The Covid-19 pandemic as an inflection point in US mortality trends from hematologic and circulatory system diseases: a piecewise trend analysis Authors: Amirali Koohi Bachemir, Zhenshan Sun, Melika Safari, Nasrin Borumandnia, Hamid Alavi Majd Published: 2026 ABSTRACT AbstractBackgroundThe COVID-19 pandemic disrupted healthcare systems and may have influenced mortality trends in hematologic-immune disorders (D50-D89) and Diseases of the circulatory system (I00-I99). This study evaluates US mortality trends from 2010 to 2023 to assess potential pandemic-related changes.MethodsWe utilized spline regression modeling on CDC WONDER mortality data (2010–2023) to quantify post-pandemic trend shifts in both age-adjusted and sex/age-stratified mortality rates.ResultPost-2019 mortality trends for D50-D89 and I00-I99 diseases showed significant increases, particularly among older adults and younger males. D50-D89 mortality in females ≥ 65 increased ~ 2.6% (95% CI 1.70% to 3.68%) annually, while males ≥ 65 increased ~ 1.6% (95% CI 0.28% to 3.07%) annually; younger males (15–34 years) showed a 4.5% (95% CI 0.49% to 8.87%) annual increase. I00-I99 mortality increases were most pronounced in older females (~ 1.4% annual increase, 95% CI 0.49 to 2.31). Analysis of standardized rates in all groups except for men in I00–I99 showed a positive slope-shift after 2019. (p < 0.05)”.ConclusionThese findings highlight the persistent impact of the COVID-19 pandemic on hematologic and circulatory system disease mortality, emphasizing the need for continued monitoring of vulnerable populations.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12889-026-28064-y.Keywords: COVID-19 pandemic, Hematologic disorders, Blood disorders, Circulatory System disorders Introduction Blood diseases and diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism (D50-D89) and diseases of the circulatory system (I00-I99) are significant global health burdens whose mortality trends are influenced by advancements in diagnostic technology, therapeutic progress, and demographic changes. The COVID-19 pandemic emerged in the U.S. in early 2020, quickly escalating to alarming levels. It significantly impacted American society, economy, and international relations [1]. COVID-19, as a systemic endothelial damage syndrome [2, 3], devastated global healthcare systems, creating novel complications in the form of long-COVID, thromboembolic events, and immune dysregulation. This virus has significantly affected the blood coagulation system. This has led to disorders of hemostasis, such as disseminated intravascular coagulation. Moreover, high concentrations of D-dimer and fibrinogen can predict severe progression, as well as the development of thrombi, due to inflammation [4]. SARS-CoV-2 can cause damage to tissues and organs, with cardiovascular diseases such as myocarditis, myocardial infarction, and arrhythmia being among the most significant manifestations [5]. Additionally, mRNA COVID-19 vaccines, although crucial for pandemic control, have been linked to rare immune-mediated adverse effects, such as myocarditis [6, 7]. Previous studies have shown that before the COVID-19 pandemic, mortality from ischemic heart disease and cerebrovascular disease was declining, and the number of cardiovascular patients remained stable. In 2020, mortality from hypertensive diseases increased by 26% and from atrial fibrillation by 18%, while mortality rates from cardiovascular disease and coronary heart disease returned to pre-pandemic levels by 2021–2022. Additionally, cases of congestive heart failure and acute myocardial infarction increased significantly in 2020–2021, indicating a higher burden of these diseases during the pandemic [8–10]. This study aims to fill the existing gap in understanding how the COVID-19 pandemic has influenced long-term mortality trends in I00-I99 and D50-D89 diseases in the United States. This analysis enables us to identify whether the pandemic represents a significant shift in mortality patterns and which populations were most affected. Methodology Dataset In this study, we analyzed temporal mortality trends for diseases of the blood and blood-forming organs (ICD-10 codes D50-D89) and circulatory system diseases (ICD-10 codes I00-I99) from 2010 to 2023. Mortality data were obtained from the CDC WONDER database, which combines two datasets: Underlying Cause of Death by Bridged-Race Categories (from 2010 to 2018) and Underlying Cause of Death by Single-Race Categories (from 2019 to 2023). The study utilized de-identified, publicly available data (https://wonder.cdc.gov/deaths-by-underlying-cause.html) and did not require Institutional Review Board approval. Analysis for the 2010–2023 years was segmented into two primary analytical approaches: Total analysis, based on age-adjusted death rates, and stratified analysis that utilized unadjusted death rates with stratification by sex and age groups. For stratified analysis, data were grouped by sex and year (2010 to 2023). Age stratification (children: <15 years, young adults: 15–34 years, middle-aged adults: 35–64 years, older adults: ≥65 years) was employed to identify biological, clinical, and epidemiological variations among these distinct age groups. Eight sex-age subgroups were considered for each disease category in the stratified analysis. Age-adjusted mortality rates (per 100,000 population) for the overall population were obtained directly from the CDC WONDER database. These rates were calculated using the 2000 US standard population for age adjustment. Statistical analysis We modeled age-specific mortality counts using count-model spline regression with a log-population offset to account for differing population sizes. We formally assessed over-dispersion for each model. (Supplementary Table 1) To evaluate temporal changes in mortality, we applied a linear spline structure with a single knot at 2019. The breakpoint for this spline was strategically placed at the year 2019. This selection is justified because the global outbreak and widespread effects of COVID-19 predominantly began in early 2020. Although the widespread effects of the pandemic became evident from 2020 onward, setting the knot in 2019 allows the model to quantify any trend changes occurring after 2020 as a structural break at the threshold of the pandemic period. If we place the knot in 2020, the increase in mortality in 2020 itself might be considered part of the pre‑knot trend and thus overlooked. We reported the annual percentage change before 2019, the annual percentage change after 2019 (2020 to 2023), and the difference between these slopes (the change in slope after 2019 relative to before 2019), along with 95% confidence intervals and p-values. For age-standardized rates we fitted a Gaussian spline-model and reported the resulting slopes. These slopes represent the absolute change in rate per 100,000 individuals. Clinically, these parameters provide a quantitative measure of post-pandemic shifts in mortality risk and help identify subgroups with heightened vulnerability. All statistical analyses were performed in R version 4.4.1, using specialized packages including lspline for spline modeling [11], and ggplot2 for visualization. For assessing the stability of our results, we performed a sensitivity analysis with a knot point in 2020 (Supplementary Table 2). Result The mortality trends of hematologic and circulatory system diseases from 2010 to 2023, based on age-adjusted rates and age-specific rates, are presented in Figs. 1 and 2. These figures display the crude mortality trends for descriptive purposes. Tables 1 and 2 present the results of the spline regression models for age-specific mortality rates, showing changes in trend slopes across sex-age subgroups. This table displays the annual slope changes for the periods before and after 2019. It also details the difference in slope post-2019, reported as a percentage change with 95% Confidence Intervals and p-values. These results highlight the post-2019 inflection in mortality trends and provide a quantitative assessment of the changes in both age-specific and age-adjusted rates following the onset of the COVID-19 pandemic. Mortality Trends for Hematologic Disorders (D50-D89) in the United States, 2010–2023 Mortality Trends for Circulatory System Disorders (I00-I99) in the United States, 2010–2023 Table 1Slope Changes in Age-Specific Mortality Rates for Hematologic-Immune Disorders (D50–D89) and Circulatory System Diseases (I00–I99), United States, 2010–2023ICD GroupSub Group%Annual Change before 2019 (95%CIa)%Annual Change after 2019 (95%CI)% Slope Change (95%CI)p-value for slope-changeD50–D89Females < 15 years-1.69 (-4.28 to 0.97), p = 0.2103.30(-3.27 to 10.30) ,p = 0.3375 (-3.37 to 14.27)0.251Males < 15 years-1.35 (-3.54 to 0.89), p = 0.237-0.17(-5.66 to 5.60) ,p = 0.9531.18(-5.80 to 8.69)0.747Females 15–34 years-0.91(-2.30 to 0.49), p = 0.2011.40(-2.08 to 4.96), p = 0.4332.31(-2.10 to 6.95)0.307Males 15–34 years-1.50 (-2.76 to -0.23), p = 0.0193.03(-0.16 to 6.29) , p = 0.0624.53(0.49 to 8.87) 0.028 Females 35–64 years0.37(-0.52 to 1.27), p = 0.4101.87 (-0.33 to 4.14), p = 0.0931.50( -1.30 to 4.39)0.293Males 35–64 years0.69 (0.16 to 1.23 ), p = 0.0110.77 (-0.51 to 2.07), p = 0.2380.08 (-1.55 to 1.74)0.923Females ≥ 65 years-2.27 (-2.57 to -1.97), p < 0.0010.35 (-0.38 to 1.11), p = 0.3612.62(1.70 to 3.68 ) < 0.001 Males ≥ 65 years-0.98 (-1.43 to -0.53) ,p < 0.0010.67(-0.39 to 1.75), p = 0.2131.65(0.28 to 3.07 ) 0.018 I00–I99Females < 15 years-1.30 (-2.25 to -0.34), p = 0.008-0.64 (-3.10 to 1.86) , p = 0.6110.66(-2.46 to 3.88)0.680Males < 15 years-2.01 (-2.88 to -1.13) ,p < 0.0010.76 (-1.53 to 3.09), p = 0.5162.77( -0.10 to 5.85)0.059Females 15–34 years1.56 (0.62 to 2.51), p = 0.001-0.97 (-3.25 to 1.36), p = 0.403-2.53 ( -5.34 to 0.43)0.091Males 15–34 years0.44(-0.38 to 1.28), p = 0.2960.65 (-1.41 to 2.77), p = 0.5320.21 (-2.39 to 2.89)0.874Females 35–64 years1.28 (0.69 to 1.87), p < 0.0010.64 (-0.80 to 2.12), p = 0.381-0.64 (-2.45 to 1.22)0.496Males 35–64 years1.13(0.50 to 1.75), p < 0.001-0.20 (-1.56 to 1.54), p = 0.979-1.33 ( -3.05 to 0.83)0.251Females ≥ 65 years-2.11 (-2.40 to -1.84), p < 0.001-0.75 (-1.45 to -0.05), p = 0.0031.36(0.49 to 2.31 ) 0.002 Males ≥ 65 years-1.08 (-1.42 to -0.76), p < 0.001-0.13 (-0.96 to 0.70), p = 0.7570.95(-0.09 to 2.04)0.074a: 95% Confidence Interval Slope Changes in Age-Specific Mortality Rates for Hematologic-Immune Disorders (D50–D89) and Circulatory System Diseases (I00–I99), United States, 2010–2023 a: 95% Confidence Interval Table 2Slope Changes in Age-Adjusted Mortality Rates for Hematologic-Immune and Circulatory System Diseases, United States, 2010–2023GroupSlope before 2019(95% CIa)Slope After 2019(95% CIa)Slope Change (95% CI)p-value for slope-changeD50–D89Females (Age-Adjusted)-0.032 (-0.48 to -0.015), p-=0.0010.056 (0.015 to 0.097), p = 0.0110.088 (0.036 to 0.139) 0.003 Males (Age-Adjusted)-0.017 (-0.036 to 0.000), p = 0.0580.056 (0.010 to 0.101), p = 0.0190.073 (0.016 to 0.131) 0.017 I00–I99Females (Age-Adjusted)-1.51 (-2.48 to -0.545), p = 0.0051.74 (-0.657 to 4.14), p = 0.1383.254 (0.200 to 6.308) 0.039 Males (Age-Adjusted)-1.27 (-2.71 to 0.175), p = 0.0792.29 (-1.29 to 5.86), p = 0.1873.556 (− 1.000 to 8.111)0.114a: 95% Confidence Interval Slope Changes in Age-Adjusted Mortality Rates for Hematologic-Immune and Circulatory System Diseases, United States, 2010–2023 a: 95% Confidence Interval Hematologic- immune disorders (D50–D89) For D50–D89 group, the most pronounced increase following 2019 was observed in the males aged 15–34 years cohort, exhibiting a significant anuual increase post-2019 of 4.5% (p = 0.028). Furthermore, both females and males aged 65 years and older demonstrated notable increases in their post-2019 slope trends, recorded at 2.6% and 1.6% respectively. (p < 0.05) These slope changes are interpreted as the percentage difference in the annual rate of mortality change before and after 2019. Extending this analysis to age-adjusted mortality rates (Table 2), both female and male populations indicated a statistically significant upward shift and increase. Specifically, males showed an increase of 0.073 per 100,000 and females demonstrated an increase of 0.088 per 100,000 (p < 0.05), underscoring a systemic impact on hematologic mortality patterns. Circulatory system diseases (I00–I99) In contrast, for I00–I99, the primary significant finding was an annual percentage increase of 1.4% (p = 0.002) observed in females aged 65 years and older. This indicates a reversal from the pre-pandemic downward trend. The age-adjusted mortality rate analysis (Table 2) further corroborated this, revealing the sole statistically significant slope change within this group for females, with an increase of 3.256 per 100,000 (p = 0.039). This suggests a heightened vulnerability and altered mortality trajectory for cardiovascular conditions among older women during the specified period. These results suggest that post-2019 mortality increases in circulatory system diseases were most pronounced in older females. Figures 3 and 4 illustrate the yearly mortality patterns and the fitted piecewise regression slopes for both disease categories. These figures visually display the changes in slope for both disease categories. The p-values shown alongside each segment indicate the statistical significance of these changes. Temporal trends in mortality rates for hematologic diseases (D50–D89) across age and sex subgroups, 2010–2023. Statistically significant slope changes after 2019 (p < 0.05) are highlighted in red Temporal trends in mortality rates for circulatory system diseases (I00–I99) across age and sex subgroups, 2010–2023. Statistically significant slope changes after 2019 (p < 0.05) are highlighted in red Sensitivity analysis The findings of our additional analyses, presented in Supplementary Material 1, offer greater insight into the robustness of our estimates and the interpretation of long-term trends. Specifically, while some subgroups (D50-D89 females ≥ 65 and D50-D89 Age-adjusted female rates ) continued to demonstrate statistically significant differences in slope after introducing a knot at 2020—indicating that the years 2021–2023 maintained a steeper trajectory compared to the pre-2020 period—other subgroups no longer showed statistically significant differences, in contrast to the results obtained using the 2019 knot. Moreover, a decline in slope was observed for certain subgroups within the I00–I99 category, particularly among females aged 15–34 and males aged 35–64. As illustrated in Fig. 4, these two subgroups experienced a marked increase in mortality rates during 2020 and 2021, followed by a return to lower levels comparable to the pre-2020 baseline in subsequent years. Although these groups did not exhibit statistically significant slope changes in the primary model with the 2019 knot—despite the evident absolute increase in mortality in 2020—relocating the knot to 2020 in the sensitivity analysis enabled a statistically significant detection of the slope decrease after 2020 (2021 to 2023). This pattern suggests that the observed increase in mortality within these subgroups was likely a transient phenomenon that subsided after one to two years. Accordingly, slope estimates should be interpreted alongside the raw data plots to provide an accurate understanding of mortality patterns. The placement of different knots may emphasize distinct aspects of the trends, underscoring the importance of visual inspection in conjunction with model-based estimates (Figs. 3 and 4). In males under 15 years of age within the I00–I99 category, a 4.2% annual increase was observed when introducing a knot at 2020 (p = 0.027), whereas this subgroup did not reach statistical significance in the 2019-knot analysis. This finding may indicate a gradual increase in disease burden over time following the pandemic. Discussion The results of this study indicate that the COVID-19 pandemic created a significant inflection point in mortality trends from hematologic and circulatory system diseases in the United States. The significant increase in the mortality slope among older adults, particularly females ≥ 65 years, as well as among young males aged 15–34 years, demonstrates that the consequences of the pandemic extended beyond the acute phase and altered the long-term mortality pattern of these diseases. In this study, we used the 2000 standard U.S. population for age adjustment in the analysis of age‑standardized rates, which should be taken into consideration when comparing our results with other studies. We know, in addition to the immediate death toll due to COVID-19 infection, the indirect effects of the pandemic contributed to unprecedented rises in deaths due to other causes among the younger age group [12]. Individuals diagnosed with COVID-19 exhibit a considerable susceptibility to the onset of coagulation abnormalities. The pathology of the disease undermines the typical hemostatic equilibrium by inducing endothelial injury, activating the inflammatory response, and promoting platelet activation, culminating in a state that markedly elevates the likelihood of thromboembolic incidents, such as venous thrombosis, deep vein thrombosis, and pulmonary embolism, particularly in cases of severe manifestation [13, 14]. Our results are based on Underlying Cause of Deaths (UCoD) data, for which the primary cause was recorded on the death certificate. Considering that COVID-19 claimed a substantial portion of the most vulnerable individuals during the early waves of the pandemic, the harvesting effect may have influenced some subgroups. However, our findings show that post-2019 mortality slopes increased significantly across most age-sex subgroups, particularly among older adults and certain other groups. This indicates that mortality related to hematologic and circulatory system diseases in the post-pandemic period remains severe and clinically significant. Therefore, these increases reflect genuine vulnerability in at-risk populations, underscoring the need for targeted preventive measures and vigilant care [15]. Studies suggest that COVID-19 survivors, including those with no prior history of heart disease, experience significant alterations in cardiac function. These changes, observed in echocardiographic assessments of both the left and right ventricles, can occur during both the post-acute recovery phase and in long COVID syndrome [16]. Furthermore, other studies have shown that at the beginning of the COVID-19 pandemic, mortality from ischemic heart disease and hypertensive diseases in the United States increased significantly. This finding indicates that the pandemic has had significant indirect consequences on patients with cardiovascular diseases. At the national level, deaths from coronary heart disease and hypertension increased during the pandemic period compared to the previous year, with the largest relative increase observed in New York City, while states such as Massachusetts and Louisiana did not experience significant increases [17]. Furthermore, previous studies showed that in 2020 in Italy, the COVID-19 pandemic led to a marked rise in deaths from pulmonary embolism and other cardiovascular diseases (such as hypertension, atrial fibrillation, and cerebrovascular disease), particularly among men and during the second wave of the pandemic [18]. Another study on the indirect impact of the pandemic across 24 countries also showed that the previous improvements in cardiovascular disease outcomes were reversed during the first two years of the pandemic, particularly in Bulgaria and Russia. The decline in life expectancy related to cardiovascular disease continued in England even up to 2022 [19]. Although mRNA vaccination is associated with a relative increase in the risk of thromboembolic events (incidence rate ratio: 1.19 after the first dose and 1.22 after the second dose), it simultaneously and significantly reduces the risk of these complications associated with COVID-19 infection itself [20]. Furthermore, studies on approximately 100 million vaccinated individuals have shown that these vaccines also increase the risk of myocarditis [7]. These findings underscore the intricate relationship between vaccination, infection, and individual health. Further, the psychological and social effects of the pandemic, including increased stress and anxiety, likely influenced patients with hematologic disorders. These psychological effects can lead to downplaying the illness severity, loss of prevention measures, or even evasion of clinical consultations. This is reflected in systematic reviews of the mental health effects during COVID-19 [21]. Biological differences between age and gender also contribute to further vulnerability. In younger males between 15 and 34 years of age, it can be exacerbated by behavioral factors like avoidance of early health care [22]. Previous studies have shown a decrease in hospital visits due to fear of infection [23]. To safeguard at-risk populations, including the elderly, immunocompromised individuals, and those with hematologic or cardiovascular conditions, a comprehensive array of interventions must be instituted. Continuous health surveillance and management of pre-existing illnesses, combined with education to help individuals quickly identify critical symptoms, can mitigate both the direct and indirect consequences of COVID-19 on mortality rates. Additionally, the provision of psychological and social support is crucial for mitigating anxiety and apprehension regarding the pursuit of medical assistance, as mental health challenges may result in diminished preventive care and delayed disease management. Environmental determinants, such as air pollution, also significantly contribute to health outcomes. Subsequent investigations should explore ecological factors, notably air pollution, given their potential to affect hematologic indicators in susceptible groups. Research has shown that even brief exposure to moderate air pollution is associated with considerable declines in leukocyte, lymphocyte, and platelet levels, as well as compromised lung function metrics [24]. Air pollution caused by PM2.5 particles is a global threat to cardiovascular health, significantly increasing the risk of cardiac events in both the short and long term, and requires urgent mitigation measures and further research [25]. Minimizing exposure to air pollution, enhancing indoor air quality, engaging in regular physical exercise, and adhering to a nutritious diet can decrease the likelihood of cardiovascular and respiratory ailments—especially among individuals with a higher propensity for longevity who might otherwise enjoy relatively good health. The amalgamation of these strategies with ongoing monitoring of excess mortality and long-term management initiatives for COVID-19 survivors can enhance public health outcomes and diminish disparities in the consequences associated with the pandemic. In conclusion, our findings indicate that the COVID-19 pandemic marked a significant inflection point in mortality trends for hematologic and circulatory system diseases in the United States. Continued monitoring and targeted prevention strategies are needed to mitigate long-term impacts. Future research should incorporate individual-level clinical data to explain better the mechanisms underlying these shifts. Limitations This study has several limitations that should be taken into account when concluding. First, death certificate-based mortality data for the analysis were obtained from the CDC WONDER database. This could lead to misclassification bias, potentially affecting the reporting accuracy of COVID-19 as a cause of death and possibly obscuring the true impact of hematologic and circulatory system disorders. The obtained data may be subject to misclassification and errors [26]. Another limitation is that individual-level data on treatment modalities, vaccination status, SARS-CoV-2 infection history, comorbidities, and access to healthcare were not available. These unmeasured confounders may have influenced mortality patterns during the pandemic and could not be assessed under the ecological design of this study. Our analysis is based on a relatively small number of annual observations (2010–2023), which limits the ability to fully account for temporal autocorrelation in the data. While residual autocorrelation may lead to underestimation of standard errors and overstatement of statistical significance, the small number of time points also constrains the reliability of formal time-series corrections (e.g., autoregressive models or heteroskedasticity- and autocorrelation-consistent estimators). As a result, findings particularly p-values—should be interpreted with caution. Also multiple sex- and age-stratified analyses were conducted, which increases the potential for type I error due to multiple comparisons. As the primary aim of this study was to describe temporal patterns rather than to formally test specific subgroup hypotheses, these stratified results should be considered exploratory. Accordingly, statistically significant findings in subgroup analyses should be interpreted with caution and may warrant confirmation in future studies. Supplementary Information Supplementary Material 1. Acknowledgements For improved writing clarity and consistency, AI-assisted tools were used; however, all analyses, models, and interpretations were developed entirely by the authors. Authors’ contributions Amirali Koohi Bachemir (A.K.B.): Statistical analysis, Secondary text review, Conceptualization.Zhenshan Sun (Z.S.): Secondary text review, Conceptualization.Melika Safari (M.S.): Initial draft writing, Reference organization.Nasrin Boroumandnia (N.B.): Statistical analysis supervision, Figure preparation.Hamid Alavi Majd (H.A.M.): Correspondence, Overall supervision of the manuscript. Funding No funding was received for this study. Data availability We express our gratitude to the Centers for Disease Control and Prevention (CDC) for providing mortality data through the CDC WONDER online database (https://wonder.cdc.gov/deaths-by-underlying-cause.html), which includes age-adjusted and death rates by cause of death, sex, age group, and other demographic characteristics for the United States. Declarations Ethics approval and consent to participate Not applicable. Data is available at https://wonder.cdc.gov/deaths-by-underlying-cause.html. Consent for publication Not applicable. This manuscript does not contain any individual person’s data in any form. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Nasrin Borumandnia, Email: Borumand.n@gmail.com. Hamid Alavi Majd, Email: alavimajd@gmail.com. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement We express our gratitude to the Centers for Disease Control and Prevention (CDC) for providing mortality data through the CDC WONDER online database (https://wonder.cdc.gov/deaths-by-underlying-cause.html), which includes age-adjusted and death rates by cause of death, sex, age group, and other demographic characteristics for the United States.