Study site and population
Data for this study were obtained from the 2015 Kenya STEPs survey. This was the first national household survey on NCD risk factors. Data was collected in all 47 counties in Kenya between April and June 2015. The aim of the overall study was to provide estimates for indicators on NCD risk factors for persons aged 18–69 years.
Sample size and sampling
A multistage stratified sampling method was used to allow national estimates by sex (male and female) and residence (urban and rural). The survey used the fifth National Sample Surveys and Evaluation Programme (NASSEP V) master sample frame that was developed by the Kenya National Bureau of Statistics (KNBS). The frame was developed using the Enumeration Areas (EAs) generated from the 2009 Kenya Population and Housing Census to form 5360 clusters split into four equal sub-samples. A total of 6000 households were sampled targeting one individual randomly selected from the eligible household members. A total 4754 households gave consent to participate in the study. To produce unbiased estimates, sampling weights were calculated as the inverse or reciprocal of all the selection probabilities at all the stages mentioned above. The weights were derived from the processes involved in the creation of sampling frame (NASSEP V) and selection of individuals in the study. Further, the weights were adjusted to cover individual non-responses. Post stratification adjustments were done to align with the population projections according to age-sex categories.
Data collection
Data were collected by trained personnel using a structured questionnaire adapted from the WHO STEPwise approach to chronic disease risk factor surveillance (STEPS) tool [25] with modifications to suit the Kenyan context. Data was collected in 2 days. On the first day, the questionnaire was administered and anthropometric measurements were also collected. Participants were asked on day one to fast overnight and fasting blood measurements were collected on day two. The questionnaire elicited information on demographic characteristics and health behaviors. The trained field personnel took anthropometric measurements (blood pressure, heart rate, height, weight, waist and hip circumference), and biochemical measurements (fasting blood glucose, and lipid profile (total cholesterol and HDL cholesterol only)).
Data were recorded on Personal Digital Assistants (PDAs) loaded with eSTEPS software. Further details on the data collection and quality assurance procedures are described elsewhere [26].
Measurements and definitions
Diabetes measurements and definitions were based on established guidelines by the WHO [10]. Fasting blood glucose levels were measured using a point of care instrument (CardiocheckPA analyzer®) from PTS Diagnostics. Pre-diabetes was defined as impaired fasting blood glucose (IFG) of 6.1 mmol/l to less than 7 mmol/l while diabetes was defined as fasting blood glucose of 7 mmol/l or more or a self-report of previous diagnosis of diabetes by a health care professional or currently receiving treatment for diabetes. Awareness was defined as prior diagnosis of diabetes by a healthcare professional among all participants. Treatment was defined as receiving pharmacologic treatment to lower blood glucose in the previous 2 weeks among all diabetic participants. Control was defined as having a fasting blood glucose of < 7 mmol/l while on pharmacologic treatment among all diabetic participants.
Fasting lipid levels were also were measured using a point of care instrument (CardiocheckPA analyzer®) from PTS Diagnostics. Abnormal lipid values were defined using National Cholesterol Education Program (NCEP) guidelines [27]. High cholesterol was defined as total cholesterol ≥5.0 mmol/L or are currently on medication for raised cholesterol. Low HDL-cholesterol (HDL-C) was defined as HDL-C < 1 mmol/l for men and < 1.3 mmol/l for women.
Blood pressure was measured using a validated blood pressure machine (OMRON M2 device). Three readings were taken for systolic blood pressure (SBP) and diastolic blood pressure (DBP) in accordance with the WHO hypertension guidelines [28]. The averages of the last two readings were then recorded. Hypertension (High/raised blood pressure) was defined as a systolic blood pressure (SBP) ≥ 140 mmHg, or diastolic blood pressure (DBP) ≥ 90 mmHg, or previous diagnosis of hypertension or being on antihypertensive therapy [28].
Standing height without shoes was measured to the nearest 0.1 cm using a portable height measuring equipment (SECA 877). Weight was also measured to the nearest 0.1 kg with participants wearing light clothing and without shoes using a pre-calibrated digital weighing scale (SECA 877).
Body mass index (BMI) was calculated as weight divided by height squared (kg/m2). Overweight was defined as a BMI ≥ 25–29.9 kg/m2 and obesity was defined as BMI ≥ 30 kg/m2 using the WHO BMI cut-off points [29].
Waist circumference was measured using a constant tension tape across the umbilicus level. Central obesity was defined as waist circumference ≥ 94 cm for men and ≥ 80 cm for women [29, 30].
Physical activity was self-reported using the WHO Global Physical Activity Questionnaire [31] which was added to the main study questionnaire. Insufficient physical activity in this study was defined as self-reports of less than 150 min of moderate intensive activity or less than 75 min of vigorous intensive physical activity per week, including walking, running and cycling.
Harmful use of alcohol was defined as consumption of more than 1 standard drink (which is the amount of alcohol found in a small beer, one glass of wine, or one tot of spirits) per day for females and more than 2 standard drinks for males [32, 33].
Tobacco use was defined as self-reported current use of smoked tobacco or smokeless tobacco products.
High sugar intake was defined as self-reports of far too much or too much consumption of sugar in a day. Bad fat intake was defined as self-reported use of saturated fats e.g. lard, margarine, butter and vegetable fat for cooking. High salt consumption was defined as self-report of far too much or too much consumption of actual salt and in processed foods, adding salt when cooking and/or to cooked food. Insufficient fruit and vegetable intake was defined as self-reported consumption of less than 5 servings/day of fruit and vegetables.
Socio-demographic variables considered for this study included age recoded into four categories (18–29, 30–44, 45–59, and 60–69). At the time of data collection, the current education system in Kenya was the 8–4-4 model which had 8 years of primary education, 4 years of secondary education and 4 years of university. In this study education was categorized into four groups [i) no schooling, ii) primary incomplete – those who did not complete the 8 years of primary school, iii) primary complete – those who completed 8 years of primary school and iv) secondary level and more – those with either secondary education and higher)]. Occupation was categorized in three categories [i) employed – those on salary employment, ii) self-employed – those with businesses including small businesses and iii) unemployed – those not working at all]. Marital status was categorized into two categories [i) in union – includes those married and those not married but living together and ii) not in union – those not married].
Socio-economic status was measured using a household asset and amenities index that assessed household ownership of various assets and amenities commonly used in Demographic and Health Surveys in low and middle income countries [34]. Standardized weight scores were generated using principal components analysis and ranked to generate wealth quintiles which were recoded into five quintiles from the lowest representing poorest households to highest quintile representing the wealthiest households.
Data analysis
The analytical dataset used in this study excluded 15 participants with inconsistent age data. We computed proportions to assess the prevalence, awareness, treatment and control of pre-diabetes and diabetes. The direct method was used to estimate the age standardized prevalence’s’ for pre-diabetes and diabetes using the 2009 Census data to adjust the proportions. We further employed logistic regression models to examine the demographic, behavioral, and body composition factors associated with the prevalence of pre-diabetes and diabetes. All the known CVD risk factors in the dataset were added to the models. Demographic variables included in the model were age, sex, education, marital status, place of residence, and household socio-economic status.
All the analysis were done separately for males and females in addition to an overall model for both. Age-standardized estimates were computed using the 2009 Census data to adjust the proportions. P-values less than 0.05 were considered to be statistically significant. All data analyses were weighted using individual weights to reflect national representativeness. All analyses were done using STATA version 14 (Stata Corporation, College Station, TX).
Ethical approval
The study protocol was reviewed and approved by the Kenya Medical Research Institute’s Ethics Review Committee (SSC No. 2607). All study participants were informed about the study aims including both potential benefits and risks associated with participation. Verbal consent was sought from the household head. All eligible participants gave informed written consent before interview and examination. During the survey, participants who were noted to have abnormalities in their laboratory tests and blood pressure measurements were referred for further care to the nearest health facility or facility of their choice with a referral form. Patient identifiers were delinked from the analytical datasets.