Abstract
Aims
Obesity and its complications contribute to the burden of cardiovascular disease (CVD). Here, we characterised individuals with high body mass index (BMI) and established CVD by assessing healthcare resource utilisation (HCRU) and costs, incidence of cardiovascular (CV) events and mortality.
Materials and Methods
This was a retrospective open cohort study using UK Discover data (study period: January 2004 to December 2019). Included were individuals aged ≥45 years with BMI ≥ 27 kg/m2, without type 1 or type 2 diabetes, and with established CVD (previous myocardial infarction, stroke or peripheral artery disease). Serial annual cross sections were assembled to generate prevalence and incidence cohorts and for mapping of HCRU, costs and the incidence of selected events. CVD and mortality trajectories were modelled using a Markov model. HCRU and costs were layered onto this model to obtain associated trajectories.
Results
In 2019, annual per‐person healthcare costs for individuals with high BMI and established CVD (n = 27 313) were £3364. During 2015–2019, the incidence of major adverse CV events was 2812 per 100,000 person‐years; the incidences of all‐cause and CV mortality were 2896 and 774 per 100,000 person‐years, respectively. Over 2022–2031, this population is projected to accrue estimated healthcare costs of £40.8 million. HCRU trajectory drivers included a history of CV events, older age, and multimorbidity.
Conclusions
Owing to a high disease and treatment burden, people with a history of CVD living with high BMI incur substantial healthcare costs and are at risk of mortality.
Keywords: cardiovascular disease; cohort study, obesity care, observational study, real‐world evidence
1. INTRODUCTION
Obesity (body mass index [BMI] ≥ 30 kg/m2) and its complications, such as dyslipidaemia, hypertension and type 2 diabetes, are linked to an elevated risk of cardiovascular disease (CVD). 1 People with obesity experience cardiovascular (CV) events at an earlier age, live with CVD for a greater proportion of their lifetime, and generally live shorter lives than individuals with a healthy weight. 2 Accordingly, CVD is the leading cause of death related to high BMI globally. 3
Obesity also increases the likelihood and frequency of healthcare resource utilisation (HCRU). Evidence from US observational studies has shown that people with obesity have higher rates of inpatient admissions 4 and longer hospital stays 5 than people with a healthy weight. Obesity‐related complications, including CVD, are a key contributor to this HCRU. 6 , 7 UK data also link HCRU and costs to BMI and CV risk. In a sample of 1.6 million individuals from the UK Clinical Practice Research Datalink, the number of general practitioner contacts and prescriptions rose with both increasing BMI category and increasing CV risk. 8 In this study, individuals with CVD had the greatest HCRU, followed by those with high CVD risk (Framingham Risk Score >20%). 8 Importantly, individuals with obesity class III (BMI ≥ 40 kg/m2) and CVD had the highest HCRU and costs. 8 Despite established links between high BMI and CVD, there is a lack of UK‐specific data on the patient and clinical burden on individuals living with any class of obesity or overweight in combination with CVD and the impact on healthcare systems, particularly in terms of projecting future costs.
This study aims to assess HCRU, healthcare cost profiles, and the incidence of CV events and mortality in individuals with high BMI and established CVD in the UK using real‐world data. Furthermore, we seek to project disease prevalence, future HCRU trajectories and associated costs over the period from 2022 to 2031, providing insights into the potential impact on healthcare systems and informing policy development and healthcare resource allocation.
2. MATERIALS AND METHODS
2.1. Patient and public involvement
There was no patient and public involvement in the conduct of this study.
2.2. Study design
This was a retrospective open cohort study using data from the Discover database, an administrative real‐world data set of linked primary and secondary electronic health records covering more than 2.7 million people residing in North West London, UK. 9 Key features of the database, including scope, architecture and governance, have been described elsewhere. 10 Because this study was a retrospective analysis of secondary de‐identified data and was not considered to be human subject research, ethical approval was not required.
There were two distinct analytical phases. First, HCRU, healthcare costs, incidence of CV events and mortality were estimated in individuals from the Discover database (see Section 2.4). Then, healthcare costs were projected over 2021–2031 for the study population using Markov models (see Section 2.5).
The study period was 1 January 2004–31 December 2019. Individuals entered the population at the time they matched all required criteria (the index date) and were followed until the month of death, transfer out date or end of the study period, whichever occurred earliest. Serial annual cross sections with calendar year as the unit of investigation over the study period were assembled to generate prevalence and incidence cohorts and for the mapping of HCRU, direct healthcare costs and the incidence of selected events.
2.3. Patient population
Individuals aged 18 years or older with known sex were assessed for inclusion; individuals aged 45 years or older with a BMI ≥ 27 kg/m2 and established CVD were included in this analysis. Established CVD was defined as at least one of previous myocardial infarction (MI), previous stroke or peripheral artery disease (PAD; defined as peripheral arterial revascularisation procedure or amputation due to atherosclerotic CVD). To assess HCRU and healthcare costs without the confounding effects of diabetes‐related events and costs, individuals who had glycated haemoglobin (HbA1c) ≥ 48 mmol/mol, had received glucose‐lowering medication in the 90 days before the index or had a history of type 1 or type 2 diabetes were excluded. Full details are reported in the supplemental methods and Table S1.
2.4. Outcomes
The primary outcomes were HCRU and healthcare costs, comprising interactions and costs for primary care (general practitioner or nurse visits), secondary care (inpatient and outpatient admissions, and emergency care) and prescriptions. Secondary outcomes were individual HCRU events and cost components, incidence of CV events (stroke, MI, major adverse CV events [MACE], heart failure [HF] or composite HF) and all‐cause and CV mortality. MACE was defined as a composite of acute MI, stroke or CV death; composite HF was defined as a composite of acute HF, urgent HF visit or CV death.
Outcomes were assessed during the period 2015–2019, and were age‐standardised to the 2013 European Standard Population. 11 Data are presented by calendar year or aggregated across 2015–2019. Costs were expressed as annual means per person during this period and were adjusted to 2019 costs using UK Consumer Price Index inflation data from the Office for National Statistics. Full details are included in the supplemental methods.
2.5. Modelling of HCRU trajectories
CV events and mortality trajectories were modelled using a five‐state Markov survival model framework (Figure S1). Markov models are used in epidemiological and biostatistical studies 12 , 13 , 14 , 15 and are a more complex multi‐outcome extension of a similar single‐outcome competing risks model. 16 Adopting a closed cohort approach, the model included individuals with CV events during the study period who also had available data on HCRU visits and costs (n = 17 115). The HCRU visits and costs data were applied to this model to obtain HCRU trajectories, which were estimated for a 10‐year time horizon (2022–2031). The five states included were MI, stroke, PAD+ (defined as peripheral vascular disease AND peripheral arterial revascularisation procedure OR amputation due to atherosclerotic disease), CVD multimorbidity (≥2 CVDs) and deceased (Figure S2). Because CVD is often progressive, backward transitions were not allowed. The model estimated transition intensities (hazards) for each allowable transition between disease states. It used age at risk as the timescale (time‐inhomogeneous) and incorporated age at risk as a time‐varying covariate and risk factor profile as fixed covariate. The optimal model incorporated 14 types of covariates (see Tables S4 and S5).
The key model outputs were: (1) number of HCRU events and healthcare costs, annually and averaged over the 10‐year time horizon; and (2) variation in HCRU events and costs per patient per year, according to CVD type and risk factors.
Details on comorbidities included as strata in the trajectory analysis are shown in Table S2; full modelling methods are reported in the Supporting Information, Tables S3–S5, and Figures S1–S3.
2.6. Statistical analyses
Descriptive data are presented as means and standard deviation (SD) for continuous variables and as numbers and percentages for categorical variables. Details on how incidence events were captured and how each component of HCRU and associated costs were calculated are provided in the supplemental methods.
3. RESULTS
3.1. Study population and baseline characteristics
This study included 27,313 individuals with high BMI and established CVD. The mean age was 64.7 (SD 12.0), 35.6% were women and the mean BMI was 30.8 kg/m2 (SD 4.3) (Table 1). The majority (61.8%) of included individuals were white. Further patient characteristics are shown in Table 1.
TABLE 1.
Baseline characteristics of individuals with high body mass index (BMI) and established cardiovascular disease (CVD).
| Baseline characteristics | High BMI and established CVD, n = 27 313 |
|---|---|
| Women | 9733 (35.6%) |
| Age, years, mean | 64.7 (SD 12.0) |
| BMI, kg/m2, mean | 30.8 (SD 4.3) |
| SBP, mmHg, mean a | 134.3 (SD 11.1) |
| Unknown | 113 (0.6%) |
| DBP, mmHg, mean a | 78.3 (SD 11.1) |
| Unknown | 114 (0.6%) |
| HbA1c, %, mean a | 6.0 (SD 0.9) |
| Unknown | 9643 (49.2%) |
| HDL, mg/dL, mean a | 1.3 (0.4) |
| Unknown | 1586 (8.1%) |
| LDL, mg/dL, mean a | 2.5 (SD 1.0) |
| Unknown | 2809 (14.3%) |
| Triglycerides, mg/dL, mean a | 138.5 (SD 87.0) |
| Unknown | 2495 (12.7%) |
| eGFR, mL/min/1.73 m2, mean | 71.5 (SD 16.8) |
| Unknown | 4034 (20.6%) |
| Race/ethnicity | |
| Asian or Asian British | 5630 (20.6%) |
| Black or Black British | 1995 (7.3%) |
| Mixed | 539 (2.0%) |
| White | 16 876 (61.8%) |
| Other | 1650 (6.0%) |
| Unknown | 623 (2.3%) |
| Index of multiple deprivation b | |
| Decile 1 | 1406 (5.1%) |
| Decile 2 | 2858 (10.5%) |
| Decile 3 | 4591 (16.8%) |
| Decile 4 | 3857 (14.1%) |
| Decile 5 | 3626 (13.3%) |
| Decile 6 | 3528 (12.9%) |
| Decile 7 | 2450 (9.0%) |
| Decile 8 | 1596 (5.8%) |
| Decile 9 | 1452 (5.3%) |
| Decile 10 | 929 (3.4%) |
| Missing | 1020 (3.7%) |
Abbreviations: DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HbA1c, glycated haemoglobin; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; SBP, systolic blood pressure; SD, standard deviation.
Reported for individuals with an index date between 2004 and 2019 (n = 19 589), due to a high level of missing data before 2004.
Decile 1 corresponds to the 10% of the population living in the most deprived areas, and decile 10 corresponds to the 10% living in the least deprived areas nationally.
3.2. HCRU over time
HCRU increased over time. In 2015, the total number of age‐standardised HCRU events per person per year was 17.5 events (95% confidence interval [95% CI]: 17.4–17.6) (Figure 1), with primary care being the largest contributor to HCRU. By 2019, the number of events increased to 20.8 events (95% CI: 20.7–20.9) (difference 2015–2019: + 3.3 events per person per year; relative increase of 18.9%) (Figure 1).
FIGURE 1.
Total annual per‐person healthcare events and components in 2015 and 2019. Values are shown to two decimal places to avoid identical values. Note that due to rounding, the sum of events may differ from the total displayed in the graph. CI, confidence interval.
3.3. Healthcare costs over time
In line with HCRU, annual per‐person healthcare costs increased gradually over time, from £2218 (95% CI: 2214–2222) in 2015 to £3364 (95% CI: 3361–3368) (+51.7%) in 2019 (Figure 2A). Inpatient admissions were the primary driver of costs, followed by prescriptions and primary care (Figure 2B).
FIGURE 2.
Progression of healthcare costs from 2015 to 2019 (A) and cost components in 2015 and 2019 (B). Note that due to rounding, the sum of cost components may differ from the total costs displayed in the graph. CI, confidence interval.
3.4. Incidence of CV events
During 2015 to 2019, the incidence of MACE in individuals with high BMI and established CVD was 2812 (95% CI: 2680–2944) events per 100,000 person‐years, and the incidence of composite HF was 2798 (95% CI: 2682–2913) events per 100,000 person‐years (Figure 3). The incidence of acute HF, stroke and MI was 1250 (95% CI: 1174–1326), 1148 (95% CI: 1058–1238) and 891 (95% CI: 813–969) events per 100,000 person‐years, respectively. The incidence of CV events increased with age, and there was a trend towards higher rates of stroke, acute and composite HF events, and MACE with increasing BMI (Table S6). The observed trends for sex were variable across outcomes (Table S6).
FIGURE 3.
Mean incidences of cardiovascular (CV) events during 2015–2019. CI, confidence interval; HF, heart failure; MACE, major adverse cardiovascular event; MI, myocardial infarction.
3.5. Incidence of mortality
The incidence of all‐cause mortality and CV mortality from 2015 to 2019 was 2896 (95% CI: 2784–3009) and 774 (95% CI: 716–832) per 100,000 person‐years, respectively (Figure 4). CV‐related mortality contributed 27% of all‐cause mortality. All‐cause and CV mortality increased with increasing age (Table S6). All‐cause mortality was similar for groups with overweight (BMI 25–29.9 kg/m2), obesity class I (30–34.9 kg/m2) and obesity class II (35–39.9 kg/m2) (range: 1811–1884 per 100 000 person‐years) but was considerably higher for groups with obesity class III (BMI ≥ 40 kg/m2) (2448 per 100 000 person‐years) (Table S6).
FIGURE 4.
All‐cause mortality and cardiovascular (CV) mortality from 2015 to 2019. CI, confidence interval.
3.6. Projection of HCRU and associated
In addition to the current HCRU and healthcare costs associated with high BMI and established CVD, a substantial burden on healthcare systems owing to increased HCRU was projected over the next decade. Over 10 years (2022–2031), it was projected that individuals with high BMI and established CVD (n = 17 115; 2015–2019) would accrue healthcare costs of £40.8 million, as a result of an estimated 0.3 million HCRU events during 2022 to 2031, or 0.03 million events per year, experienced by individuals in this population. This corresponds to 20.7 events per patient per year and costs of £2950 per patient per year.
3.7. HCRU trajectory drivers
Projected per‐patient HCRU events and costs, averaged over the 10‐year time horizon, were driven by several factors (Table 2). Age was a major factor: individuals aged 80 years or older versus 45–59 years at index had approximately 140% more events and approximately 250% higher costs. CVD was also a driving factor. Having a history of two or more of MI, stroke or PAD+, versus just one of these CVDs, increased the number of events by up to 41% and costs by up to 109%. Individuals with atrial fibrillation, HF, hypertension or revascularisation at baseline had up to 76% more events and up to 94% higher costs versus those without the condition at baseline; the biggest difference was observed for atrial fibrillation.
TABLE 2.
Healthcare resource utilisation (HCRU) trajectory drivers.
| Trajectory driver | Annual per‐person number of events | Annual per‐person healthcare costs (£) |
|---|---|---|
| CVD state | ||
| MI | 19.8 | 2677 |
| Stroke | 20.9 | 3025 |
| PAD+ | 23.1 | 3409 |
| CVD multi‐morbidity a | 27.9 | 5604 |
| Age category | ||
| 45–49 years | 14.1 | 1690 |
| 50–54 years | 15.1 | 1868 |
| 55–59 years | 16.5 | 2065 |
| 60–64 years | 19.0 | 2568 |
| 65–69 years | 22.2 | 3161 |
| 70–75 years | 25.7 | 3820 |
| 75–79 years | 29.9 | 4709 |
| ≥80 years | 33.5 | 5981 |
| BMI category | ||
| 25–29.9 kg/m2 | 20.9 | 2976 |
| 30–34.9 kg/m2 | 19.9 | 2855 |
| 35–39.9 kg/m2 | 20.8 | 2989 |
| ≥40 kg/m2 | 22.5 | 3438 |
| CKD stage b | ||
| Stage 1 | 18.6 | 2545 |
| Stage 2 | 20.0 | 2760 |
| Stage 3 | 25.9 | 4206 |
| Stages 4 and 5 | 32.1 | 6336 |
| HbA1c category | ||
| Normal (<5.7%) | 20.5 | 2959 |
| Elevated (5.7%–6.49%) | 20.6 | 2947 |
| High (≥6.5%) | 21.6 | 2989 |
| HDL‐C category | ||
| High (≥1.55 mmol/L) | 22.7 | 3430 |
| Reduced (1.03–1.54 mmol/L) | 20.5 | 2914 |
| Low (<1.03 mmol/L) | 19.6 | 2732 |
| SBP category | ||
| Normal (<120 mmHg) | 20.8 | 3009 |
| Elevated (120–139 mmHg) | 20.9 | 2899 |
| High (≥140 mmHg) | 20.3 | 3003 |
| DBP category | ||
| Normal (<80 mmHg) | 23.0 | 3339 |
| Elevated (80–89 mmHg) | 19.5 | 2751 |
| High (≥90 mmHg) | 17.2 | 2403 |
| Number of non‐CVD comorbidities | ||
| 0–2 comorbidities | 17.9 | 2459 |
| ≥3 comorbidities | 29.4 | 4525 |
| Heart failure at baseline | ||
| Absent | 20.1 | 2841 |
| Present | 30.1 | 4964 |
| Atrial fibrillation at baseline | ||
| Absent | 19.4 | 2737 |
| Present | 34.1 | 5303 |
| Revascularisation at baseline | ||
| Absent | 19.7 | 2841 |
| Present | 24.4 | 3404 |
| Hypertension at baseline | ||
| Absent | 18.3 | 2459 |
| Present | 22.5 | 3334 |
Note: CV death, LDL‐C and triglyceride data were modelled but were not key HCRU trajectory drivers.
Abbreviations: BMI, body mass index; CKD, chronic kidney disease; CV, cardiovascular; CVD, cardiovascular disease; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HbA1c, glycated haemoglobin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; LMI, myocardial infarction; PAD+, peripheral vascular disease AND peripheral arterial revascularisation procedure OR amputation due to atherosclerotic disease; SBP, systolic blood pressure.
Defined as ≥2 CVDs.
Defined using medical records and/or eGFR measurements.
Other comorbidities also drove HCRU. Individuals with ≥3 non‐CVD comorbidities versus 0–2 non‐CVD comorbidities at baseline had approximately 60% more events and 80% higher costs, and individuals with chronic kidney disease stage 4–5 versus stage 1 had approximately 70% more events and 150% higher costs.
For the other factors examined (BMI, high‐density lipoprotein cholesterol, systolic blood pressure and HbA1c), projected events and costs were similar across categories.
4. DISCUSSION
Our study used a large, representative administrative data set to explore the impact of established CVD specifically in individuals with high BMI by estimating HCRU and costs. Our results indicate that individuals with high BMI and established CVD incur substantial HCRU and healthcare costs, which is aligned with findings from a previous study in the UK. 8 By examining future trajectories, this study provides valuable insights for healthcare policymakers, enabling them to anticipate and manage the long‐term burden of obesity‐related CVD on the healthcare system. To our knowledge, our study is the first to investigate CV event rates and mortality in this at‐risk group. Our results suggest that this group offers the potential for targeted intervention and management to reduce their risk of further complications, improve longevity and reduce healthcare costs.
Weight loss via a range of methods could improve CV outcomes: care pathways for overweight and obesity include bariatric surgery 17 , 18 , 19 , 20 ; intensive lifestyle intervention or diabetes support and education 21 ; and pharmacotherapy. 22 To limit CV events and early mortality in high burden–high risk groups, refinement of these care pathways is needed, which requires the use of an objective approach to detect target groups in need of weight management. In our study, we identified factors associated with high risk of CVD incidence, mortality and increased HCRU. We found that most of the modelled demographic and physiological risk factors in our study had large effects, independent of the other modelled risk factors. HCRU events and costs were driven by age, disease profile (CVD and non‐CVD) and chronic kidney disease stage. The costs associated with some drivers, such as chronic kidney disease and atrial fibrillation, are well documented in the literature, 23 , 24 , 25 but often not in the context of overweight and obesity. Weight management could be incorporated into the standard of care for patients with CVD; this has been acknowledged in a 2021 scientific statement from the American Heart Association, which concluded that with increasing prevalence of obesity in populations with a longer lifespan, there is a need to improve the management of patients with obesity and CVD. 1 It should be noted that there are further risk factors for CVD not taken into account in this analysis, such as lifestyle and socioeconomic status, which could confound HCRU and costs.
Only approximately 13% of healthcare services in the UK are provided outside the National Health Service. 26 It was therefore possible to obtain a relatively comprehensive view of UK patient characteristics, HCRU and healthcare costs using the Discover data set. Linking records from primary and secondary care and the specific costs associated with each allowed for accurate costing estimates both within and between the populations. Longitudinal data are particularly valuable for capturing real‐world HCRU trends because they allow patient journeys to be tracked across different healthcare settings. The open cohort study design, which allowed individuals to enter the study population at any point during the study period, resulted in a relatively large study population and reduced the risk of selection bias. Importantly, this study aimed to provide a descriptive epidemiological analysis from a patient and healthcare perspective; data were not adjusted for potential confounders but instead are provided to convey patient burden and HCRU. Our projections did not take into account a variety of factors that may affect healthcare costs in the future, such as inflation, technological advances or policy changes. Further analyses comparing groups with different overweight/obesity and CVD status will be valuable to go beyond the current analysis. Data more recent than 2019 are available in the Discover data set; however, due to the disruption of healthcare systems during the COVID‐19 pandemic in early 2020, the analysis period was limited to pre‐pandemic years. Studies investigating the variation in unmet need in people with high BMI and established CVD during and after the COVID‐19 pandemic and how this differs from pre‐pandemic trends would form the basis of valuable further research. Longitudinal studies that examine the long‐term effects of weight management interventions on HCRU and mortality in a population with high BMI and established CVD are another avenue for future research. Cost‐effectiveness analyses in subpopulations of people living with obesity, such as those with CVD, may also be valuable.
5. CONCLUSION
In this UK database study, we estimated that individuals with high BMI and established CVD experience high HCRU and associated healthcare costs, and poor health outcomes, as indicated by the incidence of CV events, including MACE, and all‐cause and CV mortality. Using an innovative modelling approach, we characterized a group at risk of CV events that could potentially benefit from weight management to help prevent the progression of obesity and its complications. Actionable targets could include prioritization of weight management programmes for individuals with high BMI and established CVD and public health initiatives focusing on early identification and intervention for individuals at risk of obesity and CVD. These approaches could avert the poorer health outcomes and higher healthcare costs associated with CVD in those living with high BMI.
AUTHOR CONTRIBUTIONS
MSC, SH and AT directly accessed and verified the underlying data reported in the manuscript. JP‐S, MSC, SH and AT analysed the data. JP‐S, MSC, SH, KSM, AT and SC had access to all data and contributed to the study concept and design. All authors contributed to the drafting and critical revision of the manuscript text and approved the manuscript for submission.
CONFLICT OF INTEREST STATEMENT
JP‐S is Partner and Head of Health Analytics at Lane Clark & Peacock LLP, Chair of the Royal Society for Public Health, and reports personal fees from Novo Nordisk A/S and Pfizer Ltd. outside of the submitted work. MSC, SH and AT are employees of Lane Clark & Peacock LLP; Lane Clark & Peacock LLP received consulting fees from Novo Nordisk A/S to perform this analysis. KSM and SC are employees of Novo Nordisk A/S.
Supporting information
Data S1.Supporting information.
ACKNOWLEDGEMENTS
Medical writing support was provided by Johanna Scheinost DPhil of Oxford PharmaGenesis, Oxford, UK, in accordance with Good Publication Practice (GPP 2022) guidelines (www.ismpp.org/gpp-2022) with funding from Novo Nordisk A/S.
Pearson‐Stuttard J, Chan MS, Holloway S, Sommer Matthiessen K, Thompson A, Capucci S. Estimating healthcare resource utilisation and cardiovascular events in people with high body mass index and established cardiovascular disease. Diabetes Obes Metab. 2025;27(5):2690‐2697. doi: 10.1111/dom.16271
Parts of this analysis were previously presented at the 30th European Congress on Obesity, 17–20 May 2023, Dublin, Ireland.
DATA AVAILABILITY STATEMENT
The authors confirm that the data supporting the findings of this study are available within the article and its supplemental material. Discover does not permit sharing or publication of individual patient data. Information on how to access Discover for research purposes can be found here: https://discover-now.co.uk/how-to-access-the-data/.
REFERENCES
- 1. Powell‐Wiley TM, Poirier P, Burke LE, et al. Obesity and cardiovascular disease: a scientific statement from the American Heart Association. Circulation. 2021;143(21):e984‐e1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Khan SS, Ning H, Wilkins JT, et al. Association of body mass index with lifetime risk of cardiovascular disease and compression of morbidity. JAMA Cardiol. 2018;3(4):280‐287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Afshin A, Forouzanfar MH, Reitsma MB, et al. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med. 2017;377(1):13‐27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Musich S, MacLeod S, Bhattarai GR, et al. The impact of obesity on health care utilization and expenditures in a Medicare supplement population. Gerontol Geriatr Med. 2016;2:2333721415622004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Maradit Kremers H, Visscher SL, Kremers WK, Naessens JM, Lewallen DG. Obesity increases length of stay and direct medical costs in total hip arthroplasty. Clin Orthop Relat Res. 2014;472:1232‐1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Evans M, Anupindi VR, Dekoven M, et al. Eight‐year trends in obesity‐related complications and health care cost progression in a US population with obesity: a retrospective cohort study. Diabetes Obes Metab. 2023;25(2):536‐544. [DOI] [PubMed] [Google Scholar]
- 7. Divino V, Ramasamy A, Anupindi VR, et al. Complication‐specific direct medical costs by body mass index for 13 obesity‐related complications: a retrospective database study. J Manag Care Spec Pharm. 2021;27(2):210‐222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. le Roux CW, Hartvig NV, Haase CL, Nordsborg RB, Olsen AH, Satylganova A. Obesity, cardiovascular risk and healthcare resource utilization in the UK. Eur J Prev Cardiol. 2020;28(11):1235‐1241. [DOI] [PubMed] [Google Scholar]
- 9. Discover‐NOW . The Discover dataset [online]. Accessed 8 February, 2024. https://discover-now.co.uk/the-data/
- 10. Bottle A, Cohen C, Lucas A, et al. How an electronic health record became a real‐world research resource: comparison between London's Whole Systems Integrated Care database and the Clinical Practice Research Datalink. BMC Med Inform Decis Mak. 2020;20(1):71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. European Commission , Eurostat , Pace M, et al. Revision of the European standard population – Report of Eurostat's task force – 2013 edition. Publications Office; 2013. [Google Scholar]
- 12. Chan MS, van den Hout A, Pujades‐Rodriguez M, et al. Socio‐economic inequalities in life expectancy of older adults with and without multimorbidity: a record linkage study of 1.1 million people in England. Int J Epidemiol. 2019;48(4):1340‐1351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. O'Doherty MG, Cairns K, O'Neill V, et al. Effect of major lifestyle risk factors, independent and jointly, on life expectancy with and without cardiovascular disease: results from the Consortium on Health and Ageing Network of Cohorts in Europe and the United States (CHANCES). Eur J Epidemiol. 2016;31(5):455‐468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. van den Hout A. Multi‐state survival models for interval‐censored data. 1st ed. Chapman & Hall/CRC; 2016. [Google Scholar]
- 15. Jackson C. Multi‐state models for panel data: the msm package for R. J Stat Soft. 2011;38(8):1‐28. [Google Scholar]
- 16. Berkelmans GFN, Gudbjornsdottir S, Visseren FLJ, et al. Prediction of individual life‐years gained without cardiovascular events from lipid, blood pressure, glucose, and aspirin treatment based on data of more than 500 000 patients with type 2 diabetes mellitus. Eur Heart J. 2019;40(34):2899‐2906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Aminian A, Zajichek A, Arterburn DE, et al. Association of metabolic surgery with major adverse cardiovascular outcomes in patients with type 2 diabetes and obesity. JAMA. 2019;322:1271‐1282. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Sjöström L, Peltonen M, Jacobson P, et al. Bariatric surgery and long‐term cardiovascular events. JAMA. 2012;307(1):56‐65. [DOI] [PubMed] [Google Scholar]
- 19. Sundström J, Bruze G, Ottosson J, Marcus C, Näslund I, Neovius M. Weight loss and heart failure: a nationwide study of gastric bypass surgery versus intensive lifestyle treatment. Circulation. 2017;135(17):1577‐1585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Benotti PN, Wood GC, Carey DJ, et al. Gastric bypass surgery produces a durable reduction in cardiovascular disease risk factors and reduces the long‐term risks of congestive heart failure. J Am Heart Assoc. 2017;6(5):e005126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. The Look AHEAD Research Group . Association of the magnitude of weight loss and changes in physical fitness with long‐term cardiovascular disease outcomes in overweight or obese people with type 2 diabetes: a post‐hoc analysis of the look AHEAD randomised clinical trial. Lancet Diabetes Endocrinol. 2016;4(11):913‐921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kristensen SL, Rørth R, Jhund PS, et al. Cardiovascular, mortality, and kidney outcomes with GLP‐1 receptor agonists in patients with type 2 diabetes: a systematic review and meta‐analysis of cardiovascular outcome trials. Lancet Diabetes Endocrinol. 2019;7(10):776‐785. [DOI] [PubMed] [Google Scholar]
- 23. Sheikh A, Patel NJ, Nalluri N, et al. Trends in hospitalization for atrial fibrillation: epidemiology, cost, and implications for the future. Prog Cardiovasc Dis. 2015;58(2):105‐116. [DOI] [PubMed] [Google Scholar]
- 24. Wolowacz SE, Samuel M, Brennan VK, Jasso‐Mosqueda J‐G, Van Gelder IC. The cost of illness of atrial fibrillation: a systematic review of the recent literature. Europace. 2011;13(10):1375‐1385. [DOI] [PubMed] [Google Scholar]
- 25. Jha V, Al‐Ghamdi SMG, Li G, et al. Global economic burden associated with chronic kidney disease: a pragmatic review of medical costs for the inside CKD research Programme. Adv Ther. 2023;40(10):4405‐4420. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Office for National Statistics . The impact of winter pressures on adults in Great Britain: December 2022 [online]. 2022. Accessed February 6, 2023. https://www.ons.gov.uk/peoplepopulationandcommunity/wellbeing/articles/theimpactofwinterpressuresonadultsingreatbritain/december2022#nhs‐waiting‐lists
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1.Supporting information.
Data Availability Statement
The authors confirm that the data supporting the findings of this study are available within the article and its supplemental material. Discover does not permit sharing or publication of individual patient data. Information on how to access Discover for research purposes can be found here: https://discover-now.co.uk/how-to-access-the-data/.