance of such abnormalities.
In patients who differed in the level of Qol in physical domain of WHOQOL-BREF questionnaire by 1 degree, those with a higher score have a 1.49 times higher chance of developing anxiety than those with a lower one. In patients who differed in this assessment by 14.28, those with a higher assessment have a 585 times higher chance of anxiety disorders.
In patients who differed in the level of Qol in social relations domain of WHOQOL-BREF questionnaire by 1 degree, those with a lower score have a 1.53 times higher chance of anxiety occurrence than those with a higher one. In patients who differed in this assessment by 14.66, those with a lower one have a 522 times higher chance of anxious behaviours.
It was also discovered that in patients who differed in the level of Qol in the environment domain of WHOQOL-BREF questionnaire by 1 degree, those with a higher score have a 1.67 times higher chance of anxiety occurrence than those with a lower one. In patients who differed in this assessment by 11.92, those with a higher one have a 473 times higher chance of anxiety disorders.
It was confirmed that in patients who differed in the assessment of depression by 1 on HADS-M scale, those with a higher score have a 5.85 times higher chance of anxiety disorders than those with a lower score.
In patients who differed in the number of a family nurse visits in the last 12 months by 1, it was found that those with a lower number of visits have a 1.11 times higher chance of anxiety development than those with a greater number of visits. However, in patients who differed in the number of visits by 28, those with a smaller number have a 51.39 times higher chance of such behaviours.
Patients who differed in the level of Qol in psychological domain of WHOQOL-BREF questionnaire by 1 degree, those with a higher score have a 1.32 times higher chance of anxiety occurrence than those with a lower score. But, in patients who differed in this assessment by 14, those with a higher score have a 51.39 times higher chance of developing anxiety.
Table 5b. The odds ratio in the model of logistic regression in the village group. Explained variable: HADS-M Anxiety (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory variables
|
Per unit
|
On the Range
|
|
OR
|
95% CI
|
1/OR
|
OR
|
95% CI
|
1/OR
|
range
|
Model 1
|
|
|
|
|
|
|
|
|
|
|
|
X9
|
Age (in years)
|
0.96
|
0.919
|
–
|
0.99
|
1.04
|
0.05
|
0.002
|
–
|
0.73
|
21.2
|
76.00
|
X15
|
Camberwell
|
0.00
|
0.00001
|
–
|
0.56
|
293.86
|
0.01
|
0.00007
|
–
|
0.62
|
113
|
0.83
|
X18
|
WHOQOL-BREF physical domain
|
1.56
|
1.11
|
–
|
2.33
|
0.64
|
585
|
4.58
|
–
|
177000
|
0.002
|
14.28
|
X20
|
WHOQOL-BREF social relations domain
|
0.65
|
0.40
|
–
|
0.94
|
1.53
|
0.002
|
0.000002
|
–
|
0.46
|
522
|
14.66
|
X21
|
WHOQOL-BREF environment domain
|
1.67
|
1.06
|
–
|
3.00
|
0.59
|
473
|
2.11
|
–
|
503000
|
0.002
|
11.92
|
X28
|
HADS-M depression
|
5.85
|
1.58
|
–
|
25.66
|
0.17
|
5.85
|
1.58
|
–
|
25.60
|
0.17
|
1.00
|
Model 2
|
|
|
|
|
|
|
|
|
|
|
|
X4
|
Number of home visits by a nurse during the last 12 months
|
0.90
|
0.81
|
–
|
0.99
|
1.11
|
0.06
|
0.01
|
–
|
0.83
|
17.64
|
28
|
X19
|
WHOQOL-BREF psychological domain
|
1.32
|
1.05
|
–
|
1.71
|
0.75
|
51.39
|
2.03
|
–
|
1960.55
|
0.02
|
14
|
Legend: OR - odds ratio, CI - 95% confidence interval for OR
The analysis of logistic regression in chronically ill CVD patients living in rural areas (Table 6a) led to the identification of models which permitted calculation of the odds ratio (Table 6b).
Table 6a. The results of logistic regression analysis in the village group. Explained variable: HADS-M Depression (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory Variables
|
bi
|
SEi
|
zi
|
pi = Pr(>|zi|)
|
Models with 3 explanatory variables
|
Model 1 (n=73)
|
Chi2 = 19.55, df = 3, p<0.001, pseudo R2 = 0.19
|
Free term
|
-
|
-
|
-
|
-
|
1
|
X18
|
WHOQOL-BREF physical domain
|
-0.334
|
0.167
|
-2.003
|
0.045
|
2
|
X19
|
WHOQOL-BREF psychological domain
|
0.321
|
0.162
|
1.984
|
0.047
|
3
|
X28
|
HADS-M anxiety
|
1.528
|
0.609
|
2.507
|
0.012
|
Legend: Chi-squared – statistical hypothesis test of chi2 model adjustment; df – number of degrees of freedom;
p – calculated level of test significance (if p ≤ 0.05, model introduces relevant information as it differs significantly from free term model); pseudo R2 – value which evaluates explanatory variable anticipation according to the model (0 ≤ pseudo R2 < 1, the bigger the value the better anticipation); bi – coefficient estimation in regression model; SEi – standard error estimation for bi coefficient; zi – value of test statistics in standard distribution; (pi = Pr (>|zi|) – calculated probability value pi for double-sided critical area equal to z (if pi ≤ 0,05, null hypothesis is rejected that bi coefficient = 0 which means that i-variable is relevant in the model); n-group quantity.
Table 6b presents the results of the odds ratio in the model of logistic regression for the risk of depression in chronically ill CVD patients living in rural areas.
In patients who differed in the level of Qol in physical domain of WHOQOL-BREF questionnaire by 1 degree, those with a lower score have a 1.39 times higher chance of depression than those with a higher score. In patients who differed in this assessment by 14.28, those with a lower score have a 118.39 times higher chance of developing such conditions.
In patients who differed in the level of Qol in the psychological domain of WHOQOL-BREF questionnaire by 1 degree, those with a higher score have a 1.37 times higher chance of depressive behaviours than those with a lower score. For patients who differed in this assessment by 14, those with a higher score have 89.07 times higher chance of such abnormalities.
It was also found that in patients who differed in the assessment of anxiety on HADS-M scale by 1, those with a higher score have a 4.6 times higher chance of depression than those with a lower score.
Table 6b. The odds ratio in the model of logistic regression in the village group. Explained variable: HADS-M Depression (0—lack of abnormalities, 1—abnormality confirmed)
Explanatory Variables
|
Per unit
|
On the Range
|
|
|
OR
|
95% CI
|
1/OR
|
OR
|
95% CI
|
1/OR
|
range
|
|
Model 1
|
|
|
|
|
|
|
|
|
|
|
|
X18
|
WHOQOL-BREF physical domain
|
0.71
|
0.50
|
–
|
0.97
|
1.39
|
0.01
|
0.00005
|
–
|
0.71
|
118.39
|
14.28
|
X19
|
WHOQOL-BREF psychological domain
|
1.37
|
1.01
|
–
|
1.93
|
0.72
|
89.07
|
1.31
|
–
|
10600
|
0.01
|
14.00
|
X28
|
HADS-M anxiety
|
4.60
|
1.45
|
–
|
16.28
|
0.21
|
4.60
|
1.45
|
–
|
16.2
|
0.21
|
1.00
|
Legend: OR - odds ratio, CI - 95% confidence interval for OR