Quantifying cardiovascular disease burden attributable to risk factors in Pakistan: a call for data-driven action

Cardiovascular disease (CVD) remains the leading cause of death globally, accounting for over 18.6 million deaths in 2019, with projections exceeding 23 million by 2030.^1,2 Its clinical and economic burden is immense, and modifiable metabolic, behavioral, environmental, and social risk factors continue to drive its global spread.³ The World Health Organization estimated that over three-quarters of these deaths occur in low- and middle-income countries (LMICs), highlighting a growing epidemic in resource-constrained settings.⁴ Among the LMICs, Pakistan presents a particularly stark case. Home to 0.23 billion people and the world’s fifth most populous country, Pakistan saw its annual CVD deaths double from 170,000 to 340,000 between 1990 and 2019.^2,5 The Global Burden of Disease study reported Pakistan’s age-standardized CVD incidence at 918.18 per 100,000, substantially higher than the global rate of 684.33 per 100,000. The CVD death rate followed a similar trend: 358.88 per 100,000 in Pakistan versus 239.85 per 100,000 globally.⁶ These high rates have been echoed by local data, with 18.9% of the households in a national socioeconomic survey self-reporting a history of CVD.⁷ Several compounding risk factors fuel this trend. Pakistan has a 30.8% age-adjusted diabetes prevalence, 25% smoking rate among adult men, and 37% adult hypertension prevalence. Moreover, high rates of consanguinity (58% marrying first or second cousins) may also contribute to clustering of genetic risks. With a predominantly young population and increasing life expectancy, these risk factors will likely amplify CVD burden over time. Despite this, Pakistan spends only $43 per capita on healthcare (1.2% of GDP) and ranks in the global bottom decile for universal health coverage effectiveness.^7,8 The need for comprehensive, real-time surveillance and risk stratification has never been more urgent.

In this issue of Baylor University Medical Center Proceedings, Mansoor et al⁹ present a timely analysis leveraging the Global Burden of Disease 2019 study data to quantify CVD burden in Pakistan attributable to 12 modifiable risk factors across three domains: behavioral risks of smoking, low physical activity, alcohol, and diet; metabolic risks of high systolic blood pressure, low-density lipoprotein cholesterol, body mass index, fasting glucose, and kidney dysfunction; and environmental risks of air pollution, lead exposure, and temperature extremes. Dietary risks were assessed as a composite of 15 subcomponents with protective or harmful effects (e.g., fruit, fiber, sodium, trans fats). CVD burden was quantified via number of deaths, disability-adjusted life years (DALYs), and age-standardized mortality rates (ASMR) and DALY rates (ASDR) per 100,000. The authors reported estimates by age, sex, year, and sociodemographic index and analyzed temporal patterns across three intervals: 1990 to 2019, 1990 to 2010, and 2010 to 2019. Trends were summarized using estimated annual percentage change with 95% confidence intervals (CI). Estimated annual percentage change was classified as increasing, decreasing, or stable depending on CI boundaries.

Mansoor et al⁹ reported that, over the 1990 to 2019 period, Pakistan’s annual CVD mortality and DALYs approximately doubled, compared to a 1.5-fold increase globally. ASMRs and ASDRs in Pakistan followed a stable, inverted U-shape trend peaking in 2002; they remained consistently higher than declining global counterparts. Ambient particulate matter–related ASMR increased from 16 to 51 and ASDR from 375 to 1286 in Pakistan, whereas global ASMR was unchanged and ASDR modestly increased. Household air pollution, while decreasing in Pakistan (ASMR 88 to 49; ASDR 2093 to 1205), remained persistently elevated compared to global rates. Burden from lead exposure and low ambient temperature remained stable in Pakistan, in contrast to global declines. Metabolic risks were even more striking. High systolic blood pressure remained the top contributor to CVD burden in Pakistan, with ASMR rising from 172 to 202 and ASDR from 3813 to 4614, against a global decline in both. Low-density lipoprotein cholesterol–related ASMR/ASDR increased (61 to 85; 1524 to 2139), as did body mass index–related burden (ASMR 23 to 56; ASDR 643 to 1539), diverging from declining and stable global trends. Similarly, fasting glucose–related ASMR/ASDR nearly doubled in Pakistan but decreased or plateaued globally. Kidney dysfunction–related burden also rose in Pakistan while declining worldwide. Behavioral risk–related burdens were consistently higher in Pakistan. ASMR and ASDR from dietary risks, smoking, and secondhand smoke remained stable in Pakistan but declined globally. Physical inactivity–related burden rose modestly in Pakistan (ASMR 12 to 15; ASDR 196 to 250), while falling globally. Alcohol-related burden was low and stable across both.

The authors are to be commended for conducting a timely and comprehensive analysis that quantifies CVD burden attributable to key risk factors in Pakistan. However, several limitations inherent to secondary data analyses from global modeling studies should be acknowledged. The absence of high-quality, locally collected surveillance data such as validated death certificates, disease registries, or verbal autopsy systems limits the precision of these estimates. Moreover, wide uncertainty intervals for Pakistan may obscure meaningful cross-regional or temporal differences. Additionally, the study does not account for critical subnational heterogeneity. Without data stratified by province, socioeconomic status, or urban-rural divide, large equity gaps remain invisible, limiting the translational value of these findings for targeted health interventions. Furthermore, evolving definitions of risk factors, changes in surveillance methods, and reliance on population-attributable fractions complicate longitudinal interpretation, especially in high-risk LMIC settings where multirisk exposure and interaction effects are common. Global estimates that include Pakistan may further dilute national contrasts, diminishing the visibility of urgent country-specific trends. Notably, important contributors to CVD burden in Pakistan such as rheumatic heart disease and congenital heart disease conditions, tied to early life socioeconomic, environmental, and infectious risk factors, are not explicitly represented in the risk factor framework of this study. Their exclusion may underrepresent the full spectrum of CVD burden in younger populations, especially in underserved and pediatric populations. A crucial unanswered question that emerges is: To what extent do subnational, socioeconomic, and rural-urban disparities modulate the rising cardiometabolic burden in Pakistan, and how can targeted, data-driven strategies be designed to address these gaps? Looking ahead, Pakistan must transition from passive surveillance to real-time cardiovascular monitoring. Establishing electronic health records, disease registries, and representative national surveys will be key to validating and contextualizing modeled estimates. Integration of clinical, behavioral, and environmental data at the local level will empower stakeholders to design high-impact interventions targeting top risk drivers, most notably hypertension, dyslipidemia, air pollution, and physical inactivity.

In summary, Mansoor et al⁹ provide a compelling assessment of Pakistan’s worsening cardiovascular landscape. Rising absolute CVD mortality and DALYs, driven by persistently high exposures to modifiable risk factors, underscore the urgent need for proactive surveillance, policy reform, and scalable public health interventions. Without immediate investment in high-level data infrastructure and targeted action, Pakistan’s cardiometabolic crisis is succumbed to escalate further.

—Ahmed Kamal Siddiqi, MBBS^a, Zainab Samad, MD^b, and Muhammad Shahzeb Khan, MD, MSc^c
Email: ahmedsiddiqi2020@gmail.com
^aDivision of Cardiothoracic Imaging, Department of Radiology and Imaging Sciences, Emory University, Atlanta, Georgia, USA; ^bCITRIC Health Data Science Center, The Aga Khan University, Karachi, Pakistan, and Department of Medicine, The Aga Khan University, Karachi, Pakistan; ^cDepartment of Medicine, Baylor College of Medicine, Temple, Texas, USA, Baylor Scott and White Research Institute, Dallas, Texas, USA, and Baylor Scott and White Health: The Heart Hospital, Plano, Texas, USA