Multiple myeloma (MM) is a clonal plasma cell proliferative disorder originating from plasma cells. It is characterized by abnormal proliferation of plasma cells in the bone marrow and the production of monoclonal immunoglobulins, which can lead to end-organ damage such as bone pain, anemia, renal dysfunction, and hypercalcemia. MM accounts for approximately 10% of all hematologic malignancies. Despite substantial advances in diagnostic techniques and therapeutic strategies, including the development of novel agents (1, 2), MM remains an active and evolving field within clinical oncology and is still regarded as an incurable malignancy (3–7).

In the United States, MM represents approximately 2% of newly diagnosed cancers and contributes to a similar proportion of cancer-related deaths. Compared with global patterns, the incidence of MM in the U.S. has shown a gradual upward trend over the past several decades, whereas overall mortality has declined modestly (3, 8–10)737. For instance, a study based on SEER and CDC WONDER data from 1999 to 2020 demonstrated that while MM incidence steadily increased, age-adjusted mortality rates (AAMRs) consistently declined, with significant disparities observed among racial and ethnic groups (11). Additionally, an investigation into MM mortality inequities across the U.S. revealed that the rate of AAMR decline varied among population subgroups, further exacerbating public health challenges (12).

Although previous studies have focused on trends in MM incidence and survival improvement, comprehensive analyses of mortality trends and influencing factors among the general U.S. population between 1999 and 2023 remain limited. In particular, with the broader implementation of treatment modalities such as autologous stem cell transplantation, targeted therapies, and immunotherapies, it remains unclear whether emerging mortality patterns exhibit heterogeneity across sex, race/ethnicity, age groups, geographic regions, or levels of urbanization. Existing evidence suggests that different subgroups (e.g., males vs. females, Black vs. White individuals) display variations in MM incidence, prognosis, and mortality risk (3). For example, studies have shown that although African Americans are diagnosed with MM at more than twice the rate of non-Hispanic Whites, the mortality gap between these groups persists, underscoring ongoing health inequities (13, 14). Moreover, differences in treatment accessibility, healthcare resource distribution, socioeconomic status, and demographic changes may play crucial roles in shaping these divergent trends (15–17).

To better understand mortality trends and their determinants in MM, epidemiological analyses are essential. The Centers for Disease Control and Prevention (CDC) Wide-ranging Online Data for Epidemiologic Research (WONDER) database serves as a vital national public health resource, providing U.S. mortality data since 1999. This database offers valuable insights for examining temporal trends and demographic disparities in MM-related deaths (18, 19). Therefore, this study systematically analyzes underlying-cause-of-death records for MM from 1999 to 2023 using the CDC WONDER database. Specifically, we aim to evaluate temporal variations in age-adjusted mortality rates (AAMRs) overall and by sex, race/ethnicity, age group, geographic region, state, and level of urbanization; quantify percentage changes between 1999 and 2023; estimate annual and average annual percent changes (APC/AAPC); and explore potential determinants and regional heterogeneity underlying these trends. By identifying persistent health disparities among population subgroups, this study seeks to provide an evidence base for public health policy formulation, equitable resource allocation, and the prioritization of interventions—ultimately contributing to improved outcomes and survival among patients with multiple myeloma.

This study, based on the 1999–2023 underlying cause of death data from CDC WONDER, systematically delineates the spatiotemporal patterns and stratified disparities in multiple myeloma (MM)–related mortality in the United States. Overall, we observed that, except for the oldest age group (≥85 years) (20), crude mortality rates declined across most age groups during the study period, with several age groups showing accelerated declines in recent years. Concurrently, AAMR exhibited substantial heterogeneity across states and demographic strata—for example, some states experienced accelerated declines in the past decade, whereas others showed slower decreases or early oscillations. These findings are consistent with recent studies based on SEER and CDC WONDER data, which report that although MM incidence has generally increased or remained stable in recent decades, age-adjusted mortality rates have declined overall, yet significant differences persist across racial/ethnic groups and regions (21).

Such geographic and temporal heterogeneity suggests that changes in mortality are unlikely to be driven by a single biological factor; instead, they more plausibly reflect the combined effects of multilevel structural and epidemiologic determinants, for example regional incidence differences, racial and age composition, availability of medical resources and treatment access, socioeconomic status, and variation in data registration and coding practices (19, 22–24). Consequently, when interpreting state-level APC patterns, priority should be given to these mutable ecological and institutional factors rather than to merely restating observed segmented trends. First, interstate differences in incidence and population structure will directly affect AAMR: if a state has experienced a recent increase in MM incidence or a faster rate of population aging, its mortality trajectory may worsen even when treatment capacity is comparable (6, 22). Second, racial composition and socioeconomic inequities — for example the proportions of non-White groups, poverty rates, and educational attainment — can alter mortality outcomes by influencing disease exposure, timeliness of diagnosis, access to care and the comorbidity profile; prior studies show that MM incidence and outcomes are closely linked to race and socioeconomic status, and these factors are highly unevenly distributed across states (13, 25, 26). Third, healthcare resources and treatment accessibility — including the density of oncology and hematology specialists and transplant centers, the diffusion of modern targeted and immune therapies and autologous stem cell transplantation (ASCT), and differences in insurance coverage — determine whether patients receive timely, guideline-concordant therapy and thus shape mortality trends (22, 24, 27). Fourth, public-health shocks and changes in coding or registration practices, such as short-term mortality spikes related to COVID-19 or modifications in cause-of-death attribution, may induce temporary APC fluctuations (28). Finally, data quality issues and small-number variability are more pronounced in less populous states and can exaggerate APC changes near joinpoints. Given these considerations, explanations for state-level variation should move beyond descriptive accounts and toward systematic testing of the measurable covariates outlined above.

Therapeutic advances likely account for a substantial portion of the observed decline in mortality. Since the 2000s, the introduction of proteasome inhibitors, immunomodulatory drugs, monoclonal antibodies, and CAR-T therapies has significantly prolonged both progression-free and overall survival among MM patients, thereby contributing to population-level reductions in mortality. Our observations—for instance, the marked declines in several age groups and states after 2015—temporally coincide with the gradual adoption of these novel treatments, suggesting that treatment accessibility and uptake may have had important effects on population-level mortality trends (29). Despite these therapeutic advances, a significant concern remains the phenomenon of early mortality, defined as death within the first six months following diagnosis. A substantial proportion of patients still succumb to the disease during this critical early period, underscoring that the survival benefits of novel therapies are not universally accessible or effective at presentation (30–32).

However, the benefits of declining mortality have not been equitably distributed across populations. Both this study and prior research have documented pronounced demographic disparities: non-Hispanic Black individuals and males have historically borne higher incidence and mortality burdens (10), while states and urbanization levels exhibit varying temporal patterns and magnitudes of decline (e.g., New York, Ohio, and Kansas showed accelerated declines after certain breakpoints, whereas larger states such as California and Texas, despite high absolute deaths, displayed distinct AAMR trajectories). These disparities likely reflect a confluence of mechanisms, including differences in disease biology (e.g., MGUS prevalence), historical disparities in healthcare access, uneven adoption and benefit of treatments, and regional variations in socioeconomic status and healthcare infrastructure. Recent studies highlighting racial/ethnic and sex-based differences in MM incidence and outcomes further support these explanatory pathways (19, 23, 33–35). CDC WONDER–based analyses indicate that urbanization level exerted differential effects on MM mortality between 1999 and 2020, with patients residing in rural areas experiencing comparatively higher mortality than those in urban settings (36). This disparity likely reflects reduced access to healthcare resources in rural communities, including shortages of specialist physicians (e.g., hematologist–oncologists), limited availability of advanced diagnostic equipment, and fewer treatment facilities such as chemotherapy centers and hematopoietic stem-cell transplantation units. Rural patients often must travel long distances to obtain specialized care, which not only increases financial and logistical burdens but may also contribute to reduced treatment adherence and delays in diagnosis and therapy (34).

Age-stratified results warrant particular attention. Although most young and middle-aged groups (e.g., 35–74 years) showed sustained declines with recent acceleration, the ≥85-year group exhibited a persistent modest increase. This pattern may reflect multiple factors in the oldest patients, including higher comorbidity burdens, poorer functional status, limited tolerance for intensive or novel therapies, and clinical tendencies toward more conservative management. Additionally, underrepresentation of elderly patients in clinical trials may limit the translation of trial-based survival improvements to the oldest age grou These findings underscore the need for targeted assessment and intervention strategies for older MM patients in both public health and clinical practice to avoid “treatment benefit gaps” (29).

We observed a sustained increase in mortality among individuals aged 85 years and older, a trend that can be partly attributed to limited treatment eligibility and inadequate access to medical resources. Older patients are often less likely to receive specialist assessment and complex therapies. On one hand, uneven distribution of oncology and hematology specialists and workforce shortages force elderly patients in remote or resource-poor areas to travel long distances or face prolonged waits for referral, thereby delaying diagnosis and treatment decisions (37). On the other hand, access to advanced diagnostics, such as molecular testing and PET–CT, and to complex treatment facilities, including chemotherapy centers or transplant units experienced in the care of older adults, is uneven across regions and institutions; as a result, some older patients with potentially treatable disease cannot receive optimal or potentially beneficial therapies (38). Consequently, the United States and other regions face a risk of inadequate oncology and hematology workforce coverage for the elderly and systemic barriers to transplant and high-intensity therapies. These structural access issues can produce an age-based “exclusion” effect that contributes to rising mortality in the very-old age grou When interpreting the increase in mortality among those aged 85 and older, it is therefore important to consider “healthcare resource and specialist accessibility” as a key structural driver and to emphasize the inclusion of variables such as oncology and transplant center density, referral distance, and insurance coverage in national or state-level ecological analyses to test this mechanism.

Comorbidity burden provides another critical pathway explaining the rising mortality in the oldest age grou Compared with younger patients, individuals aged 85 and older carry a higher burden of multiple chronic conditions — cardiovascular disease, diabetes, chronic respiratory disease and others — and are more likely to be excluded from standard cancer treatment pathways because of these comorbidities (39, 40). The literature shows that comorbidity substantially worsens cancer outcomes and, in patients with localized or regional disease, often reduces overall survival by increasing non-cancer competing causes of death (41, 42). Thus, in interpreting the mortality increase in this age stratum, it is essential to disentangle contributions from cancer-specific deaths and deaths related to non-cancer comorbidities, and to consider treatment modifications induced by comorbidity (dose reductions or treatment discontinuation) as mediating variables. Incorporating comorbidity indices such as the Charlson Comorbidity Index into survival or temporal trend models can help quantify the contribution of comorbidity to APC changes. Likewise, frailty is an independent predictor of outcomes in older oncology patients: it is associated with higher treatment-related toxicity and complication rates and with worse overall survival (43). Numerous studies have shown that baseline frailty or prefrailty in older patients receiving chemotherapy or other systemic therapies significantly increases the risk of severe adverse events, reduces quality of life, and elevates short- to mid-term mortality (44). Consequently, clinicians often adopt conservative strategies based on functional status or frailty assessments, including dose reductions or the avoidance of aggressive therapy, to minimize toxicity. Although such caution may reduce acute harms, it can also lead to potentially avoidable cancer deaths or the loss of curative opportunities at the population level, and thus contribute to the observed rise in mortality among those aged 85 and older (45). For these reasons, explanations of increasing mortality in the elderly should incorporate “treatment selection and under-treatment driven by frailty” into the causal chain. We therefore recommend routine inclusion of simple frailty screening tools, such as clinical frailty scales or the G8, in epidemiologic studies and clinical practice to better distinguish deaths resulting from treatment avoidance due to frailty from deaths attributable to end-stage disease (46).

Patients with multiple myeloma (MM) are inherently immunocompromised, rendering them more susceptible to infections, which remain among the leading causes of mortality in this population. Sepsis, in particular, represents a major contributor to infection-related deaths. A nationwide analysis of sepsis-associated mortality among MM patients from 1999 to 2020 reported a declining overall trend; however, the persistent burden of this complication warrants ongoing clinical vigilance (47). In addition, respiratory infections such as pneumonia, influenza, and COVID-19 constitute important causes of death in MM patients. An analysis of national mortality data from 1999 to 2022 revealed dynamic temporal changes in mortality attributable to these infections, with markedly elevated risks and disease severity observed among MM patients during the COVID-19 pandemic period (28, 48). Similar limitations are common to population-based death certificate studies and have been previously discussed in the literature (12).

This study leverages nationally representative CDC WONDER data over a long period (1999–2023) and examines multiple stratifications, including sex, age, race/ethnicity, state, and urbanization level. This allows a systematic assessment of MM mortality patterns and potential inequities at the national level. Such population-level, long-term ecological analyses provide baseline evidence for public health decision-making and help identify high-burden populations and regions that warrant prioritization for intervention (21).

Based on our findings, future research and policy directions should include: first, cohort or registry studies linking diagnosis, treatment, and outcomes to elucidate how treatment uptake, effectiveness, and social determinants drive population-level mortality changes; second, the design and evaluation of interventions to improve access, optimize care, and facilitate early diagnosis and treatment for high-burden populations, such as non-Hispanic Black individuals, older patients, and states with slower declines; and third, enhancing the representation of vulnerable populations (e.g., elderly, racial/ethnic minorities) in clinical trials to ensure equitable benefits from novel therapies. Overall, although population-level MM mortality has declined, research and policy efforts must prioritize “equitable benefit” to ensure that mortality improvements do not create or exacerbate health disparities across populations (29).

Several methodological limitations should be noted. First, CDC WONDER data are derived from death certificates, in which the underlying cause of death is recorded by certifiers and may be subject to diagnostic and coding inaccuracies, potentially leading to misclassification or systematic bias. Additionally, WONDER suppresses data for cells with low counts, which may affect the precision of estimates for small subpopulations or certain states. Second, mortality databases do not provide individual-level information on treatment exposure, pathological staging, comorbidities, or socioeconomic status, limiting our ability to attribute regional or population-level mortality differences directly to treatment accessibility or social determinants. Third, although our study spans 1999–2023 and captures long-term trends, short-term jumps in recent years (e.g., the rapid decline observed from 2020–2023) should be interpreted cautiously, as they may reflect reporting delays, changes in clinical practice, or statistical segmentation methods.

Using CDC WONDER national data, we analyzed U.S. multiple myeloma (MM)–related mortality from 1999–2023. Although national age-adjusted mortality rates (AAMRs) declined overall, substantial disparities persisted by age, sex, race/ethnicity, and geography: mortality increased among those aged ≥85; males had higher AAMRs than females, though declines in both accelerated recently; non-Hispanic Black individuals consistently had the highest AAMR. Geographically, the Southeast and Midwest bore higher burdens while the West and North showed larger declines; New York, Ohio, and Kansas exhibited accelerated reductions, whereas Oklahoma and West Virginia showed little change. All urbanization levels experienced declines, with large metropolitan areas seeing faster recent decreases. The findings underscore the need for targeted public-health strategies to ensure equitable access to effective therapies and for linking clinical and social determinants to mortality trends to guide interventions and resource allocation.