You are here

  • Home
  • Archive
  • Volume 13, Issue 2
  • Multimorbidity of cardiometabolic diseases: a cross-sectional study of patterns, clusters and associated risk factors in sub-Saharan Africa

Article Text

Article menu
Cardiovascular medicine
Original research
Multimorbidity of cardiometabolic diseases: a cross-sectional study of patterns, clusters and associated risk factors in sub-Saharan Africa
  1. Peter Otieno1,2,3,
  2. Gershim Asiki1,4,
  3. Frederick Wekesah1,5,
  4. Calistus Wilunda1,
  5. Richard E Sanya1,
  6. Welcome Wami3,
  7. Charles Agyemang2
  1. 1Chronic Diseases Management Unit, African Population and Health Research Center, Nairobi, Kenya
  2. 2Department of Public & Occupational Health, Amsterdam Public Health Research Institute, Amsterdam, The Netherlands
  3. 3Amsterdam Institute for Global Health and Development, Amsterdam, The Netherlands
  4. 4Department of Women’s and Children’s Health, Karolinska Institutet, Stockholm, Sweden
  5. 5Lown Scholars Program, Harvard University T H Chan School of Public Health, Boston, Massachusetts, USA
  1. Correspondence to Peter Otieno; pootienoh{at}gmail.com

Abstract

Objective To determine the patterns of cardiometabolic multimorbidity and associated risk factors in sub-Saharan Africa (SSA).

Design We used data from the WHO STEPwise approach to non-communicable disease risk factor surveillance cross-sectional surveys conducted between 2014 and 2017.

Participants The participants comprised 39, 658 respondents aged 15–69 years randomly selected from nine SSA countries using a multistage stratified sampling design.

Primary outcome measure Using latent class analysis and agglomerative hierarchical clustering algorithms, we analysed the clustering of cardiometabolic diseases (CMDs) including high blood sugar, hypercholesterolaemia, hypertension and cardiovascular diseases (CVDs) such as heart attack, angina and stroke. Clusters of lifestyle risk factors: harmful salt intake, physical inactivity, obesity, tobacco and alcohol use were also computed. Prevalence ratios (PR) from modified Poisson regression were used to assess the association of cardiometabolic multimorbidity with sociodemographic and lifestyle risk factors.

Results Two distinct classes of CMDs were identified: relatively healthy group with minimal CMDs (95.2%) and cardiometabolic multimorbidity class comprising participants with high blood sugar, hypercholesterolaemia, hypertension and CVDs (4.8%). The clusters of lifestyle risk factors included alcohol, tobacco and harmful salt consumption (27.0%), and physical inactivity and obesity (5.8%). The cardiometabolic multimorbidity cluster exhibited unique sociodemographic and lifestyle risk profiles. Being female (PR=1.7, 95% CI (1.5 to 2.0), middle-aged (35–54 years) (3.9 (95% CI 3.2 to 4.8)), compared with age 15–34 years, employed (1.2 (95% CI 1.1 to 1.4)), having tertiary education (2.5 (95% CI 2.0 to 3.3)), vs no formal education and clustering of physical inactivity and obesity (2.4 (95% CI 2.0 to 2.8)) were associated with a higher likelihood of cardiometabolic multimorbidity.

Conclusion Our findings show that cardiometabolic multimorbidity and lifestyle risk factors cluster in distinct patterns with a disproportionate burden among women, middle-aged, persons in high socioeconomic positions, and those with sedentary lifestyles and obesity. These results provide insights for health systems response in SSA to focus on these clusters as potential targets for integrated care.

  • public health
  • diabetes & endocrinology
  • general diabetes
  • epidemiology
  • hypertension

Data availability statement

Data are available in a public, open access repository. Data are available in a public, open-access repository. No additional data are available.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjopen-2022-064275

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    STRENGTHS AND LIMITATIONS OF THIS STUDY

    • Data used in this analysis are from nationally representative population-based surveys conducted in nine countries in sub-Saharan Africa using a standardised WHO-STEPwise approach to non-communicable disease risk factor surveillance protocol. Hence, the findings are generalisable to the populations of these countries.

    • The screening for cardiometabolic diseases is based on direct measures of blood pressure, anthropometry, key biomarkers and self-reports thereby allowing for an objective assessment of multimorbidity.

    • This study provides crucial evidence on population-based cardiometabolic multimorbidity patterns among broader age ranges comprising young, middle-aged older persons.

    • Data used in the analysis are from cross-sectional studies; thus, it is not possible to draw causal inferences for cardiometabolic multimorbidity and temporal associations with sociodemographic and lifestyle risk factors. However, the findings provide useful insights for policy makers and health service providers to prioritise risk-centred approaches for prevention, early detection and treatment of cardiometabolic multimorbidity.

    Introduction

    Cardiometabolic diseases (CMDs) are the leading cause of global mortality and disability.1 Over half of the global burden of non-communicable diseases (NCDs) are attributable to CMDs that share four major risk factors: harmful alcohol use, unhealthy diet, tobacco use and physical inactivity.2 3 Three-quarters of the global CMD-related mortality occurs in low-income and middle-income countries with 30% classified as premature deaths.4 People living with CMDs often present with multiple conditions including diabetes mellitus, hypercholesterolaemia, hypertension and stroke.5–7 Hence the concept of cardiometabolic multimorbidity.

    Although CMDs are estimated to take away up to 12 years in life expectancy,8 9 their onset can be prevented or postponed by the elimination of lifestyle risk factors.10 11 Most studies have consistently identified cardiometabolic multimorbidity as a common cluster.12–15 However, there are still several knowledge gaps on the multimorbidity spectrum in sub-Saharan Africa (SSA). First, literature is scarce on clustering CMDs and behavioural risk factors because extant evidence has not applied statistical methods to separate the random and non-random co-occurrence.16–19 Moreover, most studies have mainly focused on analysing the most prevalent CMD combinations.16–20 The prevalence of different disease combinations is significantly associated with the prevalence of the individual comorbid diseases in question. Hence, a high prevalence of a particular multimorbidity combination does not provide sufficient evidence to support a non-random co-occurrence.21 Accounting for the random coexistence of multimorbidity requires rigorous assessment of multimorbidity clusters and lifestyle risk factors.22

    Evidence of clustering of CMDs and lifestyle risk factors is needed to prioritise healthcare services for the most frequently co-occurring combinations.23 However, the methodological differences in the conceptualisation of multimorbidity in previous studies have hindered the comparison of findings.13 24 Furthermore, most studies have been conducted among older age groups in primary care settings where multimorbidity are more likely to occur.13 24 To date, little research in SSA has investigated multimorbidity patterns among broader age ranges from the general population.

    This study, therefore, sought to investigate patterns and clusters of cardiometabolic multimorbidity and associated factors among persons aged 15–69 years using nationally representative WHO STEPwise surveys from nine countries in SSA.

    Methods

    Study design

    We used data from the WHO STEPwise approach to NCD risk factor surveillance (STEPS) surveys in SSA. Details of the data collection methods have been published elsewhere.25 Briefly, STEPS surveys are nationally representative population-based cross-sectional surveys of risk factors for NCDs in participants aged 15–69 years. The WHO STEPS surveys aims to provide baseline national estimates for NCD indicators to inform health policies in the study countries.26

    A standard sampling protocol for the WHO STEPS survey was used in all the study countries.27 The participants were selected using a multistage stratified sampling design. First, sampling clusters or enumeration areas (EAs) were selected using probability proportional to the size of the number of households in the cluster. Second, samples of households were randomly drawn from a household listing in the cluster. Eligible participants comprised all listed household members aged 18–69 years residing in the sampled households for at least 6 months preceding the survey. In some countries, the minimum age was 15 years.28–30 Finally, the Kish sampling grid was used to randomly select one study participant from a list of all eligible household members.25

    Data abstraction

    Data used in the current study were collected using interviewer-administered structured questionnaires modified from the WHO STEPS tool.25 The data were from self-reports and direct measurements of anthropometry and key biomarkers. The variables were measured using the WHO criteria.27 Self-reported information comprised sociodemographic characteristics such as age, sex, education and employment; behavioural risk factors such as physical inactivity, tobacco and alcohol use, harmful salt consumption; and clinical histories of high blood sugar, hypercholesterolaemia, hypertension and cardiovascular diseases (CVDs) (heart attack, angina and stroke). The physical measurements comprised screening for blood pressure and anthropometrics such as weight (kg) and height (m). The biochemical measurements comprised fasting blood samples for cholesterol and blood sugar measurements in mmol/L or mg/dL. Random blood glucose was collected for 117/39 658 participants who failed to fast as instructed. The measurements of the variables used in this study are shown in table 1.

    View this table:
    Table 1

    Measurements of study of the variables

    Data inclusion and exclusion criteria

    We used a two-stage inclusion criteria for the present study (see online supplemental file 1). The first step involved the selection of eligible study countries while in the second step, eligible participants were selected from the latest round of the STEPS survey in each of the study countries (see online supplemental file 2). In the first step, a country was eligible for inclusion in the analysis if it met the following criteria: (1) Had data collected between 2000 and 2020; (2) The survey was nationally representative and (3) Had data on key variables comprising blood sugar, cholesterol, body mass index, blood pressure and CVD status. Others comprised age, sex, education, employment, alcohol consumption, smoking, diet and physical activity. In the second step, eligible participants were included in the analysis if they were not pregnant. In the end, data from 9 countries with 39 658 participants were included in the analysis (see online supplemental file 3).

    Definition of variables

    The outcome variable for the present analysis was defined as the clustering of CMDs comprising high blood sugar, hypercholesterolaemia, hypertension and CVDs (heart attack, angina and stroke). The explanatory variables included sociodemographic characteristics: age, sex, education level and employment status. Other covariates comprised clusters of multiple lifestyle risk factors including harmful salt intake, physical inactivity, obesity, tobacco use and alcohol consumption. Clusters were named based on their unique dominant CMDs and risk profiles.

    Data analysis

    Applying sample weights

    We accounted for the complex survey design using the svyset command in Stata, by defining clusters and sampling weights. The original country-level datasets were weighted using the probability of selection at each stage of sampling. Thus, the participants’ weight was equal to the inverse of the product of the probability of the selection of the EA or sampling cluster, the probability of household selection, the probability of selection within the household and the age–sex distribution of the study country.

    Characteristics of participants were summarised using frequencies and percentages. Given the exclusion of participants with missing data on the key variables, we compared the characteristics of the participants with complete data and those with incomplete data and found no differences between the two groups based on age, sex and employment status (See online supplemental file 4).

    Latent class analysis

    We used latent class analysis (LCA) to identify distinct groups of cardiometabolic multimorbidity and lifestyle risk factors. The LCA is a structural equation modelling-based approach used to identify groups of participants with homogeneous response patterns to a set of

    observed variables.31 We determined the optimal number of latent classes using the adjusted Bayesian information criterion (BIC). The BIC has been previously used as a robust indicator for determining the optimal number of classes for latent variables.32 33 First, the BIC was used to compare several plausible models. The model with the lowest values of BIC was finally selected as the best-fitting model.34 35 The posterior probabilities were used to determine the likelihood of class membership. Finally, the participants were grouped into the classes with the highest-class probability.

    Hierarchical cluster analysis

    We conducted a supplementary analysis of the cardiometabolic multimorbidity patterns and clustering of lifestyle risk factors using the agglomerative hierarchical cluster analysis.36 First, the proximity index grouped individual CMDs and lifestyle risk factors into a single cluster. Second, the individual clusters were gradually merged with the most closely related clusters until a single cluster with all the elements was obtained. We used the average linkage method to accommodate the spread of the clusters.36 The number of clusters was assessed using a dendrogram and Jaccard similarity coefficient. In summary, the Jaccard coefficient is a measure of similarity between two sets or clusters expressed as a percentage. The higher the percentage, the more similar the two clusters are.37

    Modified Poison regression

    We used modified Poisson regression to assess the sociodemographic variables and clusters of lifestyle risk factors associated with cardiometabolic multimorbidity. The sociodemographic variables comprising age, sex, education level, employment status and clusters of lifestyle risk factors were regressed against a dummy outcome variable indicating whether the participant was in class 1 (healthy class with minimal CMDs) or class 2 (cardiometabolic multimorbidity). The selection of the sociodemographic and lifestyle risk factors for the multivariable model was based on the variables conceptualised from the WHO’s theoretical framework on the social determinants of NCDs38 as more traditional level p values such as 0.05 used to select variables can fail in identifying variables known to be important.39 We adjusted for country of residence by including the country dummy variable in the models, as in previous publications from WHO surveys.40–42 We found no multicollinearity among the independent variables based on the assessment of variance inflation factors.43 The adjusted prevalence ratios (PRs) and 95% CI from modified Poisson regression were used to determine the sociodemographic and lifestyle risk factors associated with cardiometabolic multimorbidity.

    All statistical analyses were carried out in Stata V.15 (StataCorp).

    Patient and public involvement

    Patients and/or the public were not involved in the design, conduct, reporting or dissemination of this research.

    Results

    Characteristics of participants

    In total, 39 658 participants were included in the analysis. Table 2 shows the characteristics of the study participants. In general, most of the participants were women, belonged to the youngest age group (15–34 years), had a primary level of education and were unemployed.

    View this table:
    Table 2

    Sociodemographic and health characteristics of the study participants

    Harmful salt intake (37.5%) and alcohol use (27.9%) were the most prevalent lifestyle risk factors. The most prevalent CMD was hypertension (24.9%). Varying patterns in the distribution CMDs and lifestyle risk factors were observed among the study countries. Tobacco use (20.7%), harmful salt intake (98.6%), physical inactivity (25.6%) and hypertension (36.2%) were highest in Botswana. Ethiopia had the highest prevalence of alcohol use (43.3%). Obesity and hypercholesterolaemia were highest in Swaziland (19.5%) and Benin (13.9%). Sudan and Zambia had the highest prevalence of high blood sugar (7.4%) while CVD prevalence was highest in Uganda (9.2%).

    Findings of LCA

    Clustering of lifestyle risk factors

    The classes of lifestyle risk factors are shown in figure 1. We compared LCA models with 1–4 classes (see online supplemental file 5). The three-class model had the lowest BIC and thus was selected as the best-fit model. Class one comprised participants with minimal lifestyle risks (67.2%). Class two included participants with high probabilities of alcohol use, smoking and harmful salt consumption (27.0%). Class three comprised participants with the highest probability of physical inactivity and obesity (5.8%).

    Figure 1
    Figure 1

    Latent classes of lifestyle risk factors.

    Cardiometabolic multimorbidity classes

    The classes of CMDs are shown in figure 2. We ran LCA models from 1 to 4 classes selecting the two-class model based on indices of fit (see online supplemental file 5). The two-class model had the lowest BIC and thus was selected as the best-fit model. Class one (interpreted as the ‘relatively healthy group’) comprised participants with minimal multimorbidity (95.2%). Class two (cardiometabolic multimorbidity) included participants with the highest probability of high blood sugar, hypercholesterolaemia, hypertension and CVDs (4.8%).

    Figure 2
    Figure 2

    Latent classes of cardiometabolic diseases.

    Hierarchical cluster analysis findings

    As a supplementary analysis, we also calculated cardiometabolic multimorbidity patterns and clustering of lifestyle risk factors using the agglomerative hierarchical cluster analysis. Figure 3 shows the hierarchical tree plot (dendrogram). The dendrogram displays the agglomeration schedules at which CMDs and lifestyle risk factors are combined. In general, the findings were similar to those obtained using LCA. The hierarchical clustering algorithms revealed two distinct groupings of lifestyle risk factors. Based on the proximity coefficients, physical inactivity and obesity formed the first cluster. Alcohol use, harmful salt intake and tobacco use combined to form the second cluster. The proximity coefficients from the hierarchical cluster analysis revealed clustering of hypertension, hypercholesterolaemia, high blood sugar and CVDs.

    Figure 3
    Figure 3

    Dendrogram of lifestyle risk factors and cardiometabolic diseases.

    Distribution of cardiometabolic multimorbidity by sociodemographic and lifestyle risk factors

    The distribution of cardiometabolic multimorbidity by sociodemographic and lifestyle risk factors is presented in table 3. The clustering of cardiometabolic multimorbidity (hypertension, high blood sugar, hypercholesterolaemia and CVD) was highest in middle-aged and older participants, females, employed and those with tertiary education. Varying patterns in the distribution of cardiometabolic multimorbidity were observed among the study countries. The prevalence of cardiometabolic multimorbidity was highest in Sudan and Botswana (9.9% and 9.3%) and lowest in Ethiopia (2.0%).

    View this table:
    Table 3

    Distribution of cardiometabolic multimorbidity by sociodemographic and lifestyle risk factors

    Sociodemographic and lifestyle risk factors associated with cardiometabolic multimorbidity

    Figure 4 shows the sociodemographic and lifestyle risk factors associated with cardiometabolic multimorbidity. Being female (PR=1.7, 95% CI (1.5 to 2.0), middle-aged (35–54 years) (3.9 (95% CI 3.2 to 4.8)) and older age (55–69 years) (3.9 (95% CI 3.2 to 4.8)), compared with age 15–34 years were associated with a higher likelihood of cardiometabolic multimorbidity. Participants in the highest socioeconomic position such as tertiary education (2.5 (95% CI 2.0 to 3.3)) vs no formal schooling and those employed (1.2 (95% CI 1.1 to 1.4)) vs unemployed were more likely to have cardiometabolic multimorbidity. The likelihood of cardiometabolic multimorbidity was higher among participants with co-occurring physical inactivity and obesity (2.4 (95% CI 2.0 to 2.8)) compared with those with minimal lifestyle risk factors.

    Figure 4
    Figure 4

    Sociodemographic and lifestyle risk factors associated and cardiometabolic multimorbidity. Adjusted for study countries; cardiometabolic multimorbidity was defined as the clustering of hypercholesterolemia, hypertension, high blood sugar and CVDs.

    Discussion

    In this study, we examined the patterns of cardiometabolic multimorbidity and associated risk factors in persons aged 15 years and older in SSA. Two distinct classes of CMDs were identified: a relatively healthy group with minimal CMDs and a cardiometabolic multimorbidity class comprising participants with high blood sugar, hypercholesterolaemia, hypertension and CVDs. Three clusters of lifestyle risk factors were yielded. Class 1 comprised participants with minimal lifestyle risks. Class 2 included alcohol users, smokers and harmful salt consumers. Class 3 comprised clustering of physical inactivity and obesity. Our findings show that cardiometabolic multimorbidity cluster in distinct patterns with a disproportionate burden among women, middle-aged, persons in the highest socioeconomic positions, and those with sedentary lifestyles and obesity.

    The clusters of cardiometabolic multimorbidity identified in our study have similarities with findings from other studies in SSA.44 45 Several underlying pathophysiological mechanisms could explain the clustering of high blood sugar, hypercholesterolaemia, hypertension and CVDs. Insulin resistance is known as a possible mechanism explaining the clustering of CMDs.46–48 The abnormalities in the metabolism process could be due to in part insulin resistance which may also lead to defects in vascular reactivity.49 The clustering of lifestyle risk factors such as smoking, alcohol use and harmful salt consumption in our study could be due to the interplay among various sociobehavioural factors that affect smoking, drinking behaviour and diet.50 This was consistent with findings from other studies in different parts of the world.50–52 Given that CMDs share common lifestyle risk factors,2 3 our results provide crucial insights into the need to scale up population-level primary and secondary prevention programmes. Primary prevention should target co-occurring lifestyle risk factors. Previous studies have shown that healthy lifestyles in early life often persist into adulthood and old age, preventing or delaying CMDs.53 Secondary prevention should target persons living with CMDs, such as regular screening for multimorbidity, self-monitoring and adopting healthy lifestyles to prevent or delay the onset of multimorbidity.

    The gender disparity in the distribution of cardiometabolic multimorbidity observed in our study may be partly attributed to the inequalities in occupational and domestic activities and the gender differences in risk factors for CVDs such as lifestyle and diet.54 55 Our results suggest that cardiometabolic multimorbidity is not limited to older persons but is a common phenomenon among middle-aged persons in the study countries. These results mirror those of several multimorbidity studies in support of the emerging evidence on the high burden of multimorbidity in middle-aged persons.53 56–58 Studies conducted in Brazil, Guatemala, India, the Philippines and South Africa have also shown that people living in low-income and middle-income countries tend to experience CMDs much earlier in life than their counterparts living in high-income countries due to socioeconomic hardships and poor access to healthcare services.59–61

    Similar to previous studies,62–64 our results show that increasing levels of educational achievements and being employed were associated with a greater likelihood of cardiometabolic multimorbidity. This may be due to the fact that with higher education and employment also comes affluence,62–64 and greater access to alcohol, tobacco and unhealthy diets.65 However, other studies have also shown that higher education achievement was also associated with a low likelihood of clustering of multiple behavioural risk factors for CVDs, possibly because a higher level of education may also increase both awareness of, and capacity for lifestyle modification.50 66 Further studies on the socioeconomic determinants of cardiometabolic multimorbidity are needed to elucidate these results.

    The likelihood of cardiometabolic multimorbidity was higher among participants with co-occurring physical inactivity and obesity compared with those with minimal lifestyle risk factors. This finding is in line with previous studies that pointed in the same direction.67–69 The mechanisms through which physical activity increases the risk of cardiometabolic multimorbidity are known. Previous physical activity intervention studies have demonstrated consistent improvements in various CVD risk factors such as hypertension,70 HDL cholesterol,71 C reactive protein and other inflammatory markers.72 73 Consequently, further longitudinal studies are needed to elucidate the most critical sociodemographic factors attributable to the clustering of physical inactivity and obesity and its long-term effects on cardiometabolic multimorbidity.

    Overall, our findings have two main implications. First, there is a need for further policy discourse on the integrated management of CMDs in primary care settings in SSA. The clustering of CMDs and lifestyle risk factors in the population is important for clinicians, policy makers and researchers in prioritising the needs and care processes for patients living with or at risk of multimorbidity. A paradigm shift towards comprehensive care may enable patients presenting with CMDs at primary care to be regularly screened for other chronic conditions. A shift away from fragmented care may also improve access to quality healthcare services, especially in SSA where persons living with CMDs remain undiagnosed for several years and the majority of those on treatment often remain uncontrolled.74 Second, our study provides baseline estimates for future researchers to design longitudinal studies on the burden and aetiology of the most common clusters of CMDs in SSA.

    Strengths and limitations

    Data used in this analysis are from nationally representative population-based surveys conducted in nine countries in SSA using a standardised WHO-STEPS protocol. Hence, the findings are generalisable to the populations of these countries. Second, most previous studies were conducted among older age groups in primary care settings where multimorbidities are more likely to occur.13 24 Our study bridges this gap by providing crucial evidence on population-based multimorbidity patterns among broader age ranges comprising young, middle-aged and older persons. Third, the data used are based on direct measures of blood pressure, anthropometry, key biomarkers and self-reports allowing for a more objective screening for CMDs than self-reporting used in over three-quarters of previous studies.75

    Limitations stem from the fact that the data used in the analysis are from cross-sectional studies thus; it, is not possible to draw causal inferences for cardiometabolic multimorbidity and temporal associations with sociodemographic and multiple lifestyle risks. Moreover, our study does not give an idea of the index disease in the multimorbidity clusters identified. Second, the screening for CVDs such as heart attack, angina and stroke was based on self-reported history of clinical diagnosis, which may have led to information bias, possibly underestimating the prevalence of CVDs. Lastly, the current study is based on pooled data sets from several countries; hence, the findings may be limited by the variations in the sociocultural and economic contexts within the study countries.

    Conclusions

    Our findings show that cardiometabolic multimorbidity and lifestyle risk factors cluster in distinct patterns with a disproportionate burden among persons in the highest socioeconomic positions, women, the middle-aged and those with sedentary lifestyles and obesity. These results provide useful insights for health systems response in SSA to focus on these clusters as potential targets in the development and segmentation of integrated care for care cardiometabolic multimorbidity. We draw attention to the need for healthcare systems in SSA to prioritise risk-centred management of CMDs by incorporating aggressive approaches for prevention, early detection, treatment and promotion of healthy lifestyles to avert the occurrence of multimorbidity.

    Data availability statement

    Data are available in a public, open access repository. Data are available in a public, open-access repository. No additional data are available.

    Ethics statements

    Patient consent for publication

    Ethics approval

    Data used in this study are publicly available on the NCD microdata repository of the WHO (https://extranet.who.int/ncdsmicrodata/index.php/catalog/STEPS)76 . We obtained formal written permission from the WHO for the surveys included. Since we used publicly available data, no additional approval was required from an institutional ethics review board.

    References

    1. ↵
      1. Wang H
      . Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the global burden of disease study 2015. Lancet 2016;388:1459–544. doi:10.1016/S0140-6736(16)31012-1
      OpenUrlCrossRefPubMed
    2. ↵
      1. Frieden TR,
      2. Cobb LK,
      3. Leidig RC, et al
      . Reducing premature mortality from cardiovascular and other non-communicable diseases by one third: achieving sustainable development goal indicator 3.4.1. Glob Heart 2020;15:50. doi:10.5334/gh.531
    3. ↵
      1. Naghavi M
      . Global, regional, and national age-sex specific mortality for 264 causes of death, 1980-2016: a systematic analysis for the global burden of disease study 2016. Lancet 2017;390:1151–210. doi:10.1016/S0140-6736(17)32152-9
      OpenUrlCrossRefPubMed
    4. ↵
      1. Roth GA,
      2. Mensah GA,
      3. Johnson CO, et al
      . Global burden of cardiovascular diseases and risk factors, 1990-2019: update from the GBD 2019 study. J Am Coll Cardiol 2020;76:2982–3021. doi:10.1016/j.jacc.2020.11.010
      OpenUrlCrossRefPubMed
    5. ↵
      1. Bayliss EA,
      2. Bayliss MS,
      3. Ware JE, et al
      . Predicting declines in physical function in persons with multiple chronic medical conditions: what we can learn from the medical problem list. Health Qual Life Outcomes 2004;2:1–8.:47. doi:10.1186/1477-7525-2-47
      OpenUrlCrossRefPubMed
    6. ↵
      1. Garin N,
      2. Olaya B,
      3. Moneta MV, et al
      . Impact of multimorbidity on disability and quality of life in the Spanish older population. PLoS One 2014;9:e111498. doi:10.1371/journal.pone.0111498
    7. ↵
      1. Marengoni A,
      2. von Strauss E,
      3. Rizzuto D, et al
      . The impact of chronic multimorbidity and disability on functional decline and survival in elderly persons. A community-based, longitudinal study. J Intern Med 2009;265:288–95. doi:10.1111/j.1365-2796.2008.02017.x
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Chudasama YV,
      2. Khunti K,
      3. Davies MJ
      . Clustering of comorbidities. Future Healthc J 2021;8:e224–9. doi:10.7861/fhj.2021-0085
      OpenUrl
    9. ↵
      1. Emerging Risk Factors Collaboration,
      2. Di Angelantonio E,
      3. Kaptoge S, et al
      . Association of cardiometabolic multimorbidity with mortality. JAMA 2015;314:52–60. doi:10.1001/jama.2015.7008
      OpenUrlCrossRefPubMed
    10. ↵
      1. Boulware LE,
      2. Marinopoulos S,
      3. Phillips KA, et al
      . Systematic review: the value of the periodic health evaluation. Ann Intern Med 2007;146:289–300. doi:10.7326/0003-4819-146-4-200702200-00008
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Si S,
      2. Moss JR,
      3. Sullivan TR, et al
      . Effectiveness of general practice-based health checks: a systematic review and meta-analysis. Br J Gen Pract 2014;64:e47–53. doi:10.3399/bjgp14X676456
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Zemedikun DT,
      2. Gray LJ,
      3. Khunti K, et al
      . Patterns of multimorbidity in middle-aged and older adults: an analysis of the UK biobank data. Mayo Clin Proc 2018;93:857–66. doi:10.1016/j.mayocp.2018.02.012
      OpenUrlPubMed
    13. ↵
      1. Ng SK,
      2. Tawiah R,
      3. Sawyer M, et al
      . Patterns of multimorbid health conditions: a systematic review of analytical methods and comparison analysis. Int J Epidemiol 2018;47:1687–704. doi:10.1093/ije/dyy134
      OpenUrlCrossRefPubMed
    14. ↵
      1. Sinnige J,
      2. Braspenning J,
      3. Schellevis F, et al
      . The prevalence of disease clusters in older adults with multiple chronic diseases -- a systematic literature review. PLoS One 2013;8:e79641. doi:10.1371/journal.pone.0079641
    15. ↵
      1. Violan C,
      2. Foguet-Boreu Q,
      3. Flores-Mateo G, et al
      . Prevalence, determinants and patterns of multimorbidity in primary care: a systematic review of observational studies. PLoS One 2014;9:e102149. doi:10.1371/journal.pone.0102149
    16. ↵
      1. Chang AY,
      2. Gómez-Olivé FX,
      3. Payne C, et al
      . Chronic multimorbidity among older adults in rural south africa. BMJ Glob Health 2019;4:e001386. doi:10.1136/bmjgh-2018-001386
    17. ↵
      1. Ekoru K,
      2. Doumatey A,
      3. Bentley AR, et al
      . Type 2 diabetes complications and comorbidity in sub-Saharan Africans. EClinicalMedicine 2019;16:30–41. doi:10.1016/j.eclinm.2019.09.001
      OpenUrl
    18. ↵
      1. Lalkhen H,
      2. Mash R
      . Multimorbidity in non-communicable diseases in South African primary healthcare. S Afr Med J 2015;105:134–8. doi:10.7196/samj.8696
      OpenUrlPubMed
    19. ↵
      1. Mutyambizi C,
      2. Chola L,
      3. Groot W, et al
      . The extent and determinants of diabetes and cardiovascular disease comorbidity in south africa - results from the south african national health and nutrition examination survey (SANHANES-1). BMC Public Health 2017;17:745. doi:10.1186/s12889-017-4792-8
    20. ↵
      1. Sharman M,
      2. Bachmann M
      . Prevalence and health effects of communicable and non-communicable disease comorbidity in rural KwaZulu-Natal, South Africa. Trop Med Int Health 2019;24:1198–207. doi:10.1111/tmi.13297
      OpenUrlPubMed
    21. ↵
      1. De Francesco D,
      2. Sabin CA,
      3. Reiss P
      . Multimorbidity patterns in people with HIV. Curr Opin HIV AIDS 2020;15:110–7. doi:10.1097/COH.0000000000000595
      OpenUrlPubMed
    22. ↵
      1. van den Akker M,
      2. Buntinx F,
      3. Roos S, et al
      . Problems in determining occurrence rates of multimorbidity. J Clin Epidemiol 2001;54:675–9. doi:10.1016/s0895-4356(00)00358-9
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Valderas JM,
      2. Starfield B,
      3. Sibbald B, et al
      . Defining comorbidity: implications for understanding health and health services. Ann Fam Med 2009;7:357–63. doi:10.1370/afm.983
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Prados-Torres A,
      2. Calderón-Larrañaga A,
      3. Hancco-Saavedra J, et al
      . Multimorbidity patterns: a systematic review. J Clin Epidemiol 2014;67:254–66. doi:10.1016/j.jclinepi.2013.09.021
      OpenUrlCrossRefPubMed
    25. ↵
      1. Riley L,
      2. Guthold R,
      3. Cowan M, et al
      . The world Health organization stepwise approach to noncommunicable disease risk-factor surveillance: methods, challenges, and opportunities. Am J Public Health 2016;106:74–8. doi:10.2105/AJPH.2015.302962
      OpenUrlCrossRefPubMed
    26. ↵
      1. World Health Organization
      . The WHO stepwise approach to surveillance of noncommunicable diseases (STEPS). Geneva, Switzerland,
    27. ↵
      1. World Health Organization
      . STEPS manual, STEPS instrument; 2011.
    28. ↵
      1. Ethiopian Public Health Institute
      . Ethiopia STEPS survey 2015. Addis Ababa, Ethiopia The Ethiopian Public Health Institute; 2015.
    29. ↵
      1. Ministry of Health
      . Swaziland WHO STEPS noncommunicable disease risk factor surveillance report. Swaziland Ministry of Health; 2014.
    30. ↵
      1. Ministry Of Health & Wellness
      . Botswana 2014 STEPS survey report on non-communicable disease risk factors; 2015.
    31. ↵
      1. Formann AK,
      2. Kohlmann T
      . Latent class analysis in medical research. Stat Methods Med Res 1996;5:179–211. doi:10.1177/096228029600500205
      OpenUrlCrossRefPubMed
    32. ↵
      1. Nylund KL,
      2. Asparouhov T,
      3. Muthén BO
      . Deciding on the number of classes in latent class analysis and growth mixture modeling: a monte carlo simulation study. Structural Equation Modeling: A Multidisciplinary Journal 2007;14:535–69. doi:10.1080/10705510701575396
      OpenUrl
    33. ↵
      1. Masyn KE
      . 25 latent class analysis and finite mixture modeling. Oxford University Press Oxford, 2013: 551.
    34. ↵
      1. Akaike H
      . A new look at the statistical model identification. IEEE Trans Automat Contr 1974;19:716–23. doi:10.1109/TAC.1974.1100705
      OpenUrl
    35. ↵
      1. Schwarz G
      . Estimating the dimension of a model. Ann Statist 1978;6:461–4. doi:10.1214/aos/1176344136
      OpenUrl
    36. ↵
      1. Everitt BS,
      2. Landau S,
      3. Leese M, et al
      . Cluster analysis. 5th ed. John Wiley, 7 January 2011. doi:10.1002/9780470977811
    37. ↵
      1. Glaser N,
      2. Barnett P,
      3. McCaslin I, et al
      . Risk factors for cerebral edema in children with diabetic ketoacidosis. the pediatric emergency medicine collaborative research committee of the american academy of pediatrics. N Engl J Med 2001;344:264–9. doi:10.1056/NEJM200101253440404
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. World Health Organization
      . Preventing chronic diseases: a vital investment. Geneva World Health Organization; 2005.
    39. ↵
      1. Grant SW,
      2. Hickey GL,
      3. Head SJ
      . Statistical primer: multivariable regression considerations and pitfalls. Eur J Cardiothorac Surg 2019;55:179–85. doi:10.1093/ejcts/ezy403
      OpenUrlCrossRef
    40. ↵
      1. Koyanagi A,
      2. Oh H,
      3. Stickley A, et al
      . Risk and functional significance of psychotic experiences among individuals with depression in 44 low- and middle-income countries. Psychol Med 2016;46:2655–65. doi:10.1017/S0033291716001422
      OpenUrlPubMed
    41. ↵
      1. Stubbs B,
      2. Koyanagi A,
      3. Veronese N, et al
      . Physical multimorbidity and psychosis: comprehensive cross sectional analysis including 242,952 people across 48 low- and middle-income countries. BMC Med 2016;14:189. doi:10.1186/s12916-016-0734-z
    42. ↵
      1. Stubbs B,
      2. Siddiqi K,
      3. Elsey H, et al
      . Tuberculosis and non-communicable disease multimorbidity: an analysis of the world health survey in 48 low- and middle-income countries. Int J Environ Res Public Health 2021;18:2439. doi:10.3390/ijerph18052439
    43. ↵
      1. Fotouhi AR
      . Modelling overdispersion in longitudinal count data in clinical trials with application to epileptic data. Contemp Clin Trials 2008;29:547–54. doi:10.1016/j.cct.2008.01.005
      OpenUrlPubMed
    44. ↵
      1. Chang AY,
      2. Gómez-Olivé FX,
      3. Manne-Goehler J, et al
      . Multimorbidity and care for hypertension, diabetes and HIV among older adults in rural south africa. Bull World Health Organ 2019;97:10–23.:10. doi:10.2471/BLT.18.217000
      OpenUrlCrossRefPubMed
    45. ↵
      1. Chidumwa G,
      2. Maposa I,
      3. Corso B, et al
      . Identifying co-occurrence and clustering of chronic diseases using latent class analysis: cross-sectional findings from sage South Africa wave 2. BMJ Open 2021;11:e041604. doi:10.1136/bmjopen-2020-041604
    46. ↵
      1. Air EL,
      2. Kissela BM
      . Diabetes, the metabolic syndrome, and ischemic stroke: epidemiology and possible mechanisms. Diabetes Care 2007;30:3131–40. doi:10.2337/dc06-1537
      OpenUrlFREE Full Text
    47. ↵
      1. Li X,
      2. Li X,
      3. Lin H, et al
      . Metabolic syndrome and stroke: a meta-analysis of prospective cohort studies. J Clin Neurosci 2017;40:34–8. doi:10.1016/j.jocn.2017.01.018
      OpenUrlPubMed
    48. ↵
      1. Towfighi A,
      2. Ovbiagele B
      . Metabolic syndrome and stroke. Curr Diab Rep 2008;8:37–41. doi:10.1007/s11892-008-0008-z
      OpenUrlCrossRefPubMed
    49. ↵
      1. Arenillas JF,
      2. Moro MA,
      3. Dávalos A
      . The metabolic syndrome and stroke: potential treatment approaches. Stroke 2007;38:2196–203. doi:10.1161/STROKEAHA.106.480004
      OpenUrlFREE Full Text
    50. ↵
      1. Haregu TN,
      2. Oti S,
      3. Egondi T, et al
      . Co-Occurrence of behavioral risk factors of common non-communicable diseases among urban slum dwellers in Nairobi, Kenya. Glob Health Action 2015;8:28697. doi:10.3402/gha.v8.28697
    51. ↵
      1. Bonevski B,
      2. Regan T,
      3. Paul C, et al
      . Associations between alcohol, smoking, socioeconomic status and comorbidities: evidence from the 45 and up study. Drug Alcohol Rev 2014;33:169–76. doi:10.1111/dar.12104
      OpenUrlCrossRefPubMedWeb of Science
    52. ↵
      1. De Leon J,
      2. Rendon DM,
      3. Baca-Garcia E, et al
      . Association between smoking and alcohol use in the general population: stable and unstable odds ratios across two years in two different countries. Alcohol Alcohol 2007;42:252–7. doi:10.1093/alcalc/agm029
      OpenUrlCrossRefPubMedWeb of Science
    53. ↵
      1. Miranda JJ,
      2. Barrientos-Gutiérrez T,
      3. Corvalan C, et al
      . Understanding the rise of cardiometabolic diseases in low- and middle-income countries. Nat Med 2019;25:1667–79. doi:10.1038/s41591-019-0644-7
      OpenUrlCrossRefPubMed
    54. ↵
      1. Abassi MM,
      2. Sassi S,
      3. El Ati J, et al
      . Gender inequalities in diet quality and their socioeconomic patterning in a nutrition transition context in the middle east and north africa: a cross-sectional study in tunisia. Nutr J 2019;18:18. doi:10.1186/s12937-019-0442-6
    55. ↵
      1. Mielke GI,
      2. Brown WJ
      . Physical activity and the prevention of chronic illness in the BRICS nations: issues relating to gender equality. J Sport Health Sci 2019;8:507–8. doi:10.1016/j.jshs.2019.08.001
      OpenUrl
    56. ↵
      1. Dabelea D,
      2. Hamman RF
      . Elevated cardiometabolic risk profile among young adults with diabetes: need for action. Diabetes Care 2019;42:1845–6. doi:10.2337/dci19-0032
      OpenUrlFREE Full Text
    57. ↵
      1. Kamkuemah M,
      2. Gausi B,
      3. Oni T
      . Missed opportunities for ncd multimorbidity prevention in adolescents and youth living with HIV in urban south africa. BMC Public Health 2020;20:821. doi:10.1186/s12889-020-08921-0
    58. ↵
      1. Thienemann F,
      2. Ntusi NAB,
      3. Battegay E, et al
      . Multimorbidity and cardiovascular disease: a perspective on low- and middle-income countries. Cardiovasc Diagn Ther 2020;10:376–85.:376. doi:10.21037/cdt.2019.09.09
      OpenUrlCrossRefPubMed
    59. ↵
      1. Hoddinott J,
      2. Behrman JR,
      3. Maluccio JA, et al
      . Adult consequences of growth failure in early childhood. Am J Clin Nutr 2013;98:1170–8. doi:10.3945/ajcn.113.064584
      OpenUrlAbstract/FREE Full Text
    60. ↵
      1. Martorell R
      . Improved nutrition in the first 1000 days and adult human capital and health. Am J Hum Biol 2017;29:e22952. doi:10.1002/ajhb.22952
    61. ↵
      1. Victora CG,
      2. Adair L,
      3. Fall C, et al
      . Maternal and child undernutrition: consequences for adult health and human capital. Lancet 2008;371:340–57. doi:10.1016/S0140-6736(07)61692-4
      OpenUrlCrossRefPubMedWeb of Science
    62. ↵
      1. Chowdhury MAB,
      2. Uddin MJ,
      3. Khan HMR, et al
      . Type 2 diabetes and its correlates among adults in bangladesh: a population based study. BMC Public Health 2015;15:1070. doi:10.1186/s12889-015-2413-y
    63. ↵
      1. Mkuu RS,
      2. Gilreath TD,
      3. Barry AE, et al
      . Identifying individuals with multiple non-communicable disease risk factors in kenya: a latent class analysis. Public Health 2021;198:180–6. doi:10.1016/j.puhe.2021.07.031
      OpenUrlPubMed
    64. ↵
      1. Olack B,
      2. Wabwire-Mangen F,
      3. Smeeth L, et al
      . Risk factors of hypertension among adults aged 35-64 years living in an urban slum nairobi, kenya. BMC Public Health 2015;15:1251. doi:10.1186/s12889-015-2610-8
    65. ↵
      1. Ahmed SM,
      2. Hadi A,
      3. Razzaque A, et al
      . Clustering of chronic non-communicable disease risk factors among selected Asian populations: levels and determinants. Glob Health Action 2009;2. doi:10.3402/gha.v2i0.1986
    66. ↵
      1. Minh HV, et al
      . Risk factors for chronic disease among rural vietnamese adults and the association of these factors with sociodemographic variables: findings from the WHO STEPS survey in rural vietnam, 2005. 2007.
    67. ↵
      1. Battista F,
      2. Ermolao A,
      3. van Baak MA, et al
      . Effect of exercise on cardiometabolic health of adults with overweight or obesity: focus on blood pressure, insulin resistance, and intrahepatic fat-A systematic review and meta-analysis. Obes Rev 2021;22 Suppl 4(Suppl 4):e13269. doi:10.1111/obr.13269
    68. ↵
      1. Young DR,
      2. Coleman KJ,
      3. Ngor E, et al
      . Associations between physical activity and cardiometabolic risk factors assessed in a southern California health care system, 2010-2012. Prev Chronic Dis 2014;11:E219. doi:10.5888/pcd11.140196
    69. ↵
      1. Yu S,
      2. Xing L,
      3. Du Z, et al
      . Prevalence of obesity and associated risk factors and cardiometabolic comorbidities in rural northeast China. Biomed Res Int 2019;2019:6509083. doi:10.1155/2019/6509083
    70. ↵
      1. Cornelissen VA,
      2. Fagard RH
      . Effects of endurance training on blood pressure, blood pressure-regulating mechanisms, and cardiovascular risk factors. Hypertension 2005;46:667–75. doi:10.1161/01.HYP.0000184225.05629.51
      OpenUrlCrossRefPubMed
    71. ↵
      1. Kodama S,
      2. Tanaka S,
      3. Saito K, et al
      . Effect of aerobic exercise training on serum levels of high-density lipoprotein cholesterol: a meta-analysis. Arch Intern Med 2007;167:999–1008. doi:10.1001/archinte.167.10.999
      OpenUrlCrossRefPubMedWeb of Science
    72. ↵
      1. Hamer M
      . The relative influences of fitness and fatness on inflammatory factors. Prev Med 2007;44:3–11. doi:10.1016/j.ypmed.2006.09.005
      OpenUrlCrossRefPubMedWeb of Science
    73. ↵
      1. Kasapis C,
      2. Thompson PD
      . The effects of physical activity on serum C-reactive protein and inflammatory markers: a systematic review. J Am Coll Cardiol 2005;45:1563–9. doi:10.1016/j.jacc.2004.12.077
      OpenUrlFREE Full Text
    74. ↵
      1. Sewpaul R,
      2. Mbewu AD,
      3. Fagbamigbe AF, et al
      . Prevalence of multimorbidity of cardiometabolic conditions and associated risk factors in a population-based sample of south africans: a cross-sectional study. Public Health Pract (Oxf) 2021;2:100193. doi:10.1016/j.puhip.2021.100193
      OpenUrl
    75. ↵
      1. Diederichs C,
      2. Berger K,
      3. Bartels DB
      . The measurement of multiple chronic diseases -- a systematic review on existing multimorbidity indices. J Gerontol A Biol Sci Med Sci 2011;66:301–11. doi:10.1093/gerona/glq208
      OpenUrlCrossRefPubMedWeb of Science
    76. ↵
      1. World Health Organisation (WHO)
      . NCD microdata repository. 2021. Available: https://extranet.who.int/ncdsmicrodata/index.php/catalog/STEPS

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Twitter @Peter_Otienoh, @Wekesah, @sanyarichard

    • Contributors PO conceptualised the study, reviewed literature and analysed the data. GA, FW, CW, RES, WW and CA made substantive contributions to the conceptualisation of the study, and data analysis and reviewed the manuscript. All authors read and approved the final manuscript. PO takes full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Map disclaimer The inclusion of any map (including the depiction of any boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.