WHAT IS ALREADY KNOWN ON THIS TOPIC
The rising burden of non-communicable diseases linked to high body mass causes a large social and economic burden.
While several international studies have estimated the health and economic impact of obesity and overweight, most of these studies focused on high-income countries.
WHAT THIS STUDY ADDS
Halting the rise of overweight and obesity not only improves health outcomes but also yields healthcare costs savings and productivity gains, which are about eight times higher than the direct healthcare cost savings realised.
By 2044, US$755 million in body mass-related healthcare costs could be saved and ~US$5.8 billion in productivity gain could be realised.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
These findings support the case for government action to prevent further increases in overweight and obesity.
Introduction
Globally, the rising trend in mean body mass index (BMI) 1 2 has resulted in a substantial increase of the burden of non-communicable disease (NCD) attributable to high BMI. 3 4 This has created a double burden of disease in many low -income and middle -income countries (LMICs) that still battle with malnutrition and face a large burden of infectious disease. 3 5 6 In Kenya, high body mass is ranked seventh among the top 10 risk factors contributing to total disability-adjusted life years (DALYs) in 2019 for all ages combined. 5 Of the adult population in Kenya, 27% have overweight or obesity (~6.8 million people), 7 overweight defined as a BMI of 25.0–29.9 kg/m 2 and obesity as ≥30.0 kg/m 2 . 8 This rising BMI-related burden of NCDs causes a huge social and economic burden in the country.
Overweight is considered a consequence of sustained energy imbalance that results from excessive energy intake (diet) and/or insufficient energy expenditure (physical activity). 9 In sub-Saharan Africa countries such as Kenya, a nutrition transition, complemented by increased sedentary lifestyles, has been considered as a driver for the obesity epidemic. 10 11 People eat progressively less locally produced indigenous foods rich in fibre and low in fat, and more of products high in saturated fats and sugar, and industrially processed foods which have become increasingly available. 12 13 Rapid urbanisation, increased sedentary lifestyles and emergence of less energy-demanding jobs 12 have led to physical inactivity which is associated with an increased prevalence of overweight and obesity. 14 15
Prioritisation of NCD prevention in Kenya is an aspiration. 16 The country has set out various national strategies for the prevention and control of NCDs and adopted the Global Action Plan target to halt the rise (0% increase) of overweight and obesity by 2025. 17 , 20 Although achieving the obesity reduction target is challenging, population-wide strategies that target the modifiable components of energy intake and energy expenditure may achieve great overall health impact. 21 22
Several international studies (most conducted in high-income countries) have estimated the economic impact of obesity and overweight by either estimating the current or future cost of overweight and obesity 23 , 28 or by quantifying the potential healthcare cost savings that could be realised from reduction of overweight and obesity. 29 , 31 Various studies have estimated the impact of reduction of overweight and obesity on productivity, 32 , 37 capturing cost from a societal perspective that considers indirect costs. 38 39 Studies that estimated both direct and indirect costs of obesity found productivity-related costs higher than the direct healthcare costs. 40 So far, no research has been published on the assessment of the potential productivity gains and healthcare cost savings associated with prevention and control of overweight and obesity in adults in Kenya.
In this study, we use a proportional multistate life table model 41 42 to quantify the potential impact of achieving the Kenya national target of halting the rise (0% increase at all ages) of overweight and obesity by 2025. Quantifying these benefits may give impetus for government and development partners to prioritise the prevention of overweight and obesity.
Methods
Study design
This study is an extension of previous work where we developed the Kenya Obesity Model , 41 an adaptation of the proportional multistate life table (pMSLT) model 42 developed by Veerman et al . 43 The pMSLT is a macrosimulation model with state-transition (Markov) property that simulates multiple age cohorts of the population into the future. The model allows for the inclusion of multiple diseases simultaneously while accounting for comorbidity. 42 We modelled a total of 37 diseases associated with high BMI ( online supplemental SF table 1 ). 5 41 44 We explain the multistate life table modelling process in the online supplemental SF pages 3–5 and online supplemental SF figure 1 .
Modelled scenario
In this assessment, the model was extended to model the difference between health and economic outcomes for a reference population with ‘business as usual’ where the current BMI levels and trend continue unabated for 25 years, and a comparator (an identical ‘intervention’ population) in which ‘business as usual’ is maintained until the rising trend is halted in the year 2025 (the target is to halt the rise by 2025).
BMI distribution and disease epidemiology
As described in our previous work, 41 current BMI levels are derived from the latest published national survey results ( online supplemental SF table 2 ). 7 45 We derived the BMI trend from age-specific and sex-specific mean BMI data for Kenya from the NCD Risk Factor Collaboration (NCD-RisC) study from 1975 to 2016 2 ( online supplemental SF pages 6 and 7 , online supplemental SF figures 2 and 3 ). Our business-as-usual scenario (base case) assumes that the observed rising BMI trend continues up to 25 years into the future.
To populate our model, we used the following Kenya-specific data from the Global Burden of Disease (GBD) 2019 study: age-specific, sex-specific and cause-specific incidence, prevalence, and case fatality rates ( online supplemental SF table 3 ), all-cause mortality and disability rates ( online supplemental SF tables 4 and 5 ) and the 2019 population data ( online supplemental SF table 6 ). 5 46 We used DisMod V.II software to enforce internal consistency in the epidemiological estimates and derive remission and case fatality parameters not provided in the GBD data ( online supplemental SF pages 7 and 8 , online supplemental SF figure 4 ). 47 We calculated disability weights using disease-specific prevalence and years lived with disability (YLDs) estimates from the 2019 GBD study ( online supplemental SF page 30 , online supplemental SF table 7 ). 5 41 48 Based on the latest evidence of diseases associated with high BMI, we modelled 37 diseases and used respective relative risk estimates by age and sex as reported in the GBD 2019 study ( online supplemental SF table 8 ). 5 41 44
Life table analysis
We modelled the entire 2019 population in Kenya (50.2 million, online supplemental SF table 6 ) over their lifetime with risks rising only from age 20 and no burden among children. This is in line with the available evidence which shows high BMI (>22.5 kg/m 2 ) is associated with NCDs that are commonly seen in adults (apart from asthma). 44 Figure 1 gives a schematic overview of this study. Halting the rise in BMI translates to fewer people who have overweight and obesity in Kenya since we model a closed cohort of the 2019 population, incorporating effects of population ageing but do not incorporate new birth cohorts or immigration. This impacts on incident cases of the modelled diseases. For each age-sex group, the proportional changes in incidence after a change in BMI were calculated using the potential impact fraction (see also online supplemental SF pages 4 and 5 ), 49 using 22.5 kg/m 2 as our optimal BMI level (theoretical minimum risk exposure level). 41 Changes in incidence resulted in changes in prevalence, and ultimately disease-specific mortality followed. Changes in disease-related quality of life at every age were calculated using disease-specific disability weights. 5 48 The new disease-specific mortality and morbidity rates and cost offsets are fed into a life table. The changes in the morbidity and mortality are summarised as health-adjusted life years (HALYs) and life expectancy (HALE), and the change in health expenditure is assessed. Finally, the changes in the HALYs affect productivity. In this study, we took a societal perspective that allows for the assessment of indirect effects on the wider economy that are valued by the rest of the society, not the effects of inability to work on the well-being of the individual which may be incorporated in the measurement of quality of life. 38 39
In the section below, we describe how we extended the Kenya Obesity Model 41 to incorporate healthcare costs and productivity outcomes.
Healthcare cost data and estimating healthcare savings
Disease-specific costs in Kenya
We identified estimates of disease-specific costs from five published studies. 50 , 53 We present the cost estimates in table 1 alongside additional notes on how we calculated the annual cost per incident or prevalent case of the disease. Two studies reported only point/mean estimates for disease costs. 50 53 One study defined uncertainty boundaries by applying an arbitrary ±15% value to the central value 52 and one study provided the mean, median and SD. 51 Therefore, we assumed the SE of the mean to be 20% of the published point estimates of costs for each disease. The ±20% is often used in the literature to estimate the lower and upper limits of costs. In addition, for best alignment with the total healthcare costs estimate for Kenya, we used only the disease costs estimates from public facilities (not private hospitals) reported in the published studies. Using Kenyan data on the percentage of people who had some sickness reported but did not seek healthcare and percentage of the population requiring admission who did not get admitted, we adjusted the disease costs to account for the percentage of people with disease who are likely to seek medical care in Kenya and hence incur healthcare costs. The estimates of these two percentages were taken from the 2013 Kenya Household Health Expenditure survey report. 54
Table 1
Per capita annual disease costs in US$ derived from identified primary studies *
* All disease costs in the identified studies were hospital based (with some being tertiary hospitals) † Only estimated (CKD) cases on dialysis and transplant were costed in our model. Of the 9·.1% global CKD prevalence, dialysis accounted for 0·.041% (0·.037 to 0·.044), and kidney transplantation for 0·.011% (0·.010 to 0·.012). 65 ‡ We considered the acute presentations of IHD as acute myocardial infarction, angina and heart failure (secondary to hypertension). 50 To determine the number of people diagnosed with IHD that had experienced each of the above events (ie, myocardial infarction, angina and heart failure [(secondary to hypertension])), we used proportions identified in a study by Subramanian and colleagues 71 based on previous research. 72 73 ALLacute lymphoid leukaemiaAMLacute myeloid leukaemiaCKDchronic kidney diseaseCLLchronic lymphoid leukaemiaCMLchronic myeloid leukaemiaGBDGlobal Burden of DiseaseIHDischaemic heart disease
| Disease modelled | Incident or prevalent case per year | Cost type/notes | Percentage of patients | Point estimates of cost per incident or prevalent case per year | Year of costing | Source of data/primary study | Country of the primary study |
| Asthma | Incident cost | Diagnosis | 4.23 | 2017 | Subramanian et al 50 | Kenya | |
| Prevalent cost | Mild asthma (management (mgt)/treatment (rx)) | 0.95 | 67.93 | ||||
| Severe asthma (mgt/rx) | 0.05 | 146.74 | |||||
| Weighted average used as prevalent cost | 71.87 | ||||||
| Breast cancer | Incident cost | Breast cancer (clinical breast examination)(screening) | 3.90 | 2017 | Subramanian et al 50 | Kenya | |
| Diagnosis | 401.00 | ||||||
| Total cost used as incident cost | 404.90 | ||||||
| Prevalent cost | Breast cancer treatment | ||||||
| Stage I | 0.07 | 1340.38 | 2017 | Subramanian et al 50 | Kenya | ||
| Stage II | 0.35 | 1340.38 | |||||
| Stage III (curative approach) | 0.19 | 1542.58 | |||||
| Stage III (palliative approach) and stage IV | 0.40 | 675.35 | |||||
| Weighted average used as prevalent cost | 1126.19 | ||||||
| CKD † | Prevalent cost | CKD (dialysis) | 0.70 | 5338.00 | 2017 | Subramanian et al 50 | Kenya |
| CKD (transplant) | 0.30 | 9237.00 | |||||
| Weighted average used as prevalent cost | 6507.70 | ||||||
| CLL | Prevalent cost | Average cost of treatment | 1164.20 | 2016 | Atieno et al 53 | Kenya | |
| CML | Average cost of treatment | 686.00 | |||||
| ALL | Prevalent cost | We used total number of prevalent CLL and CML cases in Kenya from GBD to calculate a weighted average cost which was applied as the prevalent cost for ALL, AML and other leukaemia | 1027.85 | ||||
| AML | |||||||
| Other leukaemia | |||||||
| Multiple myeloma | Prevalent cost | Average cost of treatment | 1473.10 | 2016 | Atieno et al 53 | Kenya | |
| Colon and rectum cancer | Prevalent cost | Average cost of treatment | 1742.10 | ||||
| Diabetes mellitus type 2 | Incident cost | Diabetes—random blood sugar (screening) | 4.95 | 2017 | Subramanian et al 50 | Kenya | |
| Diagnosis | 41.95 | ||||||
| Total cost used as incident cost | 46.90 | ||||||
| Prevalent cost | Insulin only (mgt/rx) | 0.32 | 186.40 | ||||
| Oral medication only (mgt/rx) | 0.25 | 88.61 | |||||
| Both insulin and oral medication (mgt/rx) | 0.43 | 234.44 | |||||
| Weighted average used as prevalent cost | 182.61 | ||||||
| Gallbladder and biliary tract cancer | Prevalent cost | Average cost of treatment | 407.90 | 2016 | Atieno et al 53 | Kenya | |
| Hypertensive heart disease | Prevalent cost | Annual treatment cost for hypertension | |||||
| Treatment—1 drug | 0.20 | 25.64 | 2017 | Subramanian et al 50 | Kenya | ||
| Treatment—2 drug | 0.35 | 67.25 | |||||
| Treatment—3 drug | 0.25 | 81.20 | |||||
| Treatment—4 drug | 0.10 | 110.33 | |||||
| Treatment—resistant | 0.10 | 159.36 | |||||
| Weighted average used as prevalent cost | 75.93 | ||||||
| Intracerebral haemorrhage | Prevalent cost | Used the cost of prevalent haemorrhagic stroke | 1406.30 | 2013–2017 | Aminde et al 51 | Cameroon | |
| Subarachnoid haemorrhage | Prevalent cost | Used the cost of prevalent haemorrhagic stroke | 1406.30 | 2013–2017 | Aminde et al 51 | Cameroon | |
| IHD ‡ | Incident cost | Acute myocardial infarction | 0.2F, 0.35M | 1995.65 | 2017 | Subramanian et al 50 | Kenya |
| Angina | 0.35 | 1,236.81 | |||||
| Heart failure (secondary to hypertension) | 0.10 | 1026.07 | |||||
| Weighted average calculated using the given event distribution as cited by authors | 1633.10 | ||||||
| Prevalent cost | All CHD states (chronic annual cost for secondary prevention) | 300.00 | 2012 | Gazianio et al 52 | South Africa | ||
| Ischaemic stroke | Incident cost | Stroke cost per inpatient or outpatient episode | 1873.93 | 2017 | Subramanian et al 50 | Kenya | |
| Prevalent cost | Stroke (chronic annual cost for secondary prevention) | 900.00 | 2012 | Gazianio et al 52 | South Africa | ||
| Kidney cancer | Prevalent cost | Average cost of treatment for renal carcinoma | 671.40 | 2016 | Atieno et al 53 | Kenya | |
| Liver cancer | Prevalent cost | Average cost of treatment for liver cancer | 1303.40 | ||||
| Oesophageal cancer | Prevalent cost | Average cost of treatment | 866.80 | ||||
| Ovarian cancer | Prevalent cost | Average cost of treatment | 2217.00 | ||||
| Pancreatic cancer | Prevalent cost | Average cost of treatment | 1244.00 | ||||
| Thyroid cancer | Prevalent cost | Average cost of treatment for follicular thyroid carcinoma | 1342.70 | ||||
| Uterine cancer | Prevalent cost | Average cost of treatment for endometrial carcinoma | 1624.40 |
Total healthcare costs
We used latest estimates from the 2020 WHO Global Health Expenditure Database that estimated the total healthcare costs for Kenya as US$4.5 billion for the year 2018. 55 Although we did not use cost estimates from the 2013 Household Health Expenditure survey, the survey report provided data on health-seeking behaviour, use of healthcare services and the corresponding health expenditures (by age groups and sex), out-of-pocket health spending and health insurance coverage. 54 In terms of utilisation of healthcare services by health facility type/ownership, the survey results 54 corresponded well to the characteristics of the health system in Kenya as documented in the Kenya National Health Accounts. 56 The health accounts indicate that government contributes to 42.8% of healthcare financing scheme followed by household out-of-pocket (26.1%), donors (17.9%) and voluntary healthcare payment systems at 13.2%. 56 The country’s current health expenditure is 4.59% of the gross domestic product and 11% of the country’s total budget. 55 To apportion the total healthcare costs across specific age groups and sex, we used proportional cost distribution informed by the age-specific annual per capita health spending and sex-specific spending on healthcare published in the 2013 Kenya Household Health Expenditure and Utilisation Survey ( table 2 ). 54 To derive the yearly costs of all other diseases per person, we subtracted the age-specific and sex-specific costs of all modelled diseases from the total health expenditure in the respective sex and age groups ( table 2 ). Overall healthcare costs for all other health conditions are included because as preventive interventions prolong life, additional health expenditure is expected in those added years of life. 57 58
Table 2
2019 healthcare costs in US$
Total healthcare costs for Kenya were derived from the 2020 WHO Global Health Expenditure Database that published 2018 costs. 55 We inflated the reported 2018 total healthcare costs to 2019 (our model base year) based on Kenya consumer price indices. 61 To convert the reported US dollars (US$) to Kenya shillings, the reader to use the world Bank’s official 2019 exchange rate of 102. 74
| Age group | 2019 total healthcare costs in US$ | 2019 costs of all other diseases not included in the model | ||||
| Costs in US$ | Costs per person in US$ | Costs per person in US$ | ||||
| Male | Female | Male | Female | Male | Female | |
| 0–4 | 211 311 453 | 297 171 638 | 65 | 94 | 61 | 89 |
| 5–14 | 100 841 169 | 141 815 008 | 16 | 23 | 10 | 16 |
| 15–24 | 120 840 100 | 169 939 915 | 22 | 31 | 15 | 22 |
| 25–34 | 130 257 586 | 183 183 919 | 34 | 46 | 22 | 31 |
| 35–44 | 171 101 963 | 240 624 205 | 63 | 86 | 39 | 61 |
| 45–54 | 262 208 203 | 368 748 782 | 153 | 218 | 101 | 169 |
| 55–64 | 404 422 822 | 568 748 122 | 401 | 557 | 290 | 465 |
| 65+ | 578 064 332 | 812 943 744 | 846 | 985 | 622 | 802 |
| Total | 1 979 047 628 | 2 783 175 333 |
The change in high-BMI-related NCDs healthcare expenditure was based on the predicted reduction in disease mortality and morbidity following achievement of the target. Cost data are often skewed and higher costs concentrated among a small group of ‘high-cost’ patients. 59 However, at the population level modelled in this study, we assumed that at the aggregate level, the cost data averages will be approximately normally distributed—in line with the central limit theorem. 60 Therefore, we modelled healthcare costs as normally distributed assuming the SE of the mean to be 20% of the point estimates. We extracted the costs in US$ as published in the primary studies and then inflated all cost estimates to 2019 (our model base year) based on Kenya consumer price indices. 61
Estimating productivity gains
We estimated productivity outcomes for the 2019 working population in Kenya (20 years up to retirement age of 65 years) over their lifetime. We used the Human Capital Approach 39 to estimate productivity gains resulting from a reduction in high BMI-related (1) mortality, (2) mortality and morbidity (combined), and (3) morbidity. First, to estimate productivity gains resulting from a reduction in high BMI-related mortality, we use the methodology previously used in Nomaguchi et al . 32 Productivity gains are estimated from the age of premature death until the age of traditional retirement in Kenya (65 years). In Equation 1, we describe how we quantified productivity changes. People obtain additional life years (LYs) due to halting the rise of overweight and obesity with the potential to earn higher income as follows:
where pM is productivity changes due to reduced high BMI-related mortality, ( L'x-Lx)i is the difference in years lived between employed population post ‘intervention,’ that is, after halting the rise in overweight and obesity ( L'x ) and the employed population pre ‘intervention’ ( Lx), and WA is the average annual wage rate in the working population in Kenya (US$655.05). 62 In the model, we apply an adjustment for the percentage of the 2019 working population that is employed in Kenya (58.4%). 62 Due to limitations in the data availability on sex-age-specific average wage rate in Kenya, we apply one estimate across all age groups and sex. While this limits factual accuracy, it is recommended for promotion of equity in economic assessments. 39
Second, to estimate productivity gains resulting from a reduction in high BMI-related mortality and morbidity (combined), we use estimated HALYs gained by the working population in Kenya (age 20–65). We used the disability weight as a proxy for work-ability. Disability weights indicate the average loss of quality of life due to disease. 48 In Equation 2, people obtain additional HALYs due to halting the rise of overweight and obesity with the potential to earn higher income as follows:
where pM is productivity changes due to reduced high BMI-related mortality and morbidity (combined), ( Lw'x-Lwx)i is the difference in health-adjusted years lived between employed population post ‘intervention,’ that is, after halting the rise in overweight and obesity ( Lw'x ) and the employed population pre ‘intervention’ ( Lwx ), and WA is the average annual wage rate in the working population in Kenya. Third, we estimate productivity gains resulting from a reduction in high BMI-related morbidity. We use the same equation as above but in place of Lw'x-Lwx (difference in health-adjusted years lived) by the working population in Kenya, we use the high BMI-related morbidity calculated as the difference between HALYs gained and added years lived by employed populations, that is, (( Lw'x-Lwx )-( L'x-Lx )).
In our main analysis, we applied a 3% discount rate for health and economic (productivity and healthcare costs) outcomes as recommended in Drummond et al . 39
Uncertainty analyses
The simultaneous and combined effect of the uncertainty in model inputs on our outcomes was quantified using Ersatz software. 63 We performed a Monte Carlo simulation with bootstrapping (2000 iterations) while incorporating probabilistic uncertainty from model inputs (BMI, relative risks, costs). For all outputs, the 95% uncertainty intervals (UIs) were calculated. These are 2.5 and 97.5 percentiles capturing sampling error with input data.
Sensitivity analyses
In one-way sensitivity analyses, we examined the implications of alternative discount rates (0% and 5%).
We used Microsoft Excel for Microsoft V.365 and two software add-ins: the EpiGearXL V.5.0 add-in for calculation of the potential impact fraction. 63 We used the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist to guide our reporting. 64 Ethics approval was not required for this analysis.
Patient and public involvement
Patients or the public were not involved in the design, or conduct, or reporting or dissemination plans of our research. However, this study was carried out as part of a larger stakeholder engaged modelling study that has ethical approval from the Griffith University Human Research Ethics Committee (GU Ref No: 2019/707) and the Kenyatta National Hospital/University of Nairobi Ethics & Research Committee (P81/02/2021).
Results
In this paper, although we modelled the 2019 population in Kenya over their lifetime, as appropriate, we present results for various policy relevant periods, that is, year 2030, over the first 25 years (year 2044), and lifetime estimates.
Health outcomes
To illustrate the modelled scenarios, figure 2 presents a graph for two age groups (20–24 and 45–49 years) showing the business as usual for base case (trend going up for 25 years into the future) and the intervention case (trend stable from the year 2025).
Figure 3 shows the changes in incidence, prevalence and mortality results for the modelled diseases over the first 25 years by disease group. By disease group, up to the year 2044, the leading contributors of the health gains were musculoskeletal diseases (537 052 new cases avoided; 95% UI 444 361–639 460), T2DM (466 030; 95% UI 360 103–597 378) and cardiovascular diseases (301 729; 95% UI 248 559–367 949). In the year 2044, the largest proportional reduction in disease prevalent cases would be seen in type 2 diabetes followed by musculoskeletal diseases and cardiovascular diseases. The cumulative numbers of new cases averted by 2044 (incidence), in 2044 (prevalence) and corresponding proportional changes are presented in online supplemental SF tables 9 and 10 ). By 2044, deaths averted would be highest for cardiovascular diseases, type 2 diabetes and cancers ( figure 3 , online supplemental SF table 11 ).
The change in health outcomes translates to approximately 7756 HALYs (95% UI 6978–8608) saved by the year 2030 and 449 276 HALYs (95% UI 398 202–5 06 152) by the year 2044. Over the lifetime of the 2019 Kenyan population, approximately 6.8 million HALYs (95% UI 5.8–7.9 million) (135 HALYs per 1000 persons) or 4.8 million LYs (95% UI 3.8–6.1 million) could be saved. For these time periods, greater gains would be seen in females than males for both HALYs and LYs ( table 3 ).
Table 3
Effects of the achievement of the national target (halt the rise of overweight and obesity by 2025 in Kenya)
We modelled the entire 2019 population in Kenya with risks rising only from age 20 and no burden among children. Productivity gains estimates: gains resulting from a reduction in high BMI-related morbidity, mortality and related mortality and morbidity (combined). A 3% discount rate was applied for health and economic outcomes. * This is the overall mortality number from life table where we count both the reduction in mortality from BMI-related diseases and the increase in mortality due to other causes. BMI, body mass index; HALYs, health-adjusted life years; LYs, life years; UI, uncertainty interval; US$, US dollars
| Variable | Female, mean | Male, mean | Total, mean |
| (95% UI) | (95% UI) | (95% UI) | |
| HALYs gained over the lifetime of the 2019 population | 4 769 376 | 1 995 590 | 6 764 966 |
| (3 884 260–5 8 67 936) | (1 678 158–2 337 407) | (5 819 478–7 911 381) | |
| LYs gained over the lifetime of the 2019 population | 3 517 443 | 1 306 714 | 4 824 156 |
| (2 597 441– 4 790 994) | (1 037 816–1 600 505) | (3 824 980– 6 149 268) | |
| HALYs gained over different time periods | |||
| By 2030 | 5134 | 2622 | 7756 |
| (4425– 5879) | (2269–2996) | (6978–8608) | |
| By 2044 | 300 991 | 148 285 | 449 276 |
| (254 362–352 204) | (127 121 –170 805) | (398 202–506 152) | |
| Deaths avoided over different time periods * | |||
| By 2030 | 884 | 380 | 1265 |
| (62–1217) | (303–467) | (983–1607) | |
| By 2044 | 46 292 | 19 843 | 66 135 |
| (34 099–62 317) | (15 699–24 144) | (52 764–83 073) | |
| Healthcare cost savings over the lifetime of the 2019 population (in million US$) | 1672 | 1424 | 3096 |
| (464–2777) | (1057–1834) | (1753–4292) | |
| Healthcare cost savings by the year 2044 (in million US$) | 480 | 275 | 755 |
| (384–594) | (226–333) | (636–894) | |
| Cumulative productivity gains by year 2044 (in million US$) | |||
| Productivity gains (morbidity) | 3162 | 1721 | 4883 |
| (2651–3782) | (1416–2057) | (4277–5594) | |
| Productivity gains (mortality) | 983 | 524 | 1507 |
| (722–1304) | (395–667) | (1209–1865) | |
| Productivity gains (mortality and morbidity) | 3742 | 2031 | 5773 |
| (3104–4537) | (1659–2431) | (5012–6650) |
Healthcare cost savings
Halting the rise of overweight and obesity could save US$755 million in overweight and obesity-related healthcare costs by 2044 ( table 3 ). By 2044, cardiovascular diseases contributed the most to the healthcare cost savings ( figure 4 ). Over the lifetime of the 2019 population, the reduction in the healthcare costs (savings) is estimated at US$3096 million (US$62 per capita).
Productivity gains
By the year 2044, the total productivity gains resulting from a reduction in high BMI-related morbidity in the 2019 working-age population aged 20–65 years was estimated to be US$4883 million (95% UI 4277–5594) ( table 3 ). Productivity gains resulting from high BMI-related mortality were US$1507 million (95% UI 1209–1865), while gains from high BMI-related mortality and morbidity (combined) were US$5773 million (95% UI 5012–6650), ~US$237 per capita.
Sensitivity analyses
Compared with the base case scenario (3% discount rate), applying a 5% discount reduced the lifetime HALYs gained by ~42% ( table 4 ). When no discounting was applied, the HALYs gained were 4.5 times greater than those reported in base case (3% discounting). This relatively large impact is because the health gains materialise over time in future. Applying a 5% discount reduced the healthcare cost savings by about 59% when compared with the base case scenario of 3% discount rate. When no discounting was applied the savings were 1.5 times greater than when 3% discount was applied. In comparison to HALYs, the impact of discounting on healthcare cost savings was modest. Into the future, the cost savings are outweighed by additional healthcare costs as more people are alive resulting from prevented mortality from high BMI-related diseases. See online supplemental SF figure 5 for healthcare cost savings over different time periods with varying discount rates. Across all productivity gain estimates, no discounting increased the estimates by about 80% while a 5% discount rate resulted to a 30% reduction in all three productivity gain estimates ( figure 5 ).
Table 4
Effects of the achievement of the national target with different discount rates applied
Total represents values for both male and female. National target scenario modelled was halting the rise of overweight and obesity by 2025 in Kenya. We modelled the entire 2019 population in Kenya with risks rising only from age 20 and no burden among children. Diseases not costed: low back pain, osteoarthritis hip, osteoarthritis knee, gout, Alzheimer’s disease and other dementias, cataract, gallbladder and biliary diseases and atrial fibrillation and flutter (AFF). AFFatrial fibrillation and flutterBMIbody mass indexHALYshealth-adjusted life yearsLYslife yearsUIuncertainty intervalUS$US dollars
| Modelled scenarios | 0% discount rate | 3% discount rate(base case) | 5% discount rate |
| Variable | Total, mean | Total, mean | Total, mean |
| (95% UI) | (95% UI) | (95% UI) | |
| HALYs gained over the lifetime of the 2019 population | 30 513 170(25 976 578–36 240 896) | 6 764 966(5 819 478–7 911 381) | 2 853 103(2 452 791–3 311 707) |
| HALYs gained by 2030 | 10 259(9204–11 348) | 7756(6978–8608) | 6477(5817–7185) |
| HALYs gained by 2044 | 812 755(718 261–915 870) | 449 276(398 202–506 152) | 308 062(272 307– 347 585) |
| Healthcare cost savings over the lifetime of the 2019 population (in million US$) | 4857(−2027 to 10 321) | 3096(1753–4292) | 1830(1283–2361) |
| Healthcare cost savings by the year 2044 (in million US$) | 1340(1133–1581) | 755(636–894) | 522(442–612) |
Discussion
Our results show that halting the rise of overweight and obesity in the year 2025 may save 6.8 million HALYs (95% UI 5.8–7.9 million) over the lifetime of the 2019 Kenyan population (135 HALYs per 1000 persons). By the year 2044, 449 276 HALYs (95% UI 398 202–506 152) may be saved (9 HALYs per 1000 persons) with the leading contributors to the health gains (new cases avoided) being musculoskeletal diseases, followed by T2DM and cardiovascular diseases. The health impact is slow to eventuate. This is because BMI influences incidence, decreasing prevalence in years/decades after, with mortality modelled as a function of prevalence. Therefore, the added years of life can be decades into the future. The potential health gains are consistent with our previous study where we estimated the avoidable disease burden if exposure to high BMI was eliminated in Kenya. 41 However, the results are different in magnitude because in this study we assess a different scenario (ie, achieving target—halting the rise of overweight and obesity in the year 2025). Both studies showed greater health gains would be seen in females than males. This is largely due to the baseline distribution of overweight and obesity where a greater percentage of women have overweight or obesity in Kenya. 7 45
We found that a total of US$755 million in high BMI-related healthcare costs could be saved by 2044. (For context, this translates to 16% of Kenya’s annual healthcare expenditure or 1% of gross domestic product.) By 2044, the gains in productivity from high BMI-related mortality and morbidity (combined) (~US$5.8 billion) were about eight times higher than the direct healthcare cost savings realised in the same period of time.
Our findings are in line with previous studies that have explored the economic impact of overweight and obesity. 23 , 2933 35 37 However, no study is directly comparable. This is particularly because studies use different methodologies to estimate direct and indirect costs. 37 Total population size and population size by age groups vary across countries. Additionally, the prevalence of overweight and obesity is country-specific, and studies assess different interventions which have different impacts on overweight and obesity prevalence. Time periods for these assessments vary between studies. Moreover, studies may differ in whether they apply future BMI trends in their assessments or not.
A South African study estimated that the total cost of overweight and obesity for the year 2020 was ZAR33 194 million (~US$1826 million) which represented 15.4% of government health expenditure and was equivalent to 0.67% of the country’s gross domestic product. 23 A recent study on Brazil estimated that annually, US$654 million direct healthcare costs related to NCDs were attributable to overweight and obesity. 27 Studies that have assessed the impact on productivity are from high-income countries where authors have assessed the impact of BMI reducing interventions. For example, in a New Zealand study, the total healthcare costs attributable to overweight and obesity amounted to NZ$686 million (~US$425 million), 4.4% of New Zealand’s total healthcare expenditure in 2006 and substantial costs of lost productivity were reported. 34 The obesity and overweight-related costs were largest for type 2 diabetes (38%), followed by hypertension (27%). 34 In an Australian study, 20% tax on sugar-sweetened beverages was found to reduce the number of people with obesity which translated into substantial reductions in healthcare costs and productivity gains in both the paid and unpaid sectors of the economy. 32 In these two studies, 32 34 the productivity changes are estimated using both the Human Capital Approach and Friction Cost Approach. 39 In our study, we used the Human Capital Approach as it better reflects our societal perspective where we estimate changes in productivity using the gross earnings of those in employment. 39 In another Australian study, the authors found that the productivity gains associated with a 10% tax on unhealthy foods (an obesity prevention policy) over the period from 2003 to 2030 accounted for almost twice the value of the estimated savings to the healthcare system. 36 Our findings also showed that productivity gains were greater than healthcare cost savings with potential productivity gains resulting from a reduction in high BMI-related mortality and morbidity (combined) being 7.6 times higher than the direct healthcare costs savings by the year 2044. In sum, our findings corroborate previous work and show that reduction of overweight and obesity would yield substantial healthcare cost savings and productivity gains in not only high-income countries but also in LMICs.
Strengths and limitations of our study
We used an established proportional multistate life table model ( Kenya Obesity Model ) 41 , 43 to assess the potential impact of achieving global and national target of halting the rise (0% increase) of overweight and obesity by 2025. A detailed report on the strengths and limitations of the Kenya Obesity Model is provided in our previous work. 41 We include a summary of this in online supplemental SF section 4 .
There were specific strengths and limitations related to the healthcare costs and productivity modelling. A major strength is that we capture costs from a societal perspective by including productivity in our impact assessment. When compared against the health sector perspective, a societal perspective provides a more comprehensive estimate of cost savings related to reduction of overweight and obesity. Further, we estimate productivity gains resulting from a reduction in high BMI-related mortality, related mortality and morbidity (combined), and related morbidity.
Cost estimates were drawn from Kenyan studies 50 53 for all costed diseases apart from four cardiovascular diseases where estimates were drawn from Cameroon and South Africa. 51 52 This ensured that cost data were specific to Kenya and similar settings. A limitation is that all disease costs were hospital based (with some being tertiary hospitals) which reflect cost of treatment for advanced disease cases on referral, hence costs may be high. The case definitions used by GBD are broader where prevalent numbers, for instance, may include people not aware they have the condition. One study included some estimates based on recommended care in treatment guidelines. 50 Recommended care may differ from actual care given to patients. We took several measures to reflect plausible cost estimates. This included using published costs estimates from public facilities as opposed to private facilities and adjusting the disease costs to account for the percentage of people unwell who did not seek care hence did not incur healthcare costs. Also, we costed only CKD cases on dialysis and transplant in our model. No costs were attributed to the vast majority of cases in earlier disease stages (1, 2 and 3), which confer little to no functional health loss 65 and hence may not have formed part of the hospital-based CKD costs reported in the Kenya studies. We did not identify any literature with costs for low back pain, osteoarthritis hip, osteoarthritis knee, gout, Alzheimer’s disease and other dementias, cataract, gallbladder and biliary diseases and atrial fibrillation and flutter (AFF). The exclusion of the costs of these diseases leads to underestimation of the cost savings reported in this study particularly for the musculoskeletal diseases that were the leading contributors of the potential health gains for new cases avoided. Our work highlights the need for costing studies in Kenya and similar settings.
Comorbidity increases disease-specific costs. 51 66 This means that though the envelope total healthcare costs for Kenya was captured in the model, the disease-specific costs for both modelled and non-modelled diseases may be higher. However, this is unlikely since our cost data was sourced from primary costing studies that determined the average annual disease costs established from study samples, likely capturing those with comorbidity and those without. Studies on disease costs with comorbidity present and data on number of people with comorbidity in the population would offer additional insights in the cost of NCDs. On balance, our study is likely to have underestimated healthcare cost savings resulting from the prevention of high BMI.
Regarding the employed population, the Kenya economic survey computes the average annual wage rate considering people employed as comprising of wage employees, self-employed and unpaid family workers and informal sector. 62 This covers a broad scope of those from 16 years of age who can legally engage in paid work to those beyond 65 years of age in some sectors. In our case, for the productivity-related outputs we modelled the 2019 working population in Kenya aged 20–65 years. This may mean that some additional gains in productivity could be realised from prevented obesity-related diseases for those older than 65 who are still productive. Although our model did not incorporate projected trends in the percentage of people employed in Kenya, we assessed the available data and established a similar percentage of people employed (58%) in 2013 report and computation from the 2021 economic survey report. 54 62 We also did not model projected trends for income levels due to changes in trends seen in available data that showed increases in annual wage rate between 2016 and 2019 (~5% annual increase) and a 4% reduction in the year 2020. 62 If the increasing trend in annual wage rate before COVID resumes, our current findings are an underestimation of the potential productivity gains that could result from halting the rise of overweight and obesity in Kenya.
Meaning of the study/implication for policy makers
To our knowledge, this is the first study to assess the potential productivity gains and healthcare cost savings associated with prevention and control of overweight and obesity in adults in Kenya. Data from future national surveys in Kenya will inform the progress made so far towards achieving the national target of halting the rise (0% increase at all ages) of overweight and obesity by 2025. Quantifying these benefits at this stage may give impetus for government and development partners to prioritise the prevention of overweight and obesity. Our findings are important to development partners, governments and stakeholders across multiple sectors as we show that prevention of overweight and obesity is a means of increasing economic productivity. In Kenya, the healthcare cost savings realised from the reduction of overweight and obesity would also directly benefit households. This is because household out-of-pocket payments currently contribute to approximately 28% of current health expenditure funds in the country. 56 The findings also may provide evidence that could help generate increased demand by civil societies and the general public for government policies and interventions that reduce the prevalence of obesity. In Kenya, most of the government NCD policy actions are still at the development stage. 67 These include policy options that address the obesogenic environment, that is, the surroundings, opportunities or conditions of life whose influences promote obesity in individuals or populations. 68 For example, creation of avenues for increased physical activity such as urban planning that allows for active transport, fiscal and regulatory interventions that ensure provision of healthy and nutritious foods to all in the population, creation of healthy settings, health and nutritional education specifically targeted to the factors that influence BMI in the country. Stakeholders engaged in our larger obesity study identified and ranked a total of 24 broad strategies for the prevention of overweight and obesity in Kenya that may inform future policies in Kenya and similar settings. 69
Future research
Subject to data availability, future modelling studies could estimate the potential impact of attainment of the national obesity reduction target on healthcare costs and productivity for each of the 47 counties in Kenya and by socioeconomic factors such as education level, wealth quintiles, urban versus rural residence. Assessment of cost-effectiveness of interventions that reduce the prevalence of overweight and obesity is also key. For instance, in our recent work, we assessed the potential impact of four selected food policy interventions for the prevention and control of overweight and obesity in Kenya. 70 Two specific policy interventions (a 20% tax on sugar sweetened beverages and mandatory kilojoule menu labelling) were assessed for cost-effectiveness and found dominant (health promoting and cost saving) offering potential policy options towards reduction of overweight and obesity. Such research supports evidence-based selection and priority setting for preventive strategies and intersectoral actions in efforts to achieve the set obesity reduction target.
Conclusion
Achieving the set overweight and obesity reduction target could improve health outcomes and also yield healthcare cost savings and productivity gains. Our estimates add to available evidence that may motivate for the prioritisation of the prevention of overweight and obesity in Kenya and similar settings that are battling with the rapid rise in mean BMI and related NCDs, alongside persisting communicable disease burden.
supplementary material
Acknowledgements
The STEPS survey microdata was supplied to us by WHO NCD microdata repository on request. We acknowledge that we received stakeholder buy-in for us to model and study high BMI as a leading risk factor in Kenya. The stakeholders engaged were individuals who are known to participate in health policymaking at the national level. A description of the stakeholders and engagement process is given in 2 of our manuscripts published earlier.
Footnotes
Funding: No funding was received for this study. MNW was supported by the Griffith University Postgraduate Research Scholarships.
Provenance and peer review: Not commissioned; externally peer reviewed.
Patient consent for publication: Not applicable.
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.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplemental information.
References
1. NCD Risk Factor Collaboration (NCD-RisC) Trends in adult body-mass index in 200 countries from 1975 to 2014: a pooled analysis of 1698 population-based measurement studies with 19.2 million participants. Lancet. 2016;387:1377–96. doi: 10.1016/S0140-6736(16)30054-X.
2. NCD Risk Factor Collaboration (NCD-RisC) Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet. 2017;390:2627–42. doi: 10.1016/S0140-6736(17)32129-3.
3. Dai H, Alsalhe TA, Chalghaf N, et al. The global burden of disease attributable to high body mass index in 195 countries and territories, 1990–2017: an analysis of the global burden of disease study. PLOS Med. 2020;17:e1003198. doi: 10.1371/journal.pmed.1003198.
4. GBD 2019 Risk Factors Collaborators Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2020;396:1223–49. doi: 10.1016/S0140-6736(20)30752-2.
5. Global Burden of Disease Collaborative Network . Seattle, United States of America: Institute for Health Metrics and Evaluation (IHME); 2020. Global burden of disease study 2019 (GBD 2019) results.
6. Miranda JJ, Kinra S, Casas JP, et al. Non-communicable diseases in low- and middle-income countries: context, determinants and health policy. Trop Med Int Health. 2008;13:1225–34. doi: 10.1111/j.1365-3156.2008.02116.x.
7. Ministry of Health Republic of Kenya . Ministry of Health Division of Non Communicable Diseases, Kenya National Bureau of Statistics, World Health Organization; 2015. Kenya stepwise survey for non-communicable diseases risk factors 2015 report.
8. World Health Organization . Geneva,Switzerland: WHO; 2014. Global status report on noncommunicable diseases 2014 “Attaining the nine global noncommunicable diseases targets; a shared responsibility”.
9. Romieu I, Dossus L, Barquera S, et al. Energy balance and obesity: what are the main drivers. Cancer Causes Control. 2017;28:247–58. doi: 10.1007/s10552-017-0869-z.
10. Mbanya JC, Assah FK, Saji J, et al. Obesity and type 2 diabetes in sub-Sahara Africa. Curr Diab Rep. 2014;14:501. doi: 10.1007/s11892-014-0501-5.
11. Popkin BM, Adair LS, Ng SW. Global nutrition transition and the pandemic of obesity in developing countries. Nutr Rev. 2012;70:3–21. doi: 10.1111/j.1753-4887.2011.00456.x.
12. Adeboye B, Bermano G, Rolland C. Obesity and its health impact in Africa: a systematic review. Cardiovasc J Afr. 2012;23:512–21. doi: 10.5830/CVJA-2012-040.
13. Yusuf S, Reddy S, Ounpuu S, et al. Global burden of cardiovascular diseases: part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization. Circulation. 2001;104:2746–53. doi: 10.1161/hc4601.099487.
14. Skidmore PML, Yarnell JWG. The obesity epidemic: prospects for prevention. QJM. 2004;97:817–25. doi: 10.1093/qjmed/hch136.
15. Lee I-M, Shiroma EJ, Lobelo F, et al. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380:219–29. doi: 10.1016/S0140-6736(12)61031-9.
16. Wanjau MN, Kivuti-Bitok LW, Aminde LN, et al. Stakeholder perceptions of current practices and challenges in priority setting for non-communicable disease control in Kenya: a qualitative study. BMJ Open. 2021;11:e043641. doi: 10.1136/bmjopen-2020-043641.
17. World Health Organisation Global action plan for the prevention and control of NCDs 2013-2020. 2013 https://www.who.int/nmh/events/ncd_action_plan/en/ Available.
18. Ministry of Health Kenya . Government of Kenya: Nairobi, Kenya; 2021. National non-communicable diseases strategic plan 2021/22-2025/26, Department of non-communicable diseases, editor.
19. Ministry of Health Republic of Kenya . Government of Kenya; 2015. Kenya national strategy for the prevention and control of non-communicable diseases 2015-2020 division of non-communicable diseases, editor. www.health.go.ke Available.
20. Ministry of Health Republic of Kenya . Republic of Kenya; 2014. Kenya health policy 2014–2030: towards attaining the highest standard of health, Ministry of health, editor. http://www.health.go.ke Available.
21. Stephens SK, Cobiac LJ, Veerman JL. Improving diet and physical activity to reduce population prevalence of overweight and obesity: an overview of current evidence. Prev Med. 2014;62:167–78. doi: 10.1016/j.ypmed.2014.02.008.
22. Swinburn BA, Sacks G, Hall KD, et al. The global obesity pandemic: shaped by global drivers and local environments. Lancet. 2011;378:804–14. doi: 10.1016/S0140-6736(11)60813-1.
23. Boachie MK, Thsehla E, Immurana M, et al. Estimating the healthcare cost of overweight and obesity in South Africa. Glob Health Action. 2022;15:2045092. doi: 10.1080/16549716.2022.2045092.
24. Wang YC, McPherson K, Marsh T, et al. Health and economic burden of the projected obesity trends in the USA and the UK. Lancet. 2011;378:815–25. doi: 10.1016/S0140-6736(11)60814-3.
25. Schneider P. The World Bank; 2020. Health and economic impacts of overweight/obesity, in obesity: health and economic consequences of an impending global challenge; pp. 69–94.
26. de Oliveira ML, Santos LMP, da Silva EN. Direct healthcare cost of obesity in Brazil: an application of the cost-of-illness method from the perspective of the public health system in 2011. PLoS One. 2015;10:e0121160. doi: 10.1371/journal.pone.0121160.
27. Ferrari G, Giannichi B, Resende B, et al. The economic burden of overweight and obesity in Brazil: perspectives for the Brazilian unified health system. Public Health. 2022;207:82–7. doi: 10.1016/j.puhe.2022.03.015.
28. Lartey S, Si L, Lung T, et al. Impact of overweight and obesity on life expectancy, quality-adjusted life years and lifetime costs in the adult population of Ghana. BMJ Glob Health. 2020;5:e003332. doi: 10.1136/bmjgh-2020-003332.
29. Okunogbe A, Nugent R, Spencer G, et al. Economic impacts of overweight and obesity: current and future estimates for eight countries. BMJ Glob Health. 2021;6:e006351. doi: 10.1136/bmjgh-2021-006351.
30. Harrison S, Dixon P, Jones HE, et al. Long-term cost-effectiveness of interventions for obesity: a mendelian randomisation study. PLOS Med. 2021;18:e1003725. doi: 10.1371/journal.pmed.1003725.
31. Ananthapavan J, Sacks G, Brown V, et al. Priority-setting for obesity prevention—the assessing cost-effectiveness of obesity prevention policies in Australia (ACE-obesity policy) study. PLOS ONE. 2020;15:e0234804. doi: 10.1371/journal.pone.0234804.
32. Nomaguchi T, Cunich M, Zapata-Diomedi B, et al. The impact on productivity of a hypothetical tax on sugar-sweetened beverages. Health Policy. 2017;121:715–25. doi: 10.1016/j.healthpol.2017.04.001.
33. Menon K, de Courten B, Ademi Z, et al. Estimating the benefits of obesity prevention on productivity: an Australian perspective. Int J Obes. 2022;46:1463–9. doi: 10.1038/s41366-022-01133-z.
34. Lal A, Moodie M, Ashton T, et al. Health care and lost productivity costs of overweight and obesity in New Zealand. Aust N Z J Public Health. 2012;36:550–6. doi: 10.1111/j.1753-6405.2012.00931.x.
35. Araujo MYC, Norberto MCCS, Mantovani AM, et al. Obesity increases costs with productivity loss due to disability retirements, independent of physical activity: a cohort study. J Occup Environ Med. 2020;62:325–30. doi: 10.1097/JOM.0000000000001808.
36. Carter HE, Schofield DJ, Shrestha R, et al. The productivity gains associated with a junk food tax and their impact on cost-effectiveness. PLoS One. 2019;14:e0220209. doi: 10.1371/journal.pone.0220209.
37. Goettler A, Grosse A, Sonntag D. Productivity loss due to overweight and obesity: a systematic review of indirect costs. BMJ Open. 2017;7:e014632. doi: 10.1136/bmjopen-2016-014632.
38. Koopmanschap MA, Rutten FFH. A practical guide for calculating indirect costs of disease. Pharmacoeconomics. 1996;10:460–6. doi: 10.2165/00019053-199610050-00003.
39. Drummond MF. Methods for the economic evaluation of health care programmes. Fourth ed. New York, United States of America: Oxford University Press; 2015. p. 43.
40. Dee A, Kearns K, O’Neill C, et al. The direct and indirect costs of both overweight and obesity: a systematic review. BMC Res Notes. 2014;7:1–9. doi: 10.1186/1756-0500-7-242.
41. Wanjau MN, Aminde LN, Veerman JL. The Avoidable disease burden associated with overweight and obesity in Kenya: a modelling study. EClinicalMedicine. 2022;50:101522. doi: 10.1016/j.eclinm.2022.101522.
42. Barendregt JJ, Van Oortmarssen GJ, Van Hout BA, et al. Coping with multiple morbidity in a life table. Math Popul Stud. 1998;7:29–49. doi: 10.1080/08898489809525445.
43. Veerman JL, Sacks G, Antonopoulos N, et al. The impact of a tax on sugar-sweetened beverages on health and health care costs: a modelling study. PLOS ONE. 2016;11:e0151460. doi: 10.1371/journal.pone.0151460.
44. Global Burden of Disease Collaborative Network . Seattle, United States of America: Institute for Health Metrics and Evaluation (IHME); 2020. Global burden of disease study 2019 (GBD 2019) relative risks.
45. World Health Organization NCD Microdata repository: 2015 Kenya stepwise approach to surveillance of NCD risk factors (STEPS) survey. 2018
46. Global Burden of Disease Collaborative Network . Seattle, United States of America: Institute for Health Metrics and Evaluation (IHME); 2020. Global burden of disease study 2019 (GBD 2019) population estimates 1950-2019.
47. Barendregt JJ, Van Oortmarssen GJ, Vos T, et al. A generic model for the assessment of disease epidemiology: the computational basis of Dismod II. Popul Health Metr. 2003;1:4. doi: 10.1186/1478-7954-1-4.
48. Salomon JA, Haagsma JA, Davis A, et al. Disability weights for the global burden of disease 2013 study. Lancet Glob Health. 2015;3:e712–23. doi: 10.1016/S2214-109X(15)00069-8.
49. Barendregt JJ, Veerman JL. Categorical versus continuous risk factors and the calculation of potential impact fractions. J Epidemiol Community Health. 2010;64:209–12. doi: 10.1136/jech.2009.090274.
50. Subramanian S, Gakunga R, Kibachio J, et al. Cost and affordability of non-communicable disease screening, diagnosis and treatment in Kenya: patient payments in the private and public sectors. PLOS ONE. 2018;13:e0190113. doi: 10.1371/journal.pone.0190113.
51. Aminde LN, Dzudie A, Mapoure YN, et al. Estimation and determinants of direct medical costs of ischaemic heart disease, stroke and hypertensive heart disease: evidence from two major hospitals in Cameroon. BMC Health Serv Res. 2021;21:140. doi: 10.1186/s12913-021-06146-4.
52. Gaziano TA, Bertram M, Tollman SM, et al. Hypertension education and adherence in South Africa: a cost-effectiveness analysis of community health workers. BMC Public Health. 2014;14:240. doi: 10.1186/1471-2458-14-240.
53. Atieno OM, Opanga S, Martin A, et al. Pilot study assessing the direct medical cost of treating patients with cancer in Kenya; findings and implications for the future. J Med Econ. 2018;21:878–87. doi: 10.1080/13696998.2018.1484372.
54. Ministry of Health . 2013 Kenya household health expenditure and utilisation survey. Nairobi, Kenya: Government of Kenya; 2014.
55. World Health Organization The global health expenditure database (GHED) 2020. https://apps.who.int/nha/database/Home/Index/en Available.
56. Ministry of Health Republic of Kenya . Kenya national health accounts 2015/2016, Ministry of Health Policy and Planning, Editor. Government of Kenya: Nairobi, Kenya; 2019.
57. van Baal PHM, Feenstra TL, Polder JJ, et al. Economic evaluation and the postponement of health care costs. Health Econ. 2011;20:432–45. doi: 10.1002/hec.1599.
58. van Baal P, Meltzer D, Brouwer W. Future costs, fixed healthcare budgets, and the decision rules of cost-effectiveness analysis. Health Econ. 2016;25:237–48. doi: 10.1002/hec.3138.
59. Wammes JJG, van der Wees PJ, Tanke MAC, et al. Systematic review of high-cost patients’ characteristics and healthcare utilisation. BMJ Open. 2018;8:e023113. doi: 10.1136/bmjopen-2018-023113.
60. Kwak SG, Kim JH. Central limit theorem: the cornerstone of modern statistics. Korean J Anesthesiol. 2017;70:144–56. doi: 10.4097/kjae.2017.70.2.144.
61. Kenya National Bureau of Statistics 2019 consumer price indices and inflation rates. 2022. https://www.knbs.or.ke/ Available.
62. Kenya National Bureau of Statistics . Nairobi, Kenya: 2021. Economic survey 2021. https://www.knbs.or.ke/ Available.
63. Barendregt JJ. Epigear international Pty Ltd. 2016. https://www.epigear.com/index.htm Available.
64. Husereau D, Drummond M, Augustovski F, et al. Consolidated health economic evaluation reporting standards 2022 (CHEERS 2022) statement: updated reporting guidance for health economic evaluations. BMC Med. 2022;20:23. doi: 10.1186/s12916-021-02204-0.
65. GBD Chronic Kidney Disease Collaboration Global, regional, and national burden of chronic kidney disease, 1990–2017: a systematic analysis for the global burden of disease study 2017. Lancet. 2020;395:709–33. doi: 10.1016/S0140-6736(20)30045-3.
66. Cortaredona S, Ventelou B. The extra cost of comorbidity: multiple illnesses and the economic burden of non-communicable diseases. BMC Med. 2017;15:216. doi: 10.1186/s12916-017-0978-2.
67. Asiki G, Wanjohi MN, Barnes A, et al. Benchmarking food environment policies for the prevention of diet-related non-communicable diseases in Kenya: national expert panel’s assessment and priority recommendations. PLoS One. 2020;15:e0236699. doi: 10.1371/journal.pone.0236699.
68. Swinburn B, Egger G, Raza F. Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Prev Med. 1999;29:563–70. doi: 10.1006/pmed.1999.0585.
69. Wanjau MN, Kivuti-Bitok LW, Aminde LN, et al. Stakeholder-engaged research: strategies for the prevention and control of overweight and obesity in Kenya. BMC Public Health. 2021;21:1622. doi: 10.1186/s12889-021-11649-0.
70. Wanjau MN, Kivuti-Bitok LW, Aminde LN, et al. The health and economic impact and cost effectiveness of interventions for the prevention and control of overweight and obesity in Kenya: a stakeholder engaged modelling study. Cost Eff Resour Alloc. 2023;21:69. doi: 10.1186/s12962-023-00467-3.
71. Subramanian S, Hilscher R, Gakunga R, et al. Cost-effectiveness of risk stratified medication management for reducing premature cardiovascular mortality in Kenya. PLOS ONE. 2019;14:e0218256. doi: 10.1371/journal.pone.0218256.
72. White AD, Folsom AR, Chambless LE, et al. Community surveillance of coronary heart disease in the Atherosclerosis risk in communities (ARIC) study: methods and initial two years' experience. J Clin Epidemiol. 1996;49:223–33. doi: 10.1016/0895-4356(95)00041-0.
73. Perman G, Rossi E, Waisman GD, et al. Cost-effectiveness of a hypertension management programme in an elderly population: a Markov model. Cost Eff Resour Alloc. 2011;9:4. doi: 10.1186/1478-7547-9-4.
74. The World Bank Group Official exchange rate (local currency unit per US dollar, period average. 2022. https://data.worldbank.org/indicator/PA.NUS.FCRF?locations=KE Available.
Associated Data
Supplementary Materials
Data Availability Statement
All data relevant to the study are included in the article or uploaded as supplemental information.