Glycemic control, associated factors, acute complications of Type 1 Diabetes Mellitus in children, adolescents and young adults in Tanzania

Abstract Objective To determine the factors associated with poor glycemic control in children (1‐10 years), adolescents (11‐18 years) and young adults (19‐40 years) with Type 1 Diabetes Mellitus (T1DM) in Kilimanjaro Christian Medical Center (KCMC) in Moshi, Mount Meru Regional Referral Hospital (MMRRH) and Meru District Hospital (MDH) in Arusha, Tanzania. Methods Cross sectional study of 150 participants conducted from January to June 2019, data was collected by structured questionnaire and analyzed using SPSS version 23. Results The mean HbA1c was 12.3 ± 2.2%, 146 had poor glycemic control (HbA1c > 7.5%). BMI, insulin regime and caretaker education were associated with poor glycemic control. There were 16 participants diagnosed in DKA and the most frequently reported complications in the prior 3 months were hyperglycemia (n = 25), DKA (n = 18) and hypoglycemia (n = 4). Conclusions Glycemic control is still very poor particularly in adolescents. Significant associations with glycemic control were higher BMI, insulin regime and guardian education. The study revealed lower prevalence of DKA at diagnosis compared to previous studies.

macrovascular complications (CVA, coronary arterial disease and peripheral vascular disease). However strict glycemic control can prevent these same complications though there may be a risk of hypoglycemia. 5 Few studies have been done looking at the cause of poor glycemic control in Sub-Saharan African countries, hence knowing the causes of poor glycemic control would assist us to improve diabetes care by addressing and targeting these underlying causes.

| ME THODS
This cross sectional study conducted from January to June 2019 looked at the factors associated with poor glycemic control.

| Subjects
T1DM study participants who totaled 432 and were registered at the following study sites: KCMC with 231, MMRRH 173 and MDH with 45.

| Selection of participants
Convenience sampling and a total of 274 participants were excluded either because they were at boarding school, lost to follow-up or the caretaker was unavailable to give consent to study participation. Therefore a total of 158 subjects were interviewed and among these 8 did not meet the study criteria and were excluded from the analysis. The remaining 150 participants were recruited by convenience sampling and were followed up over the six month period ( Figure 1).

| Inclusion criteria
The research subjects were all attending T1DM clinics in KCMC, MMRRH and MDH.

| Exclusion criteria
Those who did not consent to take part in the study and were aged less than 1 year.

| Measurements
Sample size was calculated using the following formula n = (Z 2 p (1p))/ε 2 , this gives an estimated minimum sample size of 158.
The dependent variable was glycemic control and numerous independent variables were assessed including sociodemographic, clinical and diabetes related variables.
Data was collected during a short oral interview and a questionnaire consisting of 38 questions which had been translated into the Kiswahili language and were filled out by the participant and caretaker if participant was a child or adolescent or only by the participant if he or she was a young adult. 10 questions were assigned to sociodemographic details, 30 questions were clinical or diabetes related, 1 question each was assigned to diet and exercise and 3 questions concerned psychological factors in relation to T1DM.
The participants were then taken to a separate room where they received a full explanation of the research and the aims. Filling of the questionnaire was assisted by the principal investigator and two research assistants with an interview duration of approximately 30 minutes. HbA1c levels were measured and recorded to assess the average glycemic control in the previous 3 months.

| Data analysis
Data were analyzed using SPSS version 23, numerical variables were summarized using measures of central tendency (mean + SD, median + IQR). Categorical variables were summarized using percentages and frequencies, multivariate models (logistic regression and multiple logistic regressions) were used to check for factors associated with glycemic control and a P-value < .05 was considered statistically significant. Table 1 shows the socio-demographic characteristics of the study participants. Of the 150 participants enrolled in the study, the mean age was 16 years with standard deviation of 5.9 years, 74 of the participants were male and 79 had at least achieved a secondary education. A high proportion (n = 88) of the participants were from MMRRH followed by 55 from KCMC and 7 were from MDH.

| Socio-demographic characteristics of the study participants
The average time to reach each facility was 2 hours by walking with inter-quartile range 1 to 2 hours and the primary care giver most frequently reported was the mother that accounted for 80 of the participants, followed by fathers (n = 63) of the participants and only 7 reported having another primary caretaker. The majority of the caretakers (n = 91) had achieved a primary education and most (n = 129) were unemployed.  (Table 2).

| Proportion of children with poor glycaemic control
Most of the participants (n = 146) had poor glycaemic control (HbA1c > 7.5%) and the overall average HbA1c was 12.3% with standard deviation of 2.2% which indicates that a large group of the participants had poor glycaemic control.

| Socio-demographic factors associated with poor glycemic control
Children aged below 10 years had good glycemic control when compared to adolescents and young adults which was statistically significant P = .007 in univariate analysis. Adolescents were more likely to have poor glycemic control when compared to other groups, (HbA1c 12.8%), which was even higher than the overall mean HbA1c.

| Clinical and diabetes specific factors associated with poor glycemic control
Overweight participants had significantly better glycemic control when compared to the other BMI groups, and this was statistically significant P = .025. Insulin regime was associated with glycemic control and the result was statistically significant, those who had actrapid and insulatard had better glycemic control (HbA1c, 12.3 ± 2.2%) when compared to those who had other insulin regimens (mixed or only soluble), (HbA1c 14%); (see Table 4).
In the final model, socio-demographic, clinical and diabetes mellitus related factors with p value <0.1 in the analysis of variance were run using multivariate regression analysis to control for possible confounders and modifiable effects and to study their significance risk on glycemic control. The result indicated that higher BMI, type of insulin (actrapid and insulatard) and guardian education were significantly associated with better glycemic control, P < .05.

| Acute complications among T1DM children, adolescents and young adults
There were 16 participants who presented with DKA at diagnosis of T1DM, 47 T1DM participants reported to have acute complications related to T1DM in the last 3 months, the most frequently reported complications in the preceding three months were hyperglycemia (n = 25) followed by DKA (n = 18) and hypoglycemia (n = 4), ( Figure 2). Hyperglycemia and hypoglycemia were more prevalent in adolescents and young adults, whilst DKA was more commonly reported in children ( Figure 3).

| D ISCUSS I ON
The key finding from this study was that 146 study participants had poor glycemic control (HbA1c > 7.5%) with a mean HbA1c was Note: N = absolute number = 150, P value < .05 considered statistically significant.

| CON CLUS I ON S AND RECOMMENDATION
Glycemic control in children, adolescents and young adults with T1DM attending these clinics is still very poor and the factors associated with glycemic control from this study were BMI, insulin type and guardian education achievement.

CO N FLI C T O F I NTE R E S T
The author declares no financial relationships or conflicts of interest in relation to this article.

AUTH O R CO NTR I B UTI O N S
RPM designed the study, performed data entry, contributed to discussion, wrote the manuscript and reviewed and edited the manuscript. JAA reviewed the data, assisted with data cleaning and analysis, reviewed the manuscript. GGC assisted with data collection, assisted in calling participants to remedy missing data, assisted with measuring participant HbA1c levels. Professor LJM, D.M and Professor P.C reviewed and edited the manuscript.

E TH I C A L A PPROVA L
Ethical approval and consent to participate: consent for the study was obtained from the Kilimanjaro Christian Medical University College ethics committee with certificate number 2100. Consent for study participation was also obtained from the participants and caretakers.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.