Introduction
For decades, infectious and communicable diseases, such as HIV, tuberculosis and malaria, have plagued sub-Saharan Africa (SSA) and are noted to contribute significantly to disease burden in this region.1 However, like many countries across the world, SSA currently faces a double burden of disease, primarily arising from what is known as epidemiological transition.2 That is a situation where non-communicable diseases (NCDs) outstrip the contribution of communicable diseases to disease burden. The WHO reports that nearly 41 million people die each year from NCDs with 77% of these mortalities occurring in low-income and middle-income countries.3 The same report also indicates that the leading NCD was cardiovascular diseases (CVDs), followed by cancers, chronic respiratory diseases and diabetes.3
Available evidence indicates that disability-adjusted life-years (DALYs) attributable to NCDs increased in SSA from 90.6 million in 1990 to 151 million in 2017, reflecting a 67% increase.2 In the case of Kenya, nearly 37% of DALYs lost in 2018 were attributable to NCDs.4 Thus, highlighting the significant threat that NCDs pose to the well-being and quality of life of Kenyans. Therefore, it is imperative to invest in policies and programmes that aim at reducing the incidence of NCDs among the Kenyan population.
The significance of profiling the prevalence of different NCDs in Kenya is premised on the point that this ill health condition is associated with several adverse effects. The escalating burden of NCDs places significant financial strain on Kenyan families, businesses and the government. Moreover, NCDs diminish economic productivity by curtailing life spans and causing illnesses during individuals’ most productive working years.5 Mensah et al further state that NCDs accounted for a 3.4% loss in Kenya’s gross domestic product in 2016 and that addressing NCDs would avert nearly 110 000 incremental deaths in Kenya.5
Addressing the public health challenge of NCDs requires a better appreciation of the magnitude of the problem and the associated risk factors. Indeed, some studies have estimated the burden of NCDs in Kenya.5–8 However, these studies were limited in scope. For instance, Mtintsilana et al’s study7 focused only on the association between socioeconomic status and NCD risk among young adults. Mensah et al
5 addressed the economic impact of NCDs on the Kenyan population while Sureshkumar et al’s study6 only explored stakeholder perceptions about NCDs. Available evidence suggests that women of reproductive age are prone to experiencing the simultaneous presence of behavioural and metabolic risk factors, leading to an elevated risk of NCDs.9 10 In contrast to men, women of reproductive age have been documented to often exhibit fewer symptoms and less noticeable signs of specific NCDs, such as CVDs.10 11 Consequently, they are less likely to be promptly identified, treated or targeted for NCD prevention. Yet, the existing literature on NCDs in Kenya5–8 has not prioritised this subpopulation. Hence, it is unclear the extent to which NCDs and their associated factors among women of reproductive age in Kenya. This presents a significant knowledge gap that must be filled. We address this knowledge gap by examining the burden of NCDs among women of reproductive age in Kenya by highlighting the prevalence and risk factors. using the 2022 Kenya Demographic and Health Survey (KDHS).
Methods
Data source
The data for this study were obtained from the 2022 KDHS, a nationally representative household survey designed to gather comprehensive data on population, health and nutrition indicators. The survey was conducted by the Kenya National Bureau of Statistics in collaboration with the Ministry of Health while technical assistance was provided by ICF Macro through the DHS programme. The 2022 KDHS incorporates a two-stage sampling design. In the first stage, 1692 clusters were randomly selected from the Kenya Household Master Sample Frame using equal probability with independent selection in each sampling stratum.12 Subsequently, household listing was carried out in all selected clusters, creating a sampling frame for the second stage. In the second stage, 25 households were chosen from each cluster, resulting in a total of 42 022 households being included in the survey sample.12 Data collection was performed using structured household and women’s questionnaires administered by trained enumerators. The women’s questionnaire covered various aspects, including birth history, childhood mortality, fertility preferences, NCDs, antenatal and postnatal care, place of delivery, childhood diseases, and childhood vaccinations. The analysis consisted of a weighted sample of 15 891 women between the ages 15 and 49 years old having information on the nine NCDs considered in this study.
Study variables and measurements
Outcome variable
The outcome variable for the study was NCDs. Specifically, the study looked at nine NCDs: hypertension, diabetes, heart disease, lung disease, arthritis, depression, anxiety, breast cancer and cervical cancer. To gather this information, the respondents were asked whether a doctor or health worker had informed them of having any of these diseases, and they could respond with a ‘yes’ or ‘no.’ All of these NCDs were self-reported.
Explanatory variables
Multiple factors informed by theoretical and empirical literature8–10 relating to NCDs were included in the analyses as the explanatory variables. These variables include age (15–19, 20–24, 25–29, 30–34, 35–39, 40–44 and 45–49), residence (rural and urban), level of education (no education, basic, secondary and above), wealth index (poorest, poorer, middle, richer, richest), marital status (never married, married, formerly married), occupation (not working, professional/clerical, sale/services, agriculture worker, domestic worker, unskilled/skilled), type of cooking fuel (clean fuel(liquefied petroleum gas, biogas, solar power, electricity, natural gas), unclean fuel (charcoal, wood agricultural grass etc), alcohol consumption (had never consumed alcohol, consume alcohol), religion (Christianity, Islam, other) and body mass index (BMI) (underweight, normal-weight, overweight, obese).
Data analysis
Statistical analysis was performed by using Stata V.14 to examine the prevalence of different types of NCDs among women of reproductive age in Kenya. A bar graph was used to visually represent this prevalence. The characteristics of the respondents were described based on the type of NCD, and a χ2 test was employed to determine statistically significant associations between the outcomes and explanatory variables. Further, univariate and multivariate analyses were conducted to explore the relationships between potential explanatory variables and the study outcomes (NCDs). Logistic regression models were fitted using Stata regression commands to assess adjusted risk factors associated with the study outcomes, with ORs and 95% CIs calculated. To account for any sampling bias from under or oversampling of respondents in the total population, all descriptive estimates were weighted using the individual weight variable (v005) in the dataset. Before performing the multivariable regression, the possibility of multicollinearity was examined using the variance inflation factor, which showed a mean score of 6.15, indicating no significant multicollinearity. To consider the complex survey design of the DHS data, the ‘svyset’ command in Stata was employed. To assess the fit and relative performance of the regression model, a useful metric such as Pseudo R2 was used. We followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for reporting the data (see online supplemental file 1).
Supplemental material
Patient and public involvement
Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Results
Prevalence of NCDs among women of reproductive age in Kenya
Figure 1 shows the prevalence of NCDs among women of reproductive age in Kenya. Overall, 15.9% of Kenyan women aged 15–49 years were living with at least one of the nine NCDs considered in this study. The most prevalent NCD among this cohort was hypertension (8.7%), followed by arthritis (2.9%) and depression (2.8%). Breast and cervical cancer were the least reported NCDs, 0.2%, respectively.
A bar graph showing the prevalence of non-communicable diseases (NCDs) in Kenya.
Prevalence of simultaneous/co-occurrences of NCDs
Table 1 presents the prevalence of individuals experiencing multiple/co-occurrences of NCDs among the sampled population. The majority (12.66%) was identified with a single NCD while a smaller proportion (2.52%) simultaneously deals with two NCDs (Dyad). Similarly, individuals with three NCDs constitute a smaller percentage (0.65%), and a limited number of individuals (0.09%) cope with four or more NCDs concurrently. These findings emphasise the varied landscape of NCD co-occurrences, ranging from singular conditions to complex combinations of four or more diseases.
Prevalence of simultaneous/co-occurrences of NCDs
Proportional distribution of NCDs across explanatory variables
The proportion of women of reproductive age living with at least one NCD was higher among those aged 45–49 years (33.10%), urban-dwelling women (16.92%), those with basic education (17.3%), those in the richest wealth index (19.29%), formerly married women (23.48%) and among those working in professional/clerical occupations (20.51%). Statistically significant differences were observed in terms of type of cooking fuel with a higher proportion of NCDs being reported among those using clean cooking fuels (17.88%). Also, higher proportion of living with at least one NCD was observed in those who consumed alcohol (22.72%), those who professed Christianity (16.28%) and among obese women (28.27%) (see table 2).
Proportional distribution of NCDs by explanatory variables
Factors associated with the risks of NCDs
Bivariate results
Table 3 displays the results of the bivariate logistic regression analyses investigating the factors associated with NCDs. The analysis includes crude ORs (COR) at 95% CIs for each explanatory variable. The model encompasses various types of NCDs considered in the study, and the final model specifically focuses on respondents with at least one NCD.
Bivariate regression on factors associated with non-communicable diseases (NCDs) in Kenya
Our findings reveal significant associations between age, education, wealth index, marital status, occupation, alcohol consumption, religion and BMI with most of the NCDs considered in our analysis. For instance, older women aged 44–49 had higher odds of experiencing hypertension (COR 17.92, 95% CI 12.57 to 25.55), diabetes (COR 12.49, 95% CI 5.53 to 28.19), lung disease (COR 3.24, 95% CI 1.94 to 5.43), depression (COR 3.43, 95% CI 2.25 to 5.24), anxiety (COR 1.95, 95% CI 1.17 to 3.25), arthritis (COR 13.02, 95% CI 8.42 to 20.13) and living with at least one NCD (COR 9.13, 95% CI 7.46 to 11.18) compared with young people aged 15–19 years. However, there was no significant association found between age and heart disease.
The results indicate that women residing in rural areas had a lower likelihood of having hypertension, depression, anxiety and at least one NCD compared with their urban counterparts. Women with secondary or higher levels of education were 2.12 and 1.41 times more likely to suffer from hypertension or live with at least one NCD, respectively. When compared with women with the poorest wealth index, those with the highest wealth index were 3.04, 3.03, 1.47, 2.24, 1.76, 10.23 and 2.18 times more likely to have hypertension, diabetes, depression, anxiety, arthritis, cervical cancer and live with at least one NCD, respectively.
Regarding marital status, married and formerly married women were at a higher risk of most NCDs. For instance, married and formerly married women were 2.70 and 3.68 times more likely to live with at least one NCD than those who were never married. Except for heart disease, breast and cervical cancer, women employed in various sectors faced a higher risk of hypertension, diabetes, lung disease and other NCDs compared with those who were unemployed. Women who consume alcohol have an elevated risk of hypertension, lung disease, depression and anxiety compared with those who never consumed alcohol. Additionally, women who professed Islam had lower odds of experiencing hypertension, lung disease, depression and arthritis than their Christian counterparts. Lastly, compared with women with normal body weight, overweight and obese women faced a higher risk of most NCDs, such as hypertension, diabetes, depression, anxiety and arthritis.
Multivariate results
Table 4 presents the multivariate results of the factors associated with the likelihood of living with at least one of the NCDs including hypertension, diabetes, heart disease, lung disease, depression, anxiety and arthritis. After accounting for all other explanatory factors important in predicting the outcomes, age, wealth index, marital status, occupation, alcohol consumption and BMI were significantly associated with most NCDs. The results are presented as adjusted ORs (aOR) with 95% CI. As previously shown under the bivariate results, similarly, we found that the odds of living with NCDs generally increase with age, with older women aged 44–49 years having the higher odds of 4.66 times more likely to live with at least one of the NCDs. The analyses indicate that, as the wealth index of women increases, their risk to living with NCDs also increase. The richest group has higher odds of living with at least one of the NCDs (aOR 1.60, 95% CI 1.26 to 2.04) compared with the poorest group.
Regarding marital status, individuals who were married (aOR 1.33, 95% CI 1.13 to 1.56) or formerly married (aOR 1.48, 95% CI 1.22 to 1.80) had higher odds of living with NCDs compared with those who have never married. Relatedly, individuals working in professional/clerical, sale/service, agriculture, domestic work and skilled/unskilled work have varied odds of living with NCDs compared with those not working. Specifically, women working in professional/clerical, sale/service, agriculture, domestic work and skilled/unskilled were 1.19, 1.20, 1.30, 1.27 and 1.41 times more likely living with NCDs compared with their counterparts who were not working. As expected, the results revealed a statistically significant association between alcohol consumption and the risk of NCDs. Thus, individuals who consume alcohol have higher odds of living with NCDs (aOR 1.55, 95% CI 1.27 to 1.89) compared with those who have never had alcohol. Finally, the results show a positive association between increasing body weights and the risk of NCDs. Overweight and obese individuals have higher odds of living with NCDs compared with those with a normal BMI. However, underweight individuals have reduced odds of having NCDs.
Multivariate logistic regression on factors associated with non-communicable diseases (NCDs) in Kenya
Multivariate logistic regression on factors associated with simultaneous/co-occurrences of NCDs in Kenya
Table 5 presents the results of a multivariate logistic regression analysis examining factors associated with the simultaneous occurrence of NCDs. Increasing age was significantly associated with having one NCD, two NCDs or three NCDs. Similarly, alcohol consumption was consistently associated with higher odds of a single NCD or multiple NCDs. Being obese and working in agriculture was positively significant in predicting the likelihood of having one or two NCDs, but not three. Compared with those in the poorest wealth index, being in the middle wealth index was associated with higher odds of having one or two NCDs. Being married or formerly married increased the odds of having one NCD.
Multivariate logistic regression on factors associated with simultaneous/co-occurrences of NCDs in Kenya
Discussion
Several reports6–8 suggest that addressing the burden of NCDs particularly in resource-constrained settings like Kenya is quintessential to improving the quality of life of the population. This study set out with the aim of examining the burden of NCDs among women of reproductive age in Kenya. Our study revealed that 15.9% of Kenyan women were living with at least one type of NCD, with the most prevalent disease being hypertension (8.7%). The observed burden of NCDs among Kenyan women of reproductive age is lower compared with the burden reported in countries such as Nepal10 and Bangladesh.13 Our findings that hypertension is the most prevalent NCD among women of reproductive health in Kenya are inconsistent with that of another study that found mental health conditions, depression and anxiety, to be the the most common NCD.7 Unlike Mtintsilana et al’s study7 that focused on both those aged 18–35, the scope of our study was on the reproductive age 15–49 years. This wide age group explains the differences in the condition-specific prevalence observed in the current study. Nevertheless, our findings revealed that increasing age, increasing wealth, being married or formerly married, being overweight or obese, consuming alcohol and some occupations were risk factors for NCDs among women of reproductive age in Kenya.
The risk of having at least one of the NCDs considered in this study increased significantly with age. Further analysis revealed that increasing age predicted the co-occurrence of two or three NCDs among women of reproductive age. That is, the older the woman, the more likely they are to be at risk of any of the NCDs. Similar findings have been reported in Kenya,8 Nepal,10 Brazil.14 At the condition-specific level, we observed that older women of reproductive age were 7.92 times more likely to be hypertensive compared with adolescents. A similar direction of association by age was observed for diabetes, lung disease, depression and arthritis. A possible explanation for this is that ageing is associated with various changes in the body that can impair its normal functioning and increase its susceptibility to infections, inflammation and degeneration. Sapkota et al
15 also argue that ‘behavioural risk factors for NCDs, notably smoking, alcohol consumption and insufficient physical activity, increase with age’. This potentially explains the observed positive association between age and NCDs.
Our study also revealed that women in the richest wealth index were 1.6 times more likely to have at least one of the NCDs compared with those in the poorest wealth index—a finding that aligns with the results of a multicountry study7 that found higher odds of NCDs among those in the middle and highest socioeconomic statuses in Kenya and South Africa. Similar findings have been reported in Ghana16 and Bangladesh.17 The observed association was significant for hypertension, diabetes and arthritis. Several factors may explain the findings. One, women in the highest wealth index have been documented to often engage in less physically active work.18 This situation places them in a position of a sedentary lifestyle which is a known risk factor for hypertension and diabetes.19 20 Another possible explanation could be that women on high wealth index have the financial resources that grant them easy access (in terms of ability to afford) to a more affluent lifestyle, which may include a diet rich in processed foods, high in calories, sugars and unhealthy fats.
Being overweight or obese emerged as another significant risk factor for NCDs among women of reproductive age in Kenya. Obesity was associated with increased odds of co-occurrence of two or three NCDs. Our finding is corroborated by previous literature21–23 that has also found a strong positive association between overweight/obesity and the risk of NCDs. Biologically, obesity results in excess adipose tissues which exacerbates the production of proinflammatory cytokines, a known risk factor of NCDs.24 Moreover, obesity tends to worsen insulin resistance, hyperinsulinaemia and glucose intolerance25; thus, contributing significantly to the risk of diabetes.
Our study also revealed that women who consumed alcohol were 1.55 times more likely to have at least one NCD compared with those who did not consume alcohol. A further interrogation of the data revealed that consuming alcohol predicted the co-occurrence of two or three NCDs. This association was significant for hypertension, lung disease, depression and anxiety. The result is analogous to findings from prior research conducted in the United States of America26 and Ghana.27 Biological pathways may explain this association. For instance, alcohol consumption has been reported to be significantly related to ‘pancreatic β-cell dysfunction and apoptosis’ which increases one’s risk of diabetes.
Contrary to previous evidence that found no significant association between marital status and NCDs,27 we found that married women and formerly married women had higher likelihoods of having at least one of the nine NCDs. Notwithstanding, our findings align with a study by Segawa et al
28 which showed that married and formerly married women (ie, divorced or separated) were 1.27 times and 1.18 times more likely to have an NCD, respectively, compared with those who had never married. Both marriage and marital dissolution are often associated with significant emotional stress.16 Such high levels of stress, whether due to relationship issues, financial concerns or other life changes, may lead to hormonal imbalances and unhealthy coping mechanisms such as alcohol consumption, smoking and stress-induced eating which are all underlying factors that exacerbate the risk of NCDs.29 The current study further indicates that women who are employed have higher odds of having at least one NCD. This is consistent with studies from Kenya30 and Thailand.31 The association could be explained from the perspective that women who are employed have a higher tendency to engage in high-risk dietary behaviours including high salt intake and low vegetable consumption.30
This study’s strength lies in the use of a population-based, nationally representative dataset; thus, supporting the extrapolation of the findings to the larger population of Kenyan women of reproductive age. Additionally, the study is more comprehensive because we considered more NCDs than any related study8 that has been conducted in Kenya. Nonetheless, the cross-sectional nature of the DHS precludes us from establishing any causal relationship between the explanatory variables and risks of NCDs. Furthermore, the relationship between alcohol use and NCDs could be bidirectional, however, we are unable to ascertain that from the data. Given the restriction of the analysis to women of reproductive age, the findings may not be applicable to other populations such as men and older people (60 years and older). Also, since this study was reliant on variables available in the Kenya DHS, it meant that other key confounders such as level of physical activity, sleep duration and quality, cultural norms, and dietary patterns could not be controlled for in the regression analyses.
Implications for policy and practice
Evidence from our study suggests that hypertension is the most pervasive NCD affecting women of reproductive age. It, thus, underscores a need for the Kenyan Government through its health department to prioritise hypertension. To address the increasing burden of NCDs with age, policy initiatives should consider age-specific strategies that cater to the unique healthcare needs of adolescents, young adults and older women. Given the strong positive association between overweight/obesity and NCDs, public health education should intensify awareness about the health risks associated with having a higher BMI. Moreover, women of reproductive health must be encouraged to stay physically active and conscious about maintaining a normal weight. Considering the significant association between alcohol consumption and the risk of NCDs, policies aimed at reducing alcohol consumption should be implemented. This may include public health campaigns, alcohol taxation and regulations on marketing and sales of alcoholic beverages.
This post was originally published on https://bmjopen.bmj.com