Recruitment and Attrition
A total of 130 participants were recruited, and n=90 met the inclusion and exclusion criteria. Participants were equally divided into experimental (EXP)=45(50%) and waitlist control (WLC)=45(50%) and they were statistically analyzed. There was no significant discrepancies were found between EXP vs. WLC among the demographic characteristics; such as age 36.93±6.867 vs. 37.62±6.767, t=-.479, p>.633, gender (X2= .720, p>.396), education (X2=1.090, p>.278), glucose level (X2=-1.113, p>.296), family system (X2=1.196, p>.271), occupation (X2=1.607, p>.205), socioeconomic status (X2=.611, p>.737), duration of illness (X2=4.305, .230), types of treatment (X2=.743, p>.690), duration of treatment (X2=.719, p>.698), checkup (X2=.958, p>..619), hypertension (X2=.776, p>.378), BMI (X2=.539, p>.764), family history (X2=.756, p>.381) and obesity (X2=.443, p>.506) respectively (see Table 1).
Table 1 Comparison of participants' demographic characteristics groups wise and overall
Variables
|
Category
|
N
|
Groups
|
|
|
Characteristics
|
|
|
Experimental
|
Control
|
X2/t
|
P
|
N Targeted
|
130
|
|
|
|
|
N Allocated, n
|
90
|
45
|
45
|
|
|
N final n(%)
|
62
|
27(33.3%)
|
35(43.2%)
|
|
|
Age, M(SD)
|
|
90
|
36.93(6.867)
|
37.62(6.767)
|
-.479
|
.633
|
Gender
|
Female n(%)
|
50
|
23(51.2%)
|
27(60.1%)
|
.720
|
.396
|
|
Male n(%)
|
40
|
22(48.9%)
|
18(40.0%)
|
|
|
Education
|
<Matric n(%)
|
35
|
17(37.0%)
|
18(33.3%)
|
1.090
|
.278
|
|
Matric n(%)
|
50
|
26(57.7%)
|
24(45.3%)
|
|
|
|
Above n(%)
|
5
|
2(4.3%)
|
3(4.3%)
|
|
|
Glucose Level
|
<250 n(%)
|
12
|
8(17.7%)
|
4(8.8%)
|
-1.113
|
.269
|
|
>250-350 n(%)
|
48
|
19(42.2%)
|
29(64.5%)
|
|
|
|
>350-500 n(%)
|
30
|
18(40.0%)
|
12(26.7%)
|
|
|
Family System
|
Nuclear n(%)
|
57
|
31(68.9)
|
26(57.7%)
|
1.196
|
.274
|
|
Joint n(%)
|
33
|
14(31.2%)
|
19(42.3%)
|
|
|
Occupation
|
Unemployed n(%)
|
48
|
21(46.7%)
|
27(60.0%)
|
1.607
|
.205
|
|
Employee n(%)
|
42
|
24(53.4%)
|
18(40.0%)
|
|
|
SES
|
Low n(%)
|
51
|
24(53.3%)
|
27(60.1%)
|
.611
|
.737
|
|
Middle n(%)
|
29
|
15(33.3%)
|
14(31.2%)
|
|
|
|
High n(%)
|
10
|
6(13.3%)
|
4(8.8%)
|
|
|
Duration. of.
|
>2 years n(%)
|
28
|
12(26.6%)
|
16(35.5%)
|
4.305
|
.230
|
Illness
|
3-5 years n(%)
|
40
|
18(40.0%)
|
22(48.9%)
|
|
|
|
5-10 years n(%)
|
21
|
14(31.2%)
|
7(15.5%)
|
|
|
Type of
|
Insulin n(%)
|
41
|
22(48.8%)
|
19(42.3%)
|
.743
|
.690
|
Treatment
|
Medication n(%)
|
42
|
19(42.3%)
|
23(51.2%)
|
|
|
|
Both n(%)
|
7
|
4(8.8%)
|
3(6.6%)
|
|
|
Duration of
|
>6-12months n(%)
|
38
|
19(42.3%)
|
19(42.3%)
|
.719
|
.698
|
Treatment
|
2-3 years n(%)
|
31
|
14(31.1%)
|
17(37.8%)
|
|
|
|
4-6 years n(%)
|
21
|
12(26.6%)
|
9(20.0%)
|
|
|
Checkup
|
Daily n(%)
|
37
|
19(42.3%)
|
18(40.1%)
|
.958
|
.619
|
|
Weekly n(%)
|
28
|
16(35.6%)
|
12(26.7%)
|
|
|
|
Monthly n(%)
|
25
|
14(31.2%)
|
11(24.5%)
|
|
|
Hypertension
|
yes n(%)
|
58
|
27(60.1%)
|
31(68.9%)
|
.776
|
.378
|
|
No n(%)
|
32
|
18(40.1%)
|
14(31.1%)
|
|
|
BMI
|
Healthy n(%)
|
26
|
12(26.6%)
|
14(31.1%)
|
.539
|
.764
|
|
Overweight n(%)
|
25
|
14(31.1%)
|
11(24.5%)
|
|
|
|
Obese n(%)
|
39
|
19(42.2%)
|
20(44.5%)
|
|
|
Family History
|
Present n(%)
|
56
|
26(57.8%)
|
30(66.7%)
|
.756
|
.384
|
|
Not Present n(%)
|
34
|
19(42.2%)
|
15(33.4%)
|
|
|
Obesity
|
Yes n(%)
|
59
|
31(68.9%)
|
28(62.3%)
|
.443
|
.506
|
|
No n(%)
|
31
|
14(31.2%)
|
17(37.8%)
|
|
|
Similarly, insignificant discrepancies were found between EXP vs. WLC on psychological measures; such as, PHQ-9 (16.18±3.85 vs. 15.73±3.83, p> .05), SHAI (30.11±6.55 vs. 30.02±7.16, p> .05), EBS (3.70±0.86 vs. 3.77±0.734, p> .05), PDS (4.85±0.75 vs. 4.17±0.80, p> .05), RDS (4.65±0.76 vs. 4.13±0.82, p> .05), IDS (3.52±1.14 vs. 3.82±0.86, p> .05), DDS (16.45±2.73 vs. 15.90±2.76, p> .05), PBNA (9.11±2.51 vs. 9.49±2.55, p> .05), ADPB (7.76±2.14 vs. 7.53±2.19, p> .05), CRNA (3.98±1.43 vs. 3.76±1.30, p> .05), GMAS (20.84±4.82 vs. 20.78±5.00, p> .05), SATS (67.11±13.53 vs. 70.30±13.74, p> .05), IMPS (69.89±12.32 vs. 74.11±12.94, p> .05), WORS (64.89±16.11 vs. 69.19±18.11, p> .05), DQLS (67.45±10.58 vs. 71.21±12.12, p> .05), WALS (88.24±14.35 vs. 87.80±15.20, p> .05), MPAS (78±12 vs. 79±13.10, p> .05), VPAS (65.35±12.36 vs. 64.95±11.93, p> .05) IPAQ(98.26±33.10 vs. 96.56±34.25, p> .05).
After follow-up treatment n=62 participants successfully completed the all stages of treatment (see Figure 1). The sample demographic characteristics were, as men were 27(43.6%) with EXP=14(51.9%) and WLC=13(48.1%), and females were 35(56.4%) including EXP=16(45.7%) and WLC=19(54.3%). Participants’ educational level was as, below matric 22(35.5%) with EXP=10(45.5%), and WLC=12(54.5%), matric-intermediate 19(30.6%) with EXP= 9(46.4%) and WLC=10(52.6%), and above intermediate were 21(33.9%) with EXP=12(57.2%) and WLC=9(42.8%). Subjects with the nuclear family system were 30(51.6%) with EXP=14(46.7%) and WLC=16(53.3%), and joint family system were 32(48.4%) with EXP=15(46.9%) and WLC=17(53.1%). Unemployed were 23(37.1%) EXP=13(56.5%) and WLC=10(43.5%), employees were 20 (32.3%) with EXP=13(56.5%) and WLC=10(43.5%) and running own business were 19(30.6%) with EXP=9(47.4%) and WLC=10(52.6%). Participants from middle socioeconomic status were 20(32.3%) with EXP=11(55.0%) and WLC=9(45.0%), upper-middle were 23(37.1%) with EXP=12(52.2%) and WLC=11(47.8%) and from lower high economic status were 19(30.6%) with EXP=9(47.4%) and WLC=10(52.6%). Participants visited their medical consultant on weekly basis were 26(41.9%) EXP=12(46.2%) and WLC=14(53.8%) and on monthly basis were 36(58.1%) with EXP=17(47.3%) and WLC=19(52.7%). On BMI with normal weight 29(46.8%) with EXP=16(55.2%) and WLC=13(44.8%) and overweight were 33(53.2%) with EXP=14(42.4%) and WLC=19(57.6%). Participants with family history illness were 34(54.8%) with EXP=18(52.9%) and WLC=16(47.1%) and with no history of family illness were 28(45.2%) with EXP=13(46.4%) and WLC=15(53.6%) respectively.
Table 2 Mean (standard deviation) and repeated-measure ANOVA of clinical scores during pre- and post-test interventions
Groups
|
|
|
|
|
|
η2
|
|
Experimental
|
Waitlist-Control
|
|
|
Baseline
|
Post-Test
|
Baseline
|
Post-Test
|
|
|
M(SD)
|
M(SD)
|
M(SD)
|
M(SD)
|
F
|
PHQ-9
|
15.85(3.81)
|
8.81(2.588)
|
15.80(4.115)
|
16.54(3.973)
|
94.436***
|
.611
|
SHAI
|
29.81(6.83)
|
15.26(1.933)
|
29.49(7.172)
|
32.29(6.042)
|
201.915***
|
.771
|
EBS
|
3.72(.898)
|
2.36(.550)
|
3.79(.743)
|
3.93(.779)
|
103.637***
|
.633
|
PDS
|
4.57(0.73)
|
2.67(0.723)
|
4.10(0.857)
|
4.04(0.783)
|
129.088***
|
.683
|
RDS
|
4.60(0.716)
|
2.71(0.687)
|
4.12(.875)
|
3.85(.754)
|
118.629***
|
.664
|
IDS
|
3.56(1.113)
|
2.31(.613)
|
3.75(.927)
|
3.61(.865)
|
29.704***
|
.310
|
DDS
|
16.46(2.710)
|
10.05(2.220)
|
15.77(2.985)
|
15.45(2.927)
|
222.710***
|
.788
|
PBNA
|
8.78(2.342)
|
28.19(1.406)
|
9.51(2.393)
|
9.77(1.987)
|
56.327***
|
.484
|
ADPB
|
7.48(2.190)
|
10.93(1.328)
|
7.49(2.063)
|
7.29(1.775)
|
43.651***
|
.421
|
CRNA
|
3.96(1.506)
|
5.89(.320)
|
3.77(1.308)
|
3.37(1.262)
|
61.649***
|
.507
|
GMAS
|
20.22(4.726)
|
28.19(3.039)
|
20.77(4.602)
|
20.86(3.797)
|
67.579***
|
.566
|
SATS
|
67.65(13.20)
|
49.51(11.756)
|
69.05(14.383)
|
69.71(11.444)
|
41.371***
|
.408
|
IMPS
|
68.89(12.274)
|
47.04(11.288)
|
73.43(13.654)
|
71.71(9.848)
|
44.126***
|
.424
|
WORS
|
64.69(15.886)
|
45.93(13.754)
|
68.19(17.792)
|
72(14.353)
|
40.860***
|
.405
|
DQLS
|
67.35(10.402)
|
12.85(9.986)
|
70.20(12.785)
|
70.86(9.843)
|
83.352***
|
.581
|
WALS
|
389.27(323.7)
|
1103.6(260.90)
|
550.3(505.5)
|
550.3(505.5)
|
134.7***
|
.695
|
MPAS
|
171.1(286.20)
|
456.2(453.80)
|
84.0(240.50)
|
96.0(271.10)
|
17.61***
|
.227
|
VPAS
|
41.48(149.45)
|
528.88(598.18)
|
96.0(395.60)
|
96.0(395.60)
|
29.738***
|
.331
|
IPAQ
|
601.8(512.84)
|
20.88(925.4)
|
735.6(861.8)
|
747.97(889.88)
|
164.245***
|
.736
|
Note: ***= p<.001; η2=Partial Square Eta; PHQ-9: Patients Health Questionnaire; SHAI: Short Health Anxiety Inventory; EBS: Emotional Burden Subscale; PDS: Physician Distress Subscale; RDS: Regimen Distress Subscale; IDS: Interpersonal Distress Subscale; DDS. Diabetes Distress Scale; PBNA: Patient Behavior related Non-Adherence; ADPB Additional Disease and Pill Burden Related Non-Adherence; CRNA: Cost Related Non-Adherence; GMAS: General Medication Adherence Scale; SATS: Satisfaction Subscale; IMPS: Impact Subscale; WORS: Worry Subscale; DQLSL: Diabetes Quality of Life Scale; WALS: Walking Subscale; MPAS: Moderate Physical Activity Subscale; VPAS: Vigorous Physical Activity Subscale; IPAQ: International Physical Activity Questionnaire.
Primary and Secondary outcomes
Findings reported that there is significant mean difference between EXP and WLC in the baseline scores and post-testing scores on PHQ (i.e. F (1,60) = 94.436, p<.000, η2= .611) with large effect size. This indicates CBT substantially decreased the level of depressive symptoms among patients with T2DM. Findings indicate the significant mean differences were found between EXP and WLC groups on SHAI (i.e. F(1,60)= 201.915, p<.000, η2= .771) with large effect size. The large effect size indicates the CBT 77% played a role to reduce the level of health anxiety in experimental group. Furthermore, significant mean differences were investigated between baseline and post-testing scores on the scale of EBS, PDS, RDS, IDS and overall DDS between EXP and WLC with large effect size (i.e. F(1,60)= 103.637, p<.000, η2= .633; F(1,60)= 129.088, p<.000, η2= .683; F(1,60)= 118.629, p<.000, η2= .664 F(1,60)= 29.704, p<.000, η2= .531; F(1,60)= 222.710, p<.000, η2= .788 respectively). This indicates CBT reduced 63% emotional burden, 68% physician burden, 66% regimen distress, 53% interpersonal distress and 76 % overall diabetes distress of EXP group. Similarly, EXP group was found significantly different as compared to WLC on PBNA, ADPB, CRNA and overall GMAS with large effect size (i.e. F(1,60) = 56.327, p<.000, η2= .484 F(1,60)= 43.651, p<.000, η2= .421 F(1,60)= 61.649, p<.000, η2= .507 F(1,60)= 67.579, p<.000, η2= .566 respectively). This indicates CBT improved the adherence to treatment such as, 48% patient behavior related non adherence, 42% additional disease and pill burden relate non-adherence, 51% cost related non adherence and 57% over all general medication adherence to treatment. Analysis reveals the significant difference between EXP and WLC after CBT on SATS, IMPS, WORS, and overall DQLS (i.e. F(1,60)= 41.371, p<.000, η2= .408; F(1,60)= 44.126, p<.000, η2= .424; F(1,60)= 40.860, p<.000, η2= .405; F(1,60)= 83.352, p<.000, η2= .581respectively) with effect size. It indicates, CBT session 41% improved satisfaction, 43% impact, 41% non-worried behavior, and 58% overall diabetes patients’ quality of life. Furthermore, findings show that CBT produced significant difference in changing life style as compare to WLC on WALS, MPAS, VPAS and IPAQ (i.e. F(1,60)= 134.7, p<.000, η2= .695; F(1,60)= 17.61, p<.000, η2= .227; F(1,60)= 29.738, p<.000, η2= .331; F(1,60)= 164.245, p<.000, η2= .736 respectively) with effect size. It shows that patients T2DM improved their quality of life after getting CBT session; such as, 69% walking activities, 23% moderate physical activities, 33% vigorous physical activities, and 73% overall daily physical activities (see Table 2).
The analysis reveals that scores significantly decreased throughout treatment among the experimental group on PHQ, HAI, and DDS, and scores enhanced on MAS. CBT played an influential role in addressing depressive symptoms, health-related anxiety, and diabetes distress. On the other hand, CBT played a support role in the treatment adherence (see Fig 2)
Table 3 Range of symptoms severity on Patients Health Questionnaire (PHQ-9) between experimental and waitlist control groups at pre-and post-assessment scores
Groups
|
Severity level
|
.PHQ-9
|
Pre-Scores
|
Post-Scores
|
Experimental Group (EXP=27)
|
n(%)
|
n(%)
|
|
Minimal depressive symptoms
|
-
|
1(3.7%)
|
|
|
Mild depressive symptoms
|
2(7.40%)
|
10(37.1%)
|
|
|
Moderate depressive symptoms
|
11(40.8%)
|
16(59.3%)
|
|
|
Moderate to severe depressive symptoms
|
9(33.4%)
|
-
|
|
|
Severe depressive symptoms
|
5(18.5%)
|
-
|
|
Waitlist Control Group (WLC=35)
|
|
Minimal depressive symptoms
|
-
|
-
|
|
|
Mild depressive symptoms
|
2(5.7%)
|
1(2.9%)
|
|
|
Moderate depressive symptoms
|
9(25.7%)
|
9(25.7%)
|
|
|
Moderate to severe depressive symptoms
|
18(51.4%)
|
17(48.8%)
|
|
|
Severe depressive symptoms
|
6(17.2%)
|
8(22.8%)
|
|
Findings showed a significant improvement in symptoms of depression in the EXP group. The score in the pre-vs. post of the EXP group is as follows, minimal depressive symptoms 0% vs. 3.7%, mild depressive symptoms 7.4% vs.37.1%, moderate depressive symptoms 40.8% vs. 59.3%, which indicates CBT significantly reduced the depressive symptoms. Similarly, at pre-assessment the moderate-severe depressive symptoms 33.4%, and severe depressive symptoms 18.5% reduced due CBT at post-assessment while no significant change is observed in the waitlist control group (see Table 3).
Table 4 Score difference on General Medical Adherence Scale (GMAS) between experimental and waitlist control groups at pre-and post-assessment scores
Groups
|
Levels
|
GMAS
|
Pre-Scores
|
Post-Scores
|
Experimental Group (EXP=27)
|
n(%)
|
n(%)
|
|
High Adherence
|
2(7.41%)
|
10(37.04%)
|
|
Good Adherence
|
3(11.12%)
|
12(44.45%)
|
|
Partial Adherence
|
13(48.15%)
|
3(11.12%)
|
|
Low Adherence
|
6(22.23%)
|
2(7.41%)
|
|
Poor Adherence
|
3(11.12%)
|
-
|
Waitlist-Control Group (WLC=35)
|
|
High Adherence
|
3(8.57%)
|
2(5.71%)
|
|
Good Adherence
|
6(17.14%)
|
5(14.29%)
|
|
Partial Adherence
|
15(42.86%)
|
18(51.43%)
|
|
Low Adherence
|
9(25.71%)
|
6(17.14%)
|
|
Poor Adherence
|
2(5.71%)
|
4(11.43%)
|
Findings show that CBT significantly improved adherence to treatment among individuals with T2DM. Post-testing analysis reveals that the individuals from the EXP group improved treatment adherence considerably. For example, high adherence at baseline 2(7.41%) increased to treatment 10(37.04%), good adherence at baseline 3(11.12%) improved to treatment 12(44.45%), partial adherence 13(46.15%) to treatment 3(11.12%), low adherence at baseline 6(22.23%) to treatment 2(7.41%). In the case of WLC, no significant difference was observed, such as high adherence at baseline 3(8.57%) and post-analysis 2(5.71%), good adherence at baseline 6(17.14%) and post-analysis 5(14.29%), partial adherence at baseline 15(42.86%) and post-analysis 18(51.43%), low adherence at baseline 9(25.71%) and post-analysis 6(17.14%) and poor adherence at baseline 2(5.71%) and post-analysis 4(11.43%) respectively (see Table 4).