Retrospective cohort study on risk factors for development of gestational diabetes among mothers attending antenatal clinics in Nairobi County

Retrospective cohort study on risk factors for development of gestational diabetes among mothers attending antenatal clinics in Nairobi County

 

Maureen Atieno Adoyo¹, Charles Mbakaya², Venny Nyambati³, Yeri Kombe²

 

¹Institute of Tropical Medicine and Infectious Diseases, Kenya, ²Centre for Public Health Research, Kenya Medical Research Institute, Kenya, ³College of Health Sciences, Department of Biochemistry, Jomo Kenyatta University of Agriculture and Technology, Kenya

 

 

^&Corresponding author
Maureen Atieno Adoyo, Institute of Tropical Medicine and Infectious Diseases, Kenya

 

 

 

 

A WHO report indicates that burden of Non-Communicable Diseases (NCDs) is rising at an alarming rate among low Income Countries (LICs) imposing large, avoidable costs in socio-economic terms among population and communities. Further, diabetes case is projected to rise by more than 50% by the year 2020, if urgent interventions are not put in place to reverse the trend [1]. In Africa, the movement from a rural lifestyle to a more industrial urbanized way of life is mainly responsible for the evolving problem of chronic diseases such as diabetes and other NCDs [2]. Maternal factors including medical condition, weight, age and nutrition during development of fetus may influence future health and disease susceptibility such as adverse metabolic conditions including obesity, diabetes and cardiovascular disease [3] hence the need to have intervention at individual developmental stages among high risk population. Mayo Clinic suggests that any woman is at risk of developing gestational diabetes but some have a higher risk than others, women older than 25 years have increased risks, other factors such as family or personal history is also a determining factor to the development of GDM condition [4]. Further, increased GDM risk is determined by a woman's pre-diabetes condition; a precursor to type 2 diabetes and when a family member of a pregnant woman, such as a parent or sibling, has type two 2 diabetes. A woman who is overweight before pregnancy and had had very large baby in the previous deliveries or had a still birth is also thought to be at risk of developing gestational diabetes [4]. A significant proportion of pregnancies approximately 7% are complicated by GDM related to birth defects and also the leading causes of perinatal mortality across the world [5]. Generally, studies done in Africa related to GDM are few and scanty in information therefore, the need to conduct further research to explore risk factors that lead to GDM specific to developing Countries in Africa.

 

 

A retrospective cohort study was conducted among mothers attending selected maternity facilities in Nairobi County; which comprises of six sub-counties and has a total of 92 maternity facilities put into different categories of: local Authority, Ministries of Health, Non-Governmental and Private enterprise based on institution ownership. An eligibility criterion for study participants was carefully determined by prescribed inclusion criteria before the commencement of the study.

 

Inclusion criteria: mothers attending ANC in selected maternity facilities; mother who are on third trimester of pregnancy; willingness to take Glucose intolerant test; mothers willing to participate in the study by signing the consent form. A proportionate sample size determination formulae [6] was used to derive a definite sample for the two groups (GDM and non-GDM), effect size for each group was within 10% point of true difference, and 95% Confidence. Factoring in the effect size, the estimate of proportion of among mothers with GDM was assumed to be 23%.

 

n= {Z_1-α/2√[2P(1-P)]+Z_1-β√[P₁(1- P₁+P₂(1- P₂)]}²/(P₁-P₂)²

 

Where:

 

n=Minimum required sample size;

 

α = Type I error (0.05);

 

β= Type II error (0.10);

 

At 95% confidence, α_1-α/2 = 1.96;

 

At 80% power, z_1-β= 0.842;

 

P1= Estimated proportion of mothers without GDM (13%);

 

P2= Estimated proportion of mothers mothers with GDM (23%);

 

P1-P2= Effect size (10%);

 

P=(P₁+P₂)/2

 

The minimum required sample size is 231. Allowing for 10% attrition, the sample size will be adjusted upwards to 254 with each group having a sample of 127 respondents. Six maternity facilities were selected using simple random sampling method to ensure all facilities had an equal chance of inclusion as a study site given a population of 97 maternity facilities from Nairobi County Master Facility List (MFL) spread within six sub counties of Nairobi County. Facilities codes were recorded on sheets of papers, folded carefully, mixed in a container and hand picked randomly from the container. Each code was then matched with the MFL and a total of six (6) had a chance of representation in this particular study. All women who attended clinic during the study period and had met inclusion criteria were purposively allowed to be part of study subjects until the suitable sample size was reached. A structured interviewer administered questionnaire was used to collect data and oral glucose intolerance test was done to ascertain GDM status of mothers selected to participate in the study

 

Data management and analysis: filled questionnaires were checked for completeness thereafter the code data was entered into excel sheet which was exported for analysis onto stata version 13. Risk factors were analysed as odds ratio, univariate logistic regression and multilevel logistic regression to account for confounding variable. Background information of respondents was presented as frequency and percentages.

 

 

Results on demographic characteristics of respondents were presented in descriptive statistics. Table 1 indicates that out of 238 respondents 172 (72.27%) had no GDM condition while 66 (27.73%) had GDM condition. Most mothers 112 (47.86%) had BMI above normal (Overweight and Obese), 108 (46.15%) had normal weight and few 14 (5.98%) were underweight.

 

Diabetes history in the family was highly observed at 38 (62.67%) among mothers with GDM condition while no history of diabetic condition in the family was a common observation at 88 (57.89%) among non-GDM mothers while high blood pressure condition was common 23 (38.89%) among GDM mothers compared to 27 (18.75%) among non-GDM mothers (Table 2). The mean age for mothers with GDM was high with a mean of 33.06 (95% C.I: 31.59-34.52) compared to a mean of 27.9 (95% C.I: 27.01-28.78) for non- GDM mothers. Complimentarily, weight before pregnancy was high with mean of 74.04 (95% C.I: 70.82-77.30) among mothers with GDM compared to mean of 60.27 (95% C.I:58.59-61.96) among non-GDM mothers (Table 3).

 

Results for BMI among mothers determined by weight before pregnancy indicates that mothers with GDM had above normal BMI (overweight and obese) at 54.55% and 24.24% respectively compared to non-GDM with 35.12% and 0.6% in the category of overweight and obese respectively (Figure 1). To further establish the relationship between gestational diabetes and individual biological characteristics, logistic regression analysis was done with results indicating that Mothers who had diabetic history in the family had twice the risk of developing GDM (OR= 2.27; 95% C.I: 1.23-4.17) compared to those who did not observe diabetic history in the family .Obese BMI status had positive association with GDM, findings indicates that obese mothers had two times increased risk of developing GDM condition (OR= 2.08; 95% C.I: 11.83-3656.7). Age and weight of the mother before pregnancy similarly had positive association with GDM (OR = 1.15; 95 % C.I:1.089-1.206) and (OR =1.10; 95 % C.I:1.07-1.13) respectively (Table 4).

 

After adjusting for age confounder ,GDM risk remains significant in age and weight (OR = 1.098; 95% C.I: 1.089 -1.206) and (OR = 1.05; 95% C.I:1.003 -1.093) respectively, although slightly reduced in magnitude where age has 9.8 % risk and weight gave 5% increased risk to the development of GDM ,while diabetic history and BMI were not significant Table 5.

 

 

The analyses in this study suggest that individual biological factors such age is independently associated with GDM, including weight of the mother before pregnancy and history of diabetes among family members. Further, the study suggests that GDM is common among women who are advanced in age, mothers with GDM had a mean age of 33.06 years while non-GDM mothers had a mean age of 27.9 years. This findings are consistent with evidence that any women above age of 25 years has increase risk of GDM. High BMI which corresponds with high weight was associated with GDM condition. This study indicates that nearly more than half (47.86%) of the study population is considered either overweight or obese. According to Macaulay et. al (2014), diabetes and other NCDs such high blood pressure among others are largely associated with sedentary life style as populations shift from rural way of life to urban life lifestyle to a more industrial urbanized way of life and hence obese condition among population. Obesity condition has serious health consequences related to NCDs including diabetes and complications during pregnancy and subsequent generations. Mothers who had a close family members sibling or parent with diabetes are likely to develop gestational diabetes this is evident by the fact that mothers with GDM condition reported 62.3% observation in diabetes in family tree while mothers with non GDM condition reported low 42.11% observation in family members who had diabetes.

 

 

Gestational diabetes cases is relatively high in the population, lifestyle including other individual biological characteristics such as advanced age in pregnancy, high weight and diabetic history in family are determining factors for development of diabetes among pregnant women.

 

 

Authors declare no competing interests.

 

 

Author (corresponding author) and 2, 3, 4 supervisors with equal contribution. All authors read and approved the final version of the manuscript.

 

 

I wish to express my sincere gratitude to my supervisors Dr. Charles Mbakaya, Dr. Venny Nyambati and Dr. Yeri Kombe without whom this work would not have been accomplished. I wholeheartedly appreciate their constant guidance, valuable advice and patience throughout the process. I also wish to acknowledge all the Management and staff of Pumwanity maternity; Amref's clinics at Kibera and St. Mary's Langata for the support and coordination. Finally, I appreciate statisticians at Kemri Mr. Moses Mwangi and Mr Lawrence Muthami for the guidance in the design and analysis of the study.

 

 

Table 1: descriptive statistics on background information

Table 2: distribution of independent variables between GDM mothers and Non-GDM mothers

Table 3: mean distribution of age and weight before pregnancy between mothers GDM and non-GDM condition

Table 4: univariate logistic regression model for the analysis of factors associated with development of gestational diabetes among mothers attending ANC in Nairobi County

Table 5: multivariate logistic regression model for the analysis of factors associated with development of GDM among mothers attending ANC

Figure 1: distribution of BMI among GDM and non-GDM mothers

 

 

1. WHO and IDF Diabetes action now: An initiative of the World Health Organization and the International Diabetes Federation. Switzerland: WHO. 2004. Google Scholar


2. Shelley Macaulay et al. Gestational Diabetes Mellitus in Africa: a systematic review. PLOS ONE. 2014 June; 9(6): e97871. PubMed | Google Scholar


3. Kellie Tomoshiro. Developmental and environmental influences on physiology and behavior. Elsevier (2014). Google Scholar


4. http://www.mayoclinic.org/diseases-conditions/diabetes/basics/risk-factors/con-2003309. Accessed on October 03, 2015. Google Scholar


5. Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening. BMJ. 2002; 325 (7356):157-160. PubMed | Google Scholar


6. Stanley Lemeshow, David Hosmer Jr W, Janelle Klar, Stephen Lwanga K. Adequacy of sample size in health studies. WHO 2009. Google Scholar