Table 1. Prevalence of Prediabetes
Status
|
n (%)
|
Prediabetes
|
34 (9.19)
|
Normal
|
336 (90.81)
|
Total
|
370 (100)
|
Table 1 shows the prevalence of prediabetes.The prevalence of Prediabetes was found to be 9.19% 95% CI (6.23-12.14). The mean fasting plasma glucose(FPG) for the participants was 4.5 mmol/l, Range (2.6-12.3) SD =1.02
Table 2. Socio-demographic characteristics of the respondents.
Characteristic
|
Total n (%)
N=370
|
No prediabetes n (%)
N= 336
|
Prediabetes n (%)
N=34
|
P-value
|
Age (mean, SD)
|
42, SD=14.3
|
Age groups
|
|
|
|
0.012**
|
18-30
|
93 (25.1)
|
91 (27.08)
|
2 (5.88)
|
|
31-45
|
123 (33.2)
|
113 (33.63)
|
10 (29.41)
|
|
46-60
|
111 (30.0)
|
94 (27.98)
|
17 (50)
|
|
61-70
|
43 (11.6)
|
38 (11.31)
|
5 (14.71)
|
|
Gender
|
|
|
|
0.129
|
Male
|
119 (32.16)
|
112 (33.33)
|
7 (20.59)
|
|
Female
|
251 (67.84)
|
224 (66.67)
|
27 (79.41)
|
|
Marital status
|
|
|
|
0.767
|
Married/Cohabiting
|
285 (77.0)
|
259 (77.08 )
|
26 (76.47)
|
|
Separated/Divorced
|
23 (6.22)
|
21 (6.25)
|
2 (5.88)
|
|
Widowed
|
38 (10.27)
|
33 (9.82)
|
5 (14.71)
|
|
Never Married
|
24 (6.29)
|
23 (6.85)
|
1 (2.94)
|
|
Level of education
|
|
|
|
0.018**
|
No formal education
|
78 (21.1)
|
69 (20.5)
|
9 (26.47)
|
|
Primary level completed
|
207 (56.0)
|
191 (56.84)
|
16 (4.76)
|
|
Secondary level completed
|
57 (15.4)
|
54 (16.07)
|
3 (0.89)
|
|
College /university completed
|
17 (4.6)
|
15 (4.46)
|
2 (5.88)
|
|
Refused
|
11 (3.0)
|
7 (2.08)
|
4 (11.76)
|
|
Monthly income (UGX)
|
|
|
|
0.069
|
<50,000
|
218 (58.92)
|
198 (58.93)
|
20 (58.82)
|
|
50,000-100,000
|
113 (30.54)
|
99 (29.46)
|
14 (41.18)
|
|
100,000-500,000
|
39 (10.54)
|
39 (11.61)
|
0
|
|
Employment
|
|
|
|
0.378
|
Self Employed
|
346 (93.51)
|
313 (93.15)
|
23 (67.65)
|
|
No formal employment
|
24 (6.49)
|
23 (6.85)
|
1 (2.94)
|
|
Family history of diabetes
|
|
|
|
0.084
|
Yes
|
86 (23.3)
|
81 (24.11)
|
5 (14.7)
|
|
No
|
259 (70)
|
230 (68.45)
|
29 (85.29)
|
|
Don’t know
|
25 (6.8)
|
25 (7.44)
|
0
|
|
** statistically significant at p< 0.05.
Table 2 shows the social demographic characteristics of the participants. A total of 378 eligible participants were contacted, out of which 370(97.8%) were enrolled and participated in the study. A total of 5 participants missed in the early morning of day two for testing for blood sugar and 3 failed to achieve fasting status in the morning. Of the 370 participants 119 (32.2% were males and 251(67.84%) were females. The overall mean age was 42 (SD=14.3), and the mean age for males was 41 (SD=13.7) while for females was 42 (SD=14.5). More than half of the participants had primary level education, Family history of diabetes was reported in only 23.4% of the participants. Among the social demographic factors, Age and level of Education were found to have a significant association with Prediabetes(p<0.05) at the chi square test analysis. Monthly income and family history of diabetes were only marginally associated with Prediabetes. (p<0.1).
Table 3. Bivariate test results showing the association between Prediabetes across the social demographic variables
Characteristic
|
|
Prediabetes n (%)
|
COR
|
95% CI
|
P-value
|
No Prediabetes
N=336
|
Has Prediabetes
N=34
|
Age
|
18-30
|
91 (27.08)
|
2 (5.88)
|
1.000*
|
|
|
|
31-45
|
113 (33.63)
|
10 (29.41)
|
4.027
|
0.859 -18.879
|
0.077
|
|
46-60
|
94 (27.98)
|
17 (50.00)
|
8.229
|
1.845 -36.704
|
0.006**
|
|
61-70
|
38 (11.31)
|
5 (14.71)
|
5.987
|
1.11 - 32.291
|
0.037**
|
Education level
|
No formal education
|
69 (20.97)
|
9 (30.00)
|
1.000*
|
|
|
|
Primary level
|
191 (58.05)
|
16 (53.33)
|
0.642
|
0.271-1.52
|
0.314
|
|
Secondary level
|
54 (16.41)
|
3 (10.00)
|
0.426
|
0.11-1.65
|
0.217
|
|
college /university
|
15 (4.56)
|
2 (6.67)
|
1.022
|
0.2-5.221
|
0.979
|
|
Refused
|
11 respondents declined to mention their level of education.
|
Household income
|
<50,000
|
198 (58.93)
|
20 (58.82)
|
1.000*
|
|
|
|
50000-100000
|
99 (29.46)
|
14 (41.18)
|
1.4
|
0.678 -2.888
|
0.363
|
|
100000-500000
|
39 (11.61)
|
0
|
1
|
|
|
Family history of diabetes
|
No
|
230 ( 73.95)
|
29 (85.29)
|
1.000*
|
|
|
|
Yes
|
81 ( 26.05)
|
5 (14.71)
|
2.042
|
0.764-5.454
|
0.154
|
** statistically significant at p< 0.05.
1.000*- odds ratio of the referent category.
Table 3 shows the results of the bivariate test results between prediabetes and social demographic characteristics. Advancing age was significantly associated with prediabetes. The results show that the odds of prediabetes are 8 times higher with increasing age (COR=8.229 CI 1.845 - 36.704 p=0.006). Age was therefore included in the multivariate analysis, other social demographic factors, that is, education level, household income and family history of diabetes did not remain significantly associated with prediabetes and were therefore not included in the multivariate analysis model.
Table 4. Bivariate test results showing the association between Prediabetes and lifestyle variables
Characteristic
|
|
Prediabetes
|
COR
|
95 % CI
|
P value
|
No prediabetes
N=336
|
Prediabetes
N=34
|
|
|
Smoking
|
No
|
288 (85.71)
|
25 (73.53)
|
1.000*
|
|
|
|
Yes
|
48 (14.29)
|
9 (26.47)
|
2.16
|
0.950- 4.908
|
0.066
|
Alcohol use
|
No
|
186 (55.36)
|
19 (55.88)
|
1.000*
|
|
|
|
Yes
|
150 (44.64)
|
15 (44.12)
|
0.978
|
0.481- 1.991
|
0.953
|
Body mass Index
|
Normal
|
217 (64.58)
|
16 (47.06)
|
1.000
|
|
|
|
Overweight
|
83 (24.70)
|
8 (23.53)
|
1.307
|
0.539-3.169
|
0.553
|
|
Obese
|
36 (10.71)
|
10 (29.41)
|
3.767
|
1.585-8.949
|
0.003**
|
Hypertension
|
Normal
|
229 (68.15)
|
17 (50.00)
|
1.000*
|
|
|
|
High
|
107 (31.85)
|
17 (50.00)
|
2.140
|
1.051-4.354
|
0.036**
|
Moderate intensity work.
|
No
|
225 (66.96)
|
15 (44.12)
|
1.000*
|
|
|
|
Yes
|
111 (33.04)
|
19 (55.88)
|
2.567
|
1.257-5.244
|
0.010**
|
Dietary Patterns
|
|
|
|
|
|
|
|
Low
|
105 (31.25)
|
4 (11.76)
|
1.000*
|
|
|
Protective diet
|
Medium
|
177 (52.68)
|
18 (52.94)
|
2.669
|
0.879-8.100
|
0.083
|
|
High
|
54 (16.07)
|
12 (35.29)
|
5.833
|
1.795-18.950
|
0.003**
|
** statistically significant at p< 0.05
1.000*- odds ratio of the referent category.
From Table 4,the results show that increasing body mass index, hypertension, moderate intensity level of work and diet categorized as protective were significantly associated with Prediabetes. Participants with higher BMI were 3 times likely to have Prediabetes (COR=3.767 CI=1.585-8.949 p=0.003). The odds of Prediabetes were two fold among those who were hypertensive (COR=2.140, CI=1.051-4.354 p=0.036). Participants with moderate intensity level of work were twice likely to have Prediabetes (COR=2.567, CI=1.257-5.244 p=0.010). The Odds of Prediabetes were 5 times higher among participants who consumed a high level of food categorized as protective (COR=5.833 CI=1.795-18.950 p=0.003). Smoking remained marginally associated with Prediabetes and alcohol use continued to have no significant association with Prediabetes they were not included in the multivariate analysis model. Increasing BMI, Hypertension, moderate level of work and protective diet were included in the multivariate analysis.
Table 5. Bivariate test results showing the association between Prediabetes and Medical and Health accesses variables
Characteristic
|
|
Prediabetes, n (%)
|
COR
|
95% CI
|
P Value
|
No Prediabetes
N=336
|
Prediabetes
N=34
|
|
|
Ever tested for diabetes
|
|
|
|
|
|
|
|
No
|
305 (91.87)
|
27 (8.13)
|
1.000
|
|
|
|
Yes
|
31 (81.58)
|
7 (18.42)
|
2.550
|
1.027-6.335
|
0.044**
|
Reported Available testing services at health facility
|
|
|
|
|
|
|
|
No
|
251 (91.27 )
|
24 ( 8.73)
|
1.000
|
|
|
|
Yes
|
59 (85.51)
|
10 (14.49)
|
1.772
|
0.804-3.907
|
0.156
|
|
Don’t know
|
26 (100)
|
0
|
1.000
|
|
|
Steroid use
|
No
|
190 (56.55)
|
13 (38.24)
|
1.000*
|
|
|
|
Yes
|
146 (43.45)
|
21 (61.76)
|
2.102
|
1.018-4.338
|
0.044**
|
** statistically significant at p< 0.05.
In table 5, the results show that; ever testing for Diabetes and steroid use were significantly associated with Prediabetes. The results show that the odds of Prediabetes were twice as high among those who had ever tested for Diabetes (COR=2.550, CI=1.027-6.335 p=0.044). The participants who used steroids were twice as likely to have Prediabetes (COR=2.102, CI=1.018-4.338 p=0.044). Ever testing for diabetes and history of steroid use were therefore included in the multivariate analysis model. Report of availability of testing services at the health facility was not significantly associated with Prediabetes and was thus not included in the multivariate analysis model.
Table 6. Multivariate analysis results of factors inherently associated with Prediabetes
Characteristic
|
|
COR
|
P value
|
AOR
|
P value
|
95% CI
|
Age
|
18-30
|
1.000
|
|
|
|
|
|
31-45
|
4.027
|
0.077
|
3.540
|
0.115
|
0.736-17.03
|
|
46-60
|
8.229
|
0.006
|
7.442
|
0.01
|
1.614-34.294
|
|
61-70
|
5.987
|
0.037
|
5.771
|
0.046**
|
1.031-32.306
|
Moderate intensity work
|
No
|
1.000
|
|
|
|
|
|
Yes
|
2.567
|
0.010
|
2.637
|
0.012**
|
1.233-5.636
|
Protective diet
|
Low
|
1.000
|
|
|
|
|
|
Medium
|
2.669
|
0.083
|
2.421
|
0.13
|
0.771-7.6
|
|
High
|
5.833
|
0.003
|
5.657
|
0.005**
|
1.679-19.054
|
BMI
|
Normal
|
1.000
|
|
|
|
|
|
overweight
|
1.307
|
0.553
|
1.049
|
0.918
|
0.416-2.644
|
|
Obese
|
3.767
|
0.003
|
3.608
|
0.007**
|
1.414-9.206
|
** statistically significant at p< 0.05.
Table 6 shows the results from the multivariate analysis. From the multivariate analysis of significant factors identified at bivariate analysis, increasing age, moderate intensity level of work, consumption of a high level of diet categorized as protective and BMI remained statistically significant with P values <0.05. We adjusted for confounding factors based on biological relevance and significant associations found at the bivariate analysis. The final model was adjusted for family history of Diabetes, smoking, hypertension, ever testing for Diabetes and the use of steroids.
The multivariate analysis showed that the Odds of Prediabetes were 5.7 higher with increasing age (AOR 5.771, p=0.046, CI=1.031-32.306). And the Odds of Prediabetes remained twice as much for participants who had moderate intensity work (AOR=2.637, 0.012, CI=1.233-5.636). Participants who consumed a high level of diet categorized as protective were 5 times likely to have Prediabetes (AOR=5.657, p=0.005, CI=1.679-19.054). The Odds of Prediabetes were 3.6 times higher among the participants with a higher BMI (AOR=3.608, p=0.007, CI=1.414-9.206), see Table 6.