Advertisement

Temporal Trends in Racial and Ethnic Disparities in Multimorbidity Prevalence in the United States, 1999-2018

Open AccessPublished:April 23, 2022DOI:https://doi.org/10.1016/j.amjmed.2022.04.010

      Abstract

      Background

      Disparities in multimorbidity prevalence indicate health inequalities, as the risk of morbidity does not intrinsically differ by race/ethnicity. This study aimed to determine if multimorbidity differences by race/ethnicity are decreasing over time.

      Methods

      Serial cross-sectional analysis of the National Health Interview Survey, 1999-2018. Included individuals were ≥18 years old and categorized by self-reported race, ethnicity, age, and income. The main outcomes were temporal trends in multimorbidity prevalence based on the self-reported presence of ≥2 of 9 common chronic conditions.

      Findings

      The study sample included 596,355 individuals (4.7% Asian, 11.8% Black, 13.8% Latino/Hispanic, and 69.7% White). In 1999, the estimated prevalence of multimorbidity was 5.9% among Asian, 17.4% among Black, 10.7% among Latino/Hispanic, and 13.5% among White individuals. Prevalence increased for all racial/ethnic groups during the study period (P ≤ .001 for each), with no significant change in the differences between them. In 2018, compared with White individuals, multimorbidity was more prevalent among Black individuals (+2.5 percentage points) and less prevalent among Asian and Latino/Hispanic individuals (−6.6 and −2.1 percentage points, respectively). Among those aged ≥30 years, Black individuals had multimorbidity prevalence equivalent to that of Latino/Hispanic and White individuals aged 5 years older, and Asian individuals aged 10 years older.

      Conclusions

      From 1999 to 2018, a period of increasing multimorbidity prevalence for all the groups studied, there was no significant progress in eliminating disparities between Black individuals and White individuals. Public health interventions that prevent the onset of chronic conditions in early life may be needed to eliminate these disparities.

      Keywords

      Clinical Significance
      • Multimorbidity prevalence is increasing across racial and ethnic groups in the United States.
      • Multimorbidity was persistently more prevalent among Black individuals from 1999-2018.
      • There was no significant change in the Black–White multimorbidity prevalence disparity.
      • Black individuals had a multimorbidity prevalence equivalent to other groups 5-10 years older.
      • Those with low income also had a persistently higher prevalence of multimorbidity.

      Introduction

      In 2010, the US Department of Health and Human Services published a public health framework for managing patients with multimorbidity, defined as the presence of 2 or more concurrent chronic conditions in an individual, highlighting the importance of identifying and addressing disparities in its prevalence.
      US Department of Health Human Services
      Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions.
      ,
      • Goodman RA
      • Posner SF
      • Huang ES
      • Parekh AK
      • Koh HK
      Defining and measuring chronic conditions: imperatives for research, policy, program, and practice.
      The overall prevalence is increasing in the United States. Moreover, studies indicate that multimorbidity is more prevalent among Black individuals compared with White individuals, even though race, a social construct, has no intrinsic relationship with morbidity risk.
      • King DE
      • Xiang J
      • Pilkerton CS
      Multimorbidity trends in United States adults, 1988-2014.
      ,
      • Johnson-Lawrence V
      • Zajacova A
      • Sneed R
      Education, race/ethnicity, and multimorbidity among adults aged 30-64 in the National Health Interview Survey.
      Since at least 1985, the US government has identified the elimination of health disparities as a national priority.

      United States Department of Health and Human Services. Task Force on Black and Minority Health. Report of the Secretary's Task Force on Black and Minority Health. Report of the Secretary's Task Force on Black & Minority Health. 1985. Available at: http://resource.nlm.nih.gov/8602912. Accessed March 17, 2022.

      Whether the US has made progress in reducing or eliminating racial disparities in common chronic conditions over the last 2 decades is not known. The examination of trends in multimorbidity has implications about accountability for what the health system in the United States has achieved in population health and health equity over the past 20 years, and what it will face in the future.
      Accordingly, we describe the 20-year trends in the prevalence of multimorbidity by race and ethnicity, estimating the annual differences between groups. We also evaluate whether the effect of race and ethnicity on multimorbidity prevalence varied by age, describing the age groups among which these differences begin and are the highest. Finally, considering that income is one of the most critical determinants of health, we stratified the main analyses by income level, which can provide insight into the role of income inequality in the disparities studied.

      Methods

      Data Source

      We used data from the annual National Health Insurance Survey (NHIS) from 1999 to 2018. The NHIS uses a complex multistage area probability design that accounts for non-response and allows for nationally representative estimates, including for underrepresented groups. We used data from the NHIS Sample Adult Core file, which contains responses from an in-depth questionnaire administered to a randomly selected adult from each family (Supplementary Methods, available online). The mean conditional and final response rates of the Sample Adult Core survey during the study period were 81% and 64.8%, respectively. We obtained the data from the Integrated Public Use Microdata Series Health Surveys (https://nhis.ipums.org/).
      • Blewett LA
      • Rivera Drew JA
      • King ML
      • Williams KCW
      IPUMS health surveys: National Health Interview Survey, version 6.4.
      The Institutional Review Board at Yale University exempted the study from review.

      Study Population

      We included individuals ≥18 years old from years 1999 to 2018. We excluded responses with missing information about chronic conditions. Due to small numbers, we also excluded those who did not identify as non-Hispanic Asian, non-Hispanic Black/African American, Latino/Hispanic, or non-Hispanic White (details in Results section).

      Demographic Variables

      Race was determined by the participants’ self-reported primary race selection. Latino/Hispanic ethnicity was defined as answering “Yes” to the question, “Do you consider yourself Latino/Hispanic?” Participants were divided into 4 mutually exclusive subgroups: non-Hispanic Asian (Asian), non-Hispanic Black/African American (Black), Latino/Hispanic, and non-Hispanic White (White).
      Respondent characteristics included in the analyses were self-reported age, sex, geographic region (Northeast, North Central/Midwest, South, West), and income. Age was also categorized by 5-year groups for stratification and reporting. Household income was categorized according to its relative percentage of the federal poverty level from the US Census Bureau

      United States Census Bureau. People in families by family structure, age, and sex, iterated by income-to-poverty ratio and race. Available at:https://www.census.gov/data/tables/time-series/demo/income-poverty/cps-pov/pov-02.html#par_textimage_30. Accessed March 17, 2022.

      into middle/high income (≥200%) and low income (<200%).
      • Mahajan S
      • Caraballo C
      • Lu Y
      • et al.
      Trends in differences in health status and health care access and affordability by race and ethnicity in the United States, 1999-2018.
      Other sociodemographic and clinical variables were used to describe the characteristics of the population.

      Chronic Conditions

      We included 9 chronic conditions that the US Centers for Disease Control and Prevention consistently included in the NHIS across the entire study period. These conditions were among those listed by the US Department of Health and Human Services Office of the Assistant Secretary of Health conceptual framework for studying multiple chronic conditions. We identified individuals with these conditions if they had ever been told by a health care professional that they had diabetes, hypertension, asthma, stroke, cancer, chronic obstructive pulmonary disease (emphysema or chronic bronchitis), or heart disease (coronary artery disease, myocardial infarction, angina, or other heart conditions), or had been told in the past 12 months that they had weak/failing kidneys or any liver condition. Participants were then classified as having 0, 1, 2, 3, or ≥4 chronic conditions. We defined multimorbidity as the presence of 2 or more of the 9 chronic conditions.
      US Department of Health Human Services
      Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions.
      ,
      • Drye EE
      • Altaf FK
      • Lipska KJ
      • et al.
      Defining multiple chronic conditions for quality measurement.

      Statistical Analysis

      We summarized the general characteristics of respondents by racial/ethnic group. Then, we estimated the annual multimorbidity prevalence for each group using multivariable logistic regression models adjusting for age, sex, and region (details in Supplementary Methods, available online).
      We subtracted the annual prevalence among White individuals from that year's prevalence among Asian, Black, and Latino/Hispanic individuals, also constructing standard error for the differences, to quantify the racial and ethnic gap. We also estimated trends over the study period by fitting weighted linear regression models (Supplementary Methods). We used a z-test to estimate the absolute difference between 1999 and 2018 in the multimorbidity prevalence within each race and ethnic group, and the differences between groups.
      Then, we estimated the adjusted annual mean number of chronic conditions by race and ethnicity using linear regressions. We also used ordered logistic regression models to estimate the proportion of individuals with 0, 1, 2, 3, or ≥4 chronic conditions over the years (Supplementary Methods).
      To assess if the association between race and ethnicity and multimorbidity differs by income, we replicated the main analysis by income group using the NHIS multiply imputed income variables for respondents who did not report income (Supplementary Methods).

      Division of Health Interview Statistics National Center for Health Statistics. Multiple imputation of family income and personal earnings in the National Health Interview Survey: methods and examples. August 2019. Available at:https://nhis.ipums.org/nhis/resources/tecdoc18.pdf. Accessed March 17, 2022.

      To evaluate if the association between race and ethnicity and multimorbidity differs by age, we used a logistic regression model that included an interaction term between age and race and ethnicity (Supplementary Methods). We replicated this using an ordered logit model to estimate the percentage with 0, 1, 2, 3, and ≥4 conditions in each age group by race and ethnicity. We reported the findings from these analyses based on the approach of Barnett et al.
      • Barnett K
      • Mercer SW
      • Norbury M
      • Watt G
      • Wyke S
      • Guthrie B
      Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study.
      Lastly, we performed a sensitivity analysis including arthritis in the list of conditions using data from 2002 to 2018, the period in which the condition was available and ascertained consistently. The purpose of this sensitivity analysis was to compare our multimorbidity prevalence estimates with those obtained from NHIS studies that included arthritis in their calculation.
      • Boersma P
      • Black LI
      • Ward BW
      Prevalence of multiple chronic conditions among US adults, 2018.
      • Ward BW
      • Schiller JS
      • Goodman RA
      Peer reviewed: multiple chronic conditions among us adults: a 2012 update.
      • Ward BW
      • Schiller JS
      Prevalence of multiple chronic conditions among US adults: estimates from the National Health Interview Survey, 2010.
      For all analyses, a 2-sided P-value < .05 was used to determine statistical significance. All analyses were performed using Stata SE version 17.0 (StataCorp LLC, College Station, Texas). All results are reported with 95% confidence intervals (CIs). All analyses incorporated strata and weights to produce nationally representative estimates using the Stata -svy- command for structured survey data. All person-weights were pooled and divided by the number of years studied, following guidance from the NHIS.

      Centers for Disease Control and Prevention. National Health Interview Survey, 1997-2018. Survey description document. Available at:https://www.cdc.gov/nchs/nhis/1997-2018.htm. Accessed March 17, 2022.

      Results

      Population Characteristics

      Of the 603,140 adults interviewed, we excluded 112 with missing race and ethnicity information and 119 with missing chronic conditions information. Due to small representation, we also excluded 6673 with ethnicity other than Latino/Hispanic who also did not select a primary race (n = 2114) or who selected primary race as Alaskan Native or American Indian (n = 4355) or Other (n = 204) (Supplementary Figure 1, available online). The final study sample comprised 596,236 adults with an estimated mean age of 46.2 (standard error 0.07) years, of which 51.8% (95% CI, 51.7-52.0) were women. Of these, 4.7% (95% CI, 4.6%-4.8%) identified as Asian, 11.8% (95% CI, 11.5%-12.1%) as Black, 13.8% (95% CI, 13.5%-14.2%) as Latino/Hispanic, and 69.7% (95% CI, 69.3%-70.2%) as White. The overall estimated prevalence of multimorbidity was 19.9% (95% CI, 19.7%-20.1%), increasing from 15.2% (95% CI, 14.8%-15.7%) in 1999 to 22.7% (95% CI, 22.1%-23.4%) in 2018. The general characteristics of the study population are shown in Table 1, the Supplementary Table (available online), and Supplementary Figure 2 (available online). The overall unadjusted distribution of number of chronic conditions by race and ethnicity is shown in Supplementary Figure 3 (available online), and the annual adjusted prevalence of each of the conditions by race and ethnicity is shown in Supplementary Figure 4 (available online).
      Table 1General Characteristics of the Study Population by Race and Ethnicity
      AsianBlackLatino/HispanicWhite
      1999-20002008-20092017-20181999-20002008-20092017-20181999-20002008-20092017-20181999-20002008-20092017-2018
      Sample size, n (Total = 596,236)n = 1597n = 2828n = 2640n = 8758n = 7804n = 5881n = 10,305n = 8838n = 6422n = 41,872n = 29,553n = 36,401
      Age in years39 (28-55)41 (30–55)43 (32-57)40 (29-52)42 (29-55)43 (30-58)37 (27-49)37 (28-50)39 (28-53)44 (33-59)47 (33-61)50 (34-64)
      Age category
       18-39 years52.4 (49.2-55.6)45.9 (43.4-48.5)41.9 (39.3-44.6)49.8 (48.3-51.3)45.1 (43.4-46.8)43.7 (41.8-45.6)56.9 (55.4-58.3)54.6 (53.0-56.1)50.1 (48.4-51.9)39.5 (38.9-40.2)34.9 (34.0-35.8)33.3 (32.6-34.1)
       40-64 years38.2 (35.1-41.4)40.8 (38.2-43.5)42.5 (40.2-44.9)38.5 (37.1-39.8)42.9 (41.4-44.4)41.1 (39.3-42.9)34.1 (33.0-35.3)37.1 (35.7-38.5)39.1 (37.6-40.6)42.2 (41.5-42.8)45.8 (45.0-46.5)42.3 (41.6-43.0)
       ≥65 years9.4 (7.8-11.4)13.3 (11.8-14.9)15.5 (13.9-17.4)11.8 (11.0-12.6)12.0 (11.2-12.9)15.3 (14.2-16.3)9.0 (8.1-10.0)8.4 (7.7-9.1)10.8 (9.9-11.8)18.3 (17.8-18.8)19.3 (18.7-20.0)24.4 (23.7-25.1)
      Women51.4 (48.1-54.6)52.6 (50.3-54.9)53.3 (51.1-55.6)55.6 (54.1-57.0)55.4 (53.9-56.9)54.5 (52.8-56.3)50.6 (49.4-51.8)48.6 (47.2-50.0)50.3 (48.7-51.9)51.9 (51.3-52.4)51.6 (51.0-52.3)51.4 (50.7-52.0)
      US Citizenship (n = 594,859)57.9 (54.3-61.4)69.6 (66.8-72.2)70.5 (67.9-73.1)95.4 (94.7-96.0)95.0 (94.2-95.6)94.9 (93.8-95.9)62.6 (60.5-64.7)62.2 (60.2-64.1)72.3 (70.6-74.0)98.4 (98.2-98.5)98.5 (98.3-98.7)98.5 (98.3-98.7)
      Education level (n = 591,667)
       <High school11.4 (9.4-13.7)9.2 (7.9-10.7)7.9 (6.6-9.5)24.4 (23.1-25.8)17.7 (16.4-19.1)14.2 (12.9-15.7)44.7 (43.0-46.3)38.5 (37.0-40.1)27.8 (26.0-29.7)13.4 (12.9-13.9)10.5 (10.0-11.1)7.4 (6.9-7.9)
       High school diploma /GED18.1 (16.1-20.2)16.3 (14.4-18.3)15.3 (13.5-17.4)31.5 (30.2-32.8)30.6 (29.2-32.1)28.9 (27.3-30.5)24.3 (23.3-25.4)26.6 (25.2-28.0)26.7 (25.1-28.3)31.7 (31.0-32.4)28.2 (27.5-28.9)23.6 (22.9-24.4)
       Some college24.2 (21.7-26.9)24.4 (22.2-26.7)20.8 (18.8-22.9)29.7 (28.3-31.0)34.0 (32.5-35.5)32.9 (31.2-34.7)21.7 (20.5-23.0)22.6 (21.4-23.8)28.1 (26.5-29.7)29.5 (29.0-30.0)31.5 (30.9-32.2)31.4 (30.6-32.1)
       ≥Bachelor's degree46.4 (43.3-49.5)50.1 (46.8-53.5)56.0 (52.9-59.0)14.5 (13.4-15.7)17.7 (16.5-18.9)24.1 (22.2-26.0)9.4 (8.5-10.3)12.4 (11.4-13.4)17.5 (16.0-19.0)25.4 (24.7-26.1)29.8 (28.9-30.7)37.6 (36.5-38.7)
      Income <200% of Federal Poverty Limit
      Annual family income was categorized relative to the respective year's Federal Poverty Level from the US Census Bureau into middle/high income (≥200%) and low income (<200%). The weighted proportion of individuals with annual income <200% of the Federal Poverty Limit was estimated using multiple imputation.
      28.0 (23.9-32.5)27.1 (23.9-30.5)26.2 (22.9-29.9)45.6 (43.3-47.8)45.0 (42.7-47.2)44.4 (41.4-47.4)51.0 (49.0-53.1)50.3 (48.1-52.5)46.7 (44.1-49.3)23.0 (22.2-23.8)23.9 (22.8-25.1)21.3 (20.3-22.3)
      Uninsured at the time of interview

      (n = 594,011)
      17.4 (15.2-19.8)13.9 (12.2-15.8)6.3 (5.3-7.6)20.3 (19.2-21.4)20.8 (19.6-22.0)12.1 (10.8-13.5)36.3 (34.5-38.1)38.9 (37.0-40.8)24.0 (22.2-26.0)11.0 (10.6-11.4)12.3 (11.8-12.9)6.5 (6.2-6.9)
      US region of residence
      Based on the US Census Bureau recognized region where the housing unit of the survey participant was located.
       Northeast21.2 (18.7-24.0)18.8 (16.4-21.5)20.8 (16.9-25.4)18.0 (16.7-19.4)16.4 (14.8-18.2)16.0 (13.5-18.8)15.7 (14.5-17.1)13.4 (11.7-15.2)13.5 (11.2-16.2)20.2 (19.5-21.0)18.3 (17.3-19.3)19.2 (17.5-21.1)
       Midwest14.7 (11.9-18.0)15.0 (12.7-17.7)11.8 (9.5-14.6)18.8 (17.2-20.5)19.3 (17.3-21.5)15.2 (13.0-17.8)7.9 (6.9-9.1)9.5 (8.1-11.2)9.4 (7.6-11.6)29.4 (28.5-30.3)28.6 (27.2-30.0)27.4 (25.7-29.3)
       South18.9 (16.4-21.8)19.3 (17.2-21.7)25.5 (21.6-30.0)55.7 (53.4-57.9)56.1 (53.5-58.7)60.5 (56.7-64.3)35.4 (33.2-37.8)35.4 (33.4-37.4)37.1 (32.5-42.0)33.9 (33.0-34.8)33.8 (32.3-35.4)33.0 (30.9-35.2)
       West45.2 (41.2-49.2)46.8 (43.6-50.1)41.8 (36.7-47.1)7.6 (6.7-8.5)8.1 (7.3-9.1)8.3 (6.9-9.9)40.9 (38.6-43.3)41.7 (39.4-44.1)40.0 (35.3-44.8)16.5 (15.8-17.2)19.3 (18.3-20.4)20.4 (18.3-22.7)
      Married or living with partner

      (n = 593,807)
      64.4 (61.3-67.3)64.1 (61.2-66.8)65.0 (62.5-67.5)37.4 (35.9-38.9)35.2 (33.6-36.8)32.6 (31.0-34.3)58.5 (57.4-59.6)54.4 (52.8-56.0)49.3 (47.6-50.9)61.6 (60.8-62.3)57.9 (56.9-58.8)56.4 (55.6-57.1)
      Employment status

      (n = 595,478)
       With a job/working66.1 (63.4-68.6)65.1 (62.6-67.5)67.2 (64.9-69.3)64.1 (62.6-65.5)60.7 (59.2-62.3)61.1 (59.0-63.1)66.3 (65.0-67.5)64.2 (62.9-65.5)66.5 (64.7-68.3)65.7 (65.0-66.4)62.5 (61.6-63.4)62.1 (61.3-63.0)
       Not in labor force31.8 (29.2-34.5)30.4 (28.0-32.8)30.0 (27.8-32.3)32.0 (30.6-33.5)31.3 (29.9-32.8)32.4 (30.5-34.4)31.2 (30.0-32.4)28.9 (27.8-30.1)29.9 (28.0-31.8)32.9 (32.3-33.6)33.4 (32.5-34.3)35.4 (34.6-36.2)
       Unemployed2.2 (1.5-3.1)4.6 (3.7-5.6)2.8 (2.1-3.8)3.9 (3.4-4.6)7.9 (7.2-8.8)6.5 (5.7-7.4)2.5 (2.2-2.9)6.9 (6.2-7.7)3.6 (3.0-4.3)1.4 (1.3-1.5)4.1 (3.8-4.4)2.5 (2.3-2.7)
      Chronic conditions
       Asthma5.5 (4.4-6.9)9.1 (7.8-10.6)8.4 (7.2-9.9)9.0 (8.2-9.9)13.9 (12.9-13.8)15.1 (14.0-16.2)7.1 (6.5-7.7)10.0 (9.2-10.8)11.6 (10.5-12.8)9.2 (8.9-9.5)13.3 (12.9-13.813.9 (13.5-14.3)
       Cancer1.7 (1.1-2.7)3.0 (2.4-3.7)4.3 (3.4-5.3)2.9 (2.5-3.4)3.8 (3.4-4.4)4.8 (4.3-5.4)2.2 (1.9-2.5)2.7 (2.3-3.2)3.4 (2.9-3.8)7.8 (7.5-8.1)10.2 (9.8-10.6)12.3 (11.9-12.8)
       COPD1.6 (1.1-2.4)1.8 (1.3-2.6)2.0 (1.4-2.7)4.3 (3.8-4.7)4.5 (4.0-5.1)4.4 (3.7-5.1)2.9 (2.5-3.4)2.8 (2.4-3.4)2.8 (2.4-3.3)6.2 (5.9-6.5)6.5 (6.2-6.9)5.3 (5.0-5.6)
       Diabetes3.9 (3.0-5.1)7.4 (6.2-8.8)8.9 (7.7-10.3)8.2 (7.5-8.9)11.3 (10.4-12.2)11.7 (10.8-12.7)6.3 (5.8-6.9)8.7 (8.0-9.6)10.5 (9.6-11.4)5.2 (5.0-5.4)8.2 (7.9-8.6)9.1 (8.7-9.4)
       Heart disease4.6 (3.6-5.9)4.9 (4.0-6.2)6.6 (5.6-7.7)8.9 (8.2-9.7)9.9 (9.1-10.8)9.8 (9.0-10.7)6.0 (5.4-6.6)6.1 (5.5-6.7)6.4 (5.7-7.2)12.1 (11.7-12.4)13.8 (13.2-14.3)14.0 (13.6-14.5)
       Hypertension14.9 (12.9-17.2)22.3 (20.2-24.6)24.2 (22.2-26.3)28.8 (27.6-30.1)35.4 (34.0-36.9)37.6 (35.9-39.3)15.3 (14.4-16.2)20.0 (18.9-21.1)22.3 (21.0-23.8)23.0 (22.5-23.4)30.4 (29.7-31.1)32.7 (31.9-33.4)
       Kidney disease0.8 (0.5-1.5)1.1 (0.8-1.7)1.8 (1.2-2.6)2.0 (1.7-2.5)2.2 (1.7-2.7)2.6 (2.2-3.1)1.6 (1.3-1.9)1.8 (1.6-2.1)2.0 (1.6-2.5)1.3 (1.2-1.5)1.8 (1.6-2.0)2.3 (2.1-2.5)
       Liver disease1.1 (0.5-2.3)1.3 (0.9-2.0)1.7 (1.2-2.4)1.0 (0.8-1.3)1.0 (0.7-1.3)1.2 (0.9-1.5)1.0 (0.8-1.2)1.8 (1.5-2.1)2.4 (2.0-2.9)1.0 (0.9-1.1)1.5 (1.3-1.7)1.8 (1.6-2.0)
       Stroke0.8 (0.6-1.2)1.2 (0.8-1.7)1.9 (1.4-2.5)2.8 (2.4-3.2)3.3 (2.8-3.8)4.0 (3.5-4.6)1.1 (0.9-1.4)1.6 (1.3-1.9)2.1 (1.7-2.6)2.2 (2.1-2.3)3.0 (2.8-3.2)3.3 (3.1-3.6)
      Current smoker14.9 (12.9-17.2)11.0 (9.5-12.7)7.2 (6.0-8.5)23.8 (22.6-25.1)21.3 (20.0-22.7)14.7 (13.5-16.1)18.3 (17.4-19.3)15.2 (14.1-16.2)9.8 (9.0-10.8)24.2 (23.7-24.8)22.1 (21.4-22.8)15.1 (14.6-15.7)
      Flu vaccine in past 12 months26.1 (23.3-29.2)33.4 (31.2-35.7)48.2 (45.6-50.7)20.6 (19.7-21.5)26.3 (25.0-27.7)35.2 (33.4-37.1)18.6 (17.5-19.6)22.1 (20.9-23.2)34.9 (33.3-36.5)30.7 (30.0-31.3)36.9 (36.2-37.6)47.1 (46.4-47.9)
      Obese (BMI ≥30 kg/m2)6.1 (4.9-7.7)8.7 (7.1-10.5)12.0 (10.5-13.6)29.1 (28.0-30.2)37.1 (35.8-38.4)39.4 (37.8-41.1)23.0 (21.9-24.1)31.2 (29.9-32.4)34.0 (32.5-35.6)20.2 (19.7-20.6)26.1 (25.4-26.8)30.3 (29.6-31.0)
      Data are presented as % (95% confidence interval) for categorical variables and median (P25–P75) for continuous variables. All percentages are unadjusted and weighted.
      BMI = body mass index; COPD = chronic obstructive pulmonary disease; GED = general equivalency diploma.
      low asterisk Annual family income was categorized relative to the respective year's Federal Poverty Level from the US Census Bureau into middle/high income (≥200%) and low income (<200%). The weighted proportion of individuals with annual income <200% of the Federal Poverty Limit was estimated using multiple imputation.
      Based on the US Census Bureau recognized region where the housing unit of the survey participant was located.

      Temporal Multimorbidity Trends, 1999 to 2018

      In 1999, the estimated age-, sex-, and region-adjusted prevalence of multimorbidity was 5.9% (95% CI, 4.3-8.0) among Asian individuals, 17.2% (95% CI, 15.7-18.8) among Black individuals, 10.5% (95% CI, 9.3-11.7) among Latino/Hispanic individuals, and 13.2% (95% CI, 12.8-13.7) among White individuals (Figure 1). Between 1999 and 2018, the adjusted multimorbidity prevalence increased significantly within each race and ethnicity group (P ≤ .001 for each). There were no significant changes in the difference between groups (Table 2). In 2018, compared with the prevalence among White individuals (18.7%; 95% CI, 18.0-19.5), the estimated multimorbidity prevalence among Black individuals was 2.5 percentage points higher (95% CI, 0.5-4.6; P = .02), whereas it was lower among Asian and Latino/Hispanic individuals by 6.6 points (95% CI, −8.8 to −4.4; P < .001) and 2.1 points (95% CI, −4.0 to −0.2; P = .03), respectively. Throughout the study period, the adjusted prevalence of multimorbidity was persistently higher among individuals with low income than among those with middle/high income, and in 2018 the difference between Black and White individuals was not significant when stratified by income (Table 2 and Supplementary Figure 5).
      Figure 1
      Figure 1Trends in adjusted multimorbidity prevalence by race and ethnicity, 1999-2018. Data source is the National Health Interview Survey from 1999 to 2018. Multimorbidity was defined as the presence of 2 or more of the following conditions: asthma, cancer, chronic obstructive pulmonary disease, diabetes, heart disease, hypertension, stroke, liver disease, and “weak or failing kidneys.” Annual prevalence for each race and ethnicity was obtained using logistic regression models adjusted by age, sex, and US region. The 95% confidence intervals are represented in brackets.
      Table 2Change in the Adjusted Multimorbidity Prevalence from 1999 to 2018, by Race and Ethnicity
      Asian

      Percentage Points (95% CI), P Value
      Black

      Percentage Points (95% CI), P Value
      Latino/Hispanic

      Percentage Points (95% CI), P Value
      White

      Percentage Points (95% CI), P Value
      Annualized rate of change in prevalence
       Overall+0.23 (+0.14-+0.31), < .001+0.31 (+0.20-+0.43), < .001+0.31 (+0.23-+0.40), < .001+0.28 (+0.21-+0.34), < .001
       Low income+0.31 (+0.15-+0.46), .001+0.38 (+0.23-+0.52), < .001+0.36 (+0.27-+0.45), < .001+0.45 (+0.34-+0.56), < .001
       Middle/high income+0.18 (+0.10-+0.27), .001+0.24 (+0.12-+0.36), .001+0.27 (+0.16-+0.38), < .001+0.22 (+0.16-+0.27), < .001
      Absolute change in prevalence from 1999 to 2018
       Overall+6.19 (+3.39-+9.00), < .001+4.09 (+1.62-+6.55), .001+6.19 (+4.08-+8.30), < .001+5.50 (+4.64-+6.36), < .001
       Low income+8.17 (+1.59-+14.8), .02+5.65 (+1.66-+9.64), .005+7.99 (+4.76-+11.22), < .001+8.57 (+6.45-+10.70), < .001
       Middle/high income+5.68 (+2.46-+8.90), < .001+3.33 (+0.25-+6.40), .03+4.88 (+1.99-+7.77), < .001+4.70 (+3.80-+5.60), < .001
      Difference with White in 1999
       Overall−7.32 (−9.27 to −5.38), < .001+3.95 (+2.34-+5.56), < .001−2.76 (−4.03 to −1.49), < .001
       Low income−11.06 (−15.60 to −6.52), < .001+2.94 (+0.20-+5.68), .04−6.67 (−8.76 to −4.58), < .001
       Middle/high income−6.34 (−8.70 to −3.97), < .001+1.89 (−0.29-+4.06), .09−2.64 (−4.39 to −0.89), .003
      Difference with White in 2018
       Overall−6.63 (−8.83 to −4.43), < .001+2.54 (+0.48-+4.59), .02−2.07 (−3.96 to −0.18), .03
       Low income−11.46 (−16.68 to −6.23), < .001+.02 (−3.58-+3.62), 0.99−7.25 (−10.50 to −4.00), < .001
       Middle/high income−5.35 (−7.72 to −2.98), < .001+0.51 (−1.85-+2.87), 0.67−2.46 (−4.92-+.01), .05
      Absolute change in difference with White from 1999 to 2018
       Overall+0.69 (−2.24-+3.63), 0.65−1.41 (−4.03-+1.20), 0.29+0.69 (−1.59-+2.97), 0.55
       Low income−0.40 (−7.32-+6.52), 0.91−2.92 (−7.44-+1.60), 0.21−0.58 (−4.44-+3.29), 0.77
       Middle/high income+0.98 (−2.36-+4.33), 0.57−1.37 (−4.58-+1.83), 0.40+0.18 (−2.84-+3.21), 0.91
      Data source is the National Health Interview Survey from 1999 to 2018. For change in multimorbidity prevalence and change in difference: a positive sign (+) means the prevalence of multimorbidity, or its difference with White individuals, increased; a negative sign (−) means it decreased. The estimated multimorbidity prevalence was adjusted by age, sex, and region.
      CI = confidence interval.
      Supplementary TableStudy Population Characteristics
      AsianBlackLatino/HispanicWhite
      Sample size

      N = 596,236
      27,75584,94299,650383,889
      Age in years41 (30-55)42 (29-55)38 (28-50)47 (33-61)
      Age category
       18-39 years45.7 (44.7-46.6)45.4 (44.8-46.0)54.2 (53.7-54.8)35.3 (35.0-35.7)
       40-64 years41.5 (40.7-42.4)41.8 (41.3-42.3)36.7 (36.2-37.1)44.4 (44.1-44.7)
       ≥65 years12.8 (12.3-13.5)12.8 (12.4-13.2)9.1 (8.8-9.4)20.3 (20.0-20.6)
      Women52.3 (51.6-53.1)55.2 (54.7-55.7)49.6 (49.1-50.0)51.7 (51.5-51.9)
      US Citizenship

      n = 594,859
      68.1 (67.0-69.1)95.3 (95.0-95.7)64.4 (63.6-65.2)98.4 (98.3-98.5)
      Education level

      n = 591,667
       <High school9.9 (9.3-10.5)18.3 (17.8-18.9)36.9 (36.2-37.6)10.3 (10.1-10.5)
       High school diploma /GED16.4 (15.7-17.1)30.6 (30.1-31.1)26.2 (25.8-26.7)27.9 (27.6-28.2)
       Some college22.4 (21.7-23.2)32.7 (32.2-33.3)24.3 (23.8-24.8)31.0 (30.7-31.3)
       ≥Bachelor's degree51.3 (50.1-52.5)18.4 (17.9-18.9)12.6 (12.2-13.0)30.8 (30.4-31.2)
      Income <200% Federal poverty level
      The annual family income was categorized into low income and middle/high income relative to the respective year's federal poverty level from the US Census Bureau (<200% and ≥200%, respectively). The weighted proportion of individuals with low income was estimated using multiple imputation.
      28.2 (24.9-31.7)46.1 (43.9-48.3)51.5 (50.2-52.7)23.9 (23.0-24.9)
      Uninsured at the time of interview

      (n = 594,011)
      12.9 (12.3-13.5)18.5 (18.1-18.9)34.1 (33.4-34.8)10.5 (10.4-10.7)
      US region of residence
      Based on the US Census Bureau recognized region of the housing unit where the survey participant was interviewed.
       Northeast20.1 (18.9-21.4)16.3 (15.5-17.0)14.0 (13.3-14.8)19.3 (18.8-19.8)
       Midwest13.3 (12.3-14.3)17.8 (17.0-18.7)9.0 (8.3-9.8)28.2 (27.6-28.8)
       South21.7 (20.4-23.0)57.8 (56.6-59.1)36.3 (35.0-37.6)33.9 (33.3-34.6)
       West44.9 (43.3-46.6)8.1 (7.7-8.5)40.7 (39.3-42.1)18.6 (18.2-19.1)
      Married or living with partner

      n = 593,807
      64.5 (63.6-65.3)35.1 (34.6-35.7)53.9 (53.4-54.4)58.5 (58.1-58.9)
      Employment status

      n = 595,478
       With a job/working65.3 (64.5-66.2)60.5 (59.9-61.0)65.4 (64.9-65.9)62.9 (62.6-63.2)
       Not in labor force30.8 (29.9-31.6)32.0 (31.4-32.5)29.3 (28.8-29.8)34.0 (33.7-34.3)
       Unemployed3.9 (3.7-4.2)7.6 (7.3-7.9)5.3 (5.1-5.5)3.2 (3.1-3.2)
      Current smoker10.2 (9.7-10.7)19.7 (19.3-20.2)13.4 (13.1-13.7)20.6 (20.4-20.9)
      Flu vaccine in past 12 months36.8 (35.9-37.6)27.1 (26.7-27.6)24.6 (24.1-25.0)37.2 (36.9-37.5)
      Obese (BMI ≥30 kg/m2)9.1 (8.7-9.6)36.3 (35.8-36.8)29.6 (29.1-30.1)25.6 (25.4-25.9)
      Conditions
       Asthma8.0 (7.6-8.5)13.2 (12.9-13.5)9.5 (9.3-9.8)12.1 (12.0-12.2)
       Cancer2.9 (2.7-3.2)4.0 (3.8-4.1)2.8 (2.7-2.9)10.0 (9.9-10.1)
       COPD1.8 (1.6-2.0)4.7 (4.5-4.9)2.8 (2.7-2.9)5.9 (5.8-6.0)
       Diabetes7.2 (6.8-7.6)11.0 (10.7-11.2)8.6 (8.3-8.9)7.7 (7.5-7.8)
       Heart disease5.7 (5.3-6.0)9.6 (9.3-9.8)6.2 (6.0-6.4)13.2 (13.0-13.4)
       Hypertension21.0 (20.3-21.7)35.0 (34.5-35.6)20.0 (19.6-20.4)28.9 (28.7-29.2)
       Kidney disease1.1 (1.0-1.3)2.2 (2.1-2.4)1.8 (1.7-1.9)1.7 (1.7-1.8)
       Liver disease1.4 (1.2-1.6)1.1 (1.0-1.2)1.7 (1.6-1.8)1.4 (1.4-1.5)
       Stroke1.5 (1.4-1.7)3.4 (3.3-3.6)1.7 (1.6-1.8)2.8 (2.7-2.8)
      Number of concurrent conditions
       066.9 (66.1-67.8)51.0 (50.4-51.5)65.7 (65.3-66.2)51.3 (51.0-51.5)
       121.4 (20.7-22.1)27.6 (27.2-28.0)12.1 (20.8-21.5)27.3 (27.1-27.4)
       ≥211.7 (11.2-12.3)21.4 (21.0-21.9)13.2 (12.8-13.5)21.5 (21.3-21.7)
      Data are presented as % (95% confidence interval) for categorical variables and median (P25-P75) for continuous variables. All percentages are weighted and unadjusted.
      BMI = body mass index; CI = confidence interval; COPD = chronic obstructive pulmonary disease; GED = general equivalency diploma.
      low asterisk The annual family income was categorized into low income and middle/high income relative to the respective year's federal poverty level from the US Census Bureau (<200% and ≥200%, respectively). The weighted proportion of individuals with low income was estimated using multiple imputation.
      Based on the US Census Bureau recognized region of the housing unit where the survey participant was interviewed.
      The estimated mean number of chronic conditions was the highest among Black individuals throughout the study period (Figure 2A). Similarly, Black individuals had the greatest prevalence of at least 1, 2, 3, or 4 conditions over the study period (Figure 2B).
      Figure 2
      Figure 2Trends in (A) estimated mean number of chronic conditions and (B) ordered number of chronic conditions by race and ethnicity, 1999-2018. Data source is the National Health Interview Survey from 1999 to 2018. The mean number of chronic conditions was estimated using linear regression. The ordered number of chronic conditions was estimated using ordered logistic regression. All estimates were adjusted by age, sex, and US region. The 95% confidence intervals are represented in brackets in Panel A. For visualization purposes, no symbols or brackets are shown in panel B. (Details in the Methods section.)

      Association Between Multimorbidity and Age

      The estimated prevalence of reporting at least 1, 2, 3, or 4 conditions all increased with age across all groups, with Black individuals having the greatest prevalence in each age category (Supplementary Figure 6, available online). Among those aged ≥30 years, Black individuals had multimorbidity prevalence equivalent to those of Latino/Hispanic and White individuals aged 5 years older, and to those of Asian individuals aged 10 years older (Figure 3A). Notably, the adjusted prevalence of multimorbidity (≥2 concurrent conditions) diverged between Black individuals and White individuals among those aged 30-34 years, reached its maximum difference of 10.1 percentage points (95% CI, 8.2-12.1) among those aged 60-64 years, and converged to reach similar prevalence among those aged ≥80 years (Figure 3B).
      Figure 3
      Figure 3(A) Adjusted multimorbidity prevalence by age and race and ethnicity, and (B) prevalence differences by age and race and ethnicity. Data source is the National Health Interview Survey from 1999 to 2018. Panel A shows the sex- and region-adjusted prevalence of multimorbidity by age for each race and ethnicity group. Panel B shows the difference in such a prevalence between Asian, Black, and Latino/Hispanic individuals with White individuals of the same age group (eg, prevalence of multimorbidity among Black individuals aged 50-54 years—prevalence of multimorbidity among White individuals aged 50-54 years). The 95% confidence intervals are represented in brackets.
      There were significant racial and ethnic differences in the association between age and prevalence of multimorbidity. Asian, Black, and Latino/Hispanic individuals had a greater increase in the prevalence of multimorbidity with age compared with White individuals (multimorbidity*decade of age interaction term odds ratio [OR] 1.17; 95% CI, 1.13-1.20 for Asian, OR 1.04; 95% CI, 1.02-1.06 for Black, and OR 1.10; 95% CI, 1.08-1.12 for Latino/Hispanic; P < .001 for each). Within each racial and ethnic group, individuals with low income were associated with a higher prevalence of multimorbidity compared with individuals in the same age group who had middle/high income (Supplementary Figure 7, available online).
      Supplementary Figure 1
      Supplementary Figure 1Study population flowchart. The mutually exclusive racial/ethnic subgroups were created based on the self-reported primary race and ethnicity combination.a These excluded individuals also did not identify as Latino/Hispanic.
      Supplementary Figure 2
      Supplementary Figure 2Unadjusted age distribution by race and ethnicity. NH = non-Hispanic.
      Supplementary Figure 3
      Supplementary Figure 3Unadjusted distribution of number of chronic conditions by race and ethnicity.
      Supplementary Figure 4
      Supplementary Figure 4Trends in adjusted prevalence of individual conditions by race and ethnicity, 1999-2018. CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease.
      Supplementary Figure 4
      Supplementary Figure 4Trends in adjusted prevalence of individual conditions by race and ethnicity, 1999-2018. CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease.
      Supplementary Figure 5
      Supplementary Figure 5Trends in adjusted multimorbidity prevalence by (A) race, ethnicity, and income and (B) by income alone, 1999-2018. Annual family income was categorized relative to the respective year's Federal Poverty Limit from the US Census Bureau into middle/high income (≥200%) and low income (<200%). The weighted proportion of individuals with annual income <200% of the Federal Poverty Limit was estimated using multiple imputation. LI = low income; MHI = middle/high income.
      Supplementary Figure 6
      Supplementary Figure 6Number of chronic conditions by age among Asian, Black, Latino/Hispanic, and White people. Data source is the National Health Interview Survey from 1999 to 2018. Estimates were obtained using ordered logistic regression models, adjusting for sex and region (details in Methods).
      Supplementary Figure 7
      Supplementary Figure 7Adjusted multimorbidity prevalence by (A) age, race, ethnicity, and income, and by (B) age and income. Annual family income was categorized relative to the respective year's Federal Poverty Limit from the US Census Bureau into middle/high income (≥200%) and low income (<200%). The weighted proportion of individuals with annual income <200% of the Federal Poverty Limit was estimated using multiple imputation. LI = low income; MHI = middle/high income.

      Sensitivity Analysis

      Results from our sensitivity analysis, including arthritis in the chronic conditions count (in the years it was available), were mostly consistent with our main results, with some notable differences (Supplementary Figures 8-11, available online). In this sensitivity analysis, the overall estimated prevalence of multimorbidity was 27.2% (95% CI, 27.0%-27.5%), and the adjusted prevalence difference between Black and White individuals decreased significantly from 2002 to 2018 (Supplementary Figure 8).
      Supplementary Figure 8
      Supplementary Figure 8Sensitivity analysis including arthritis: trends in adjusted multimorbidity prevalence by race and ethnicity, 2002-2018.
      Supplementary Figure 9
      Supplementary Figure 9Sensitivity analysis including arthritis: number of chronic conditions by age among Asian, Black, Latino/Hispanic, and White people.
      Supplementary Figure 10
      Supplementary Figure 10Sensitivity analysis including arthritis: (A) adjusted multimorbidity prevalence by age, race, and ethnicity, and (B) prevalence differences by age, race, and ethnicity.
      Supplementary Figure 11
      Supplementary Figure 11Sensitivity analysis: trends in adjusted prevalence of arthritis by race and ethnicity, 2002-2018.

      Discussion

      From 1999 to 2018, there was an increasing multimorbidity prevalence and a static racial gap in which Black individuals persistently had the highest prevalence and Asian individuals had the lowest. Moreover, multimorbidity prevalence was higher among Black individuals compared with Asian, Latino/Hispanic, and White individuals of the same age range, with differences emerging among those 30-34 years old and peaking among those 60-64 years old. Overall, and among individuals of the same racial and ethnic group, those with low income experienced a higher burden of multimorbidity. Studies have reported racial/ethnic disparities but, to the best of our knowledge, this is the first to analyze whether there was improvement during a recent 2-decade period notable for national attention on reducing health disparities.
      Our findings expand the literature in several ways. First, we found no improvement in the race/ethnicity gap in multimorbidity. The increasing prevalence of multimorbidity occurred almost in parallel across each of the racial and ethnic groups. These findings reveal a lack of progress in narrowing racial/ethnic disparities in health over a 20-year period, in which Black individuals persistently had the highest multimorbidity prevalence despite significant health care investments, substantial increases in per capita health care costs,

      Centers for Medicare & Medicaid Services. National Health Expenditure Data, Historical. Available at:https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical. Accessed March 17, 2022.

      ,

      Kamal R, McDermott D, Cox C. Peterson Center on Healthcare-Kaiser Family Foundation Health System Tracker. How has U.S. spending on healthcare changed over time? Available at:https://www.healthsystemtracker.org/chart-collection/u-s-spending-healthcare-changed-time/. Accessed March 17, 2022.

      and major nationwide policies aimed to address health disparities.
      • Koh HK
      • Graham G
      • Glied SA
      Reducing racial and ethnic disparities: the action plan from the Department of Health and Human Services.
      Notably, multimorbidity was consistently less prevalent among Asian individuals. Although unclear, this may be due to generally more favorable socioeconomic conditions among Asian individuals
      • Zhang Y
      • Misra R
      • Sambamoorthi U
      Prevalence of multimorbidity among Asian Indian, Chinese, and Non-Hispanic White Adults in the United States.
      or an underestimation of their true multimorbidity prevalence due to barriers to health care access and utilization that may prevent the detection of chronic conditions.
      • Lee S
      • Martinez G
      • Ma GX
      • et al.
      Barriers to health care access in 13 Asian American communities.
      Second, we found that the racial and ethnic gap was concentrated among non-elderly adults (<65 years old), emerging among those aged 30-34 years, and peaking among those 60-64 years old. It is unclear if such age-related divergence in multimorbidity prevalence may be due to a cumulative effect of a higher early life concentration of risk factors.
      • Mahajan S
      • Caraballo C
      • Lu Y
      • et al.
      Trends in differences in health status and health care access and affordability by race and ethnicity in the United States, 1999-2018.
      ,
      • Bailey ZD
      • Krieger N
      • Agénor M
      • Graves J
      • Linos N
      • Bassett MT
      Structural racism and health inequities in the USA: evidence and interventions.
      The convergence among the elderly, on the other hand, may be explained by a combination of a ceiling effect in chronic conditions and differences in risk of death associated with earlier multimorbidity among Black individuals.
      Third, we found that stratification by income level attenuated the disparities. This observation underscores the role of income inequality in the perpetuation of racial and ethnic disparities in health.
      • Bor J
      • Cohen GH
      • Galea S
      Population health in an era of rising income inequality: USA, 1980-2015.
      This finding is especially important given the disproportionately large percentage of Black individuals in the low-income stratum,

      Bhutta N, Chang AC, Dettling LJ, Hsu JW. FEDS Notes: Disparities in wealth by race and ethnicity in the 2019 Survey of Consumer Finances. Washington: Board of Governors of the Federal Reserve System, September 28, 2020. Available at:https://www.federalreserve.gov/econres/notes/feds-notes/disparities-in-wealth-by-race-and-ethnicity-in-the-2019-survey-of-consumer-finances-20200928.htm. Accessed March 17, 2022.

      a group that during the study period also reported the highest prevalence of poor or fair health.
      • Mahajan S
      • Caraballo C
      • Lu Y
      • et al.
      Trends in differences in health status and health care access and affordability by race and ethnicity in the United States, 1999-2018.
      There are some differences between our findings and the results from other studies that have measured multimorbidity using data from the NHIS. First, the overall unadjusted weighted prevalence of multimorbidity in our study (19.9%) was lower than the estimates from others.
      • Boersma P
      • Black LI
      • Ward BW
      Prevalence of multiple chronic conditions among US adults, 2018.
      • Ward BW
      • Schiller JS
      • Goodman RA
      Peer reviewed: multiple chronic conditions among us adults: a 2012 update.
      • Ward BW
      • Schiller JS
      Prevalence of multiple chronic conditions among US adults: estimates from the National Health Interview Survey, 2010.
      This discrepancy is likely due to the inclusion of arthritis in their calculation of multimorbidity, which was inconsistently available across the 20 years of this study. In an exploratory estimation of multimorbidity using data from 2002 to 2018 including arthritis, the overall estimates were consistent with those from previous reports and did not affect the overall findings. Second, our annual estimates of multimorbidity were adjusted for age, sex, and region, which can provide greater insight into disparities by race and ethnicity.
      This study has important implications. First, it shows that the public health challenge of multimorbidity is increasing among the 4 largest race and ethnicity groups in the United States, underscoring the growing need for preventive measures nationwide. Multimorbidity is associated with worse health status,
      • Arokiasamy P
      • Uttamacharya U
      • Jain K
      • et al.
      The impact of multimorbidity on adult physical and mental health in low- and middle-income countries: what does the study on global ageing and adult health (SAGE) reveal?.
      faster functional decline,
      • Ryan A
      • Wallace E
      • O'Hara P
      • Smith SM
      Multimorbidity and functional decline in community-dwelling adults: a systematic review.
      lower quality of life,
      • Arokiasamy P
      • Uttamacharya U
      • Jain K
      • et al.
      The impact of multimorbidity on adult physical and mental health in low- and middle-income countries: what does the study on global ageing and adult health (SAGE) reveal?.
      higher mortality,
      • Nunes BP
      • Flores TR
      • Mielke GI
      • Thumé E
      • Facchini LA
      Multimorbidity and mortality in older adults: A systematic review and meta-analysis.
      higher rates of unmet medical needs,
      • Johnson PJ
      • Mentzer KM
      • Jou J
      • Upchurch DM
      Unmet healthcare needs among midlife adults with mental distress and multiple chronic conditions.
      and higher health care costs,
      • Wang L
      • Si L
      • Cocker F
      • Palmer AJ
      • Sanderson K
      A systematic review of cost-of-illness studies of multimorbidity.
      and this study shows that the burden has been persistently higher among Black individuals.
      Although preventing the development of multimorbidity should be a public health priority, these findings suggest that complementary and targeted strategies are needed for Black individuals. The persistent gap in multimorbidity over the study period may be associated with the persistent disparity in access to preventive health care services that Black individuals experience.
      • Mahajan S
      • Caraballo C
      • Lu Y
      • et al.
      Trends in differences in health status and health care access and affordability by race and ethnicity in the United States, 1999-2018.
      Also, stressors derived from race-based discrimination are associated with a higher burden of chronic diseases and multimorbidity.
      • Mays VM
      • Cochran SD
      • Barnes NW
      Race, race-based discrimination, and health outcomes among African Americans.
      ,
      • Oh H
      • Glass J
      • Narita Z
      • Koyanagi A
      • Sinha S
      • Jacob L
      Discrimination and multimorbidity among black Americans: findings from the National Survey of American Life.
      Thus, our findings should be interpreted in the context of the vast literature that identifies systemic racism as a fundamental determinant of health inequalities through the reinforcement of an artificial racial and ethnic hierarchy that places Black individuals at a disadvantage in obtaining access to goods and opportunities necessary for a healthy life. Public health strategies should account for this in the distribution of resources and efforts to prevent the development of chronic conditions over a lifetime. Particularly, the emergence of the disparities among young adults suggests that such strategies must aim for the prevention of the onset of chronic conditions early in life.
      Our study has several limitations. First, the cross-sectional design prevents us from measuring and comparing the incidence of multimorbidity across the lifetime of an individual by racial and ethnic groups. Furthermore, when analyzing the differences with age, the study design may be introducing survivor bias as age group increases, which may underestimate the overall impact of multimorbidity on health disparities. Second, we used a limited number of chronic conditions based on their availability in the NHIS and lacked information on the severity, complexity, duration, and status of such conditions. Third, our multimorbidity measure may be an underestimation because the study relied on self-reported conditions that may go underreported due to barriers to access to care—for its diagnosis—and low health literacy.

      Conclusions

      From 1999 to 2018, multimorbidity in the United States increased, and Black individuals persistently had the highest prevalence of multimorbidity, with racial gaps that did not change. When analyzed by age, this gap emerged among young adults and peaked among those aged 60-64 years. Public health interventions that prevent the onset of chronic conditions in early life may be needed to eliminate these disparities.

      Data Sharing

      All data used in this study are publicly available at the Integrated Public Use Microdata Series Health Surveys (https://nhis.ipums.org/). The codes used to analyze these data are publicly available at http://doi.org/10.3886/E151983V1.

      Supplementary Methods

      About the National Health Interview Survey

      The National Health Interview Survey (NHIS) is a series of annual cross-sectional national surveys that provide information on the health of the noninstitutionalized population of the United States. The sample design uses a multistage area probability design, which adjusts for nonresponse and further allows for national representative sampling of households and individuals, including underrepresented groups. This survey consists of a questionnaire divided into 4 cores: Household Composition, Family Core, Sample Child Core, and Sample Adult Core. The Household Composition file collects basic and relationship information about all individuals in a household. The Family Core file collects sociodemographic characteristics, basic indicators of health status, activity limitations, injuries, health insurance coverage, and access to and utilization of health care services. From each family, one sample child and one sample adult are randomly selected to gather more in-depth information for the Sample Child Core and Sample Adult Core, respectively. In our study, we used data from the Sample Adult Core files, complemented with demographic and socioeconomic characteristics.

      Statistical Analysis

      The annual multimorbidity prevalence for each group was estimated using multivariable logistic regression models. In these models, multimorbidity was the dependent variable, and age, sex, a dummy variable for each region, and an indicator for each year of interview were the independent variables. Age, sex, and region were centered on their overall mean for the study sample; the coefficients for each year, when combined with the intercept, then represented the logit of the annual multimorbidity rates adjusted for age, sex, and region. A separate model was estimated for each racial and ethnic subgroup and the results were used to generate estimated rates for each year, using the inverse logit of each year effect as the annual rate and applying the method of parametric bootstrapping to calculate the standard error and the CI for the transformed coefficients.
      US Department of Health Human Services
      Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions.
      To assess the trends in the prevalence of each of the 9 conditions, we used the same approach as for multimorbidity overall.
      The multimorbidity prevalence and racial/ethnic differences trends over the study period were estimated by fitting weighted linear regression models where the dependent variable was the adjusted annual multimorbidity prevalence or difference, and the independent variable was time in years. Each observation was weighted by the inverse square of the SE to account for varying precision of each estimated rate or the difference over time.
      In the linear and ordered logistic regression models used to estimate the mean and proportion of individuals with 0, 1, 2, 3, or ≥4 chronic conditions over the years, respectively, the number of conditions was the dependent variable and age, sex, region, and an indicator for each year were the independent variables, obtaining the estimates for each race/ethnicity group separately.
      To evaluate if the association between race and ethnicity and multimorbidity differs by age, we used a logistic regression model with multimorbidity as the dependent variable, region, sex, age, and race/ethnicity as independent variables, and an interaction term between age and race and ethnicity. We reported the interaction term odds ratio for each race and ethnicity group. We replicated this using an ordered logit model and the number of chronic conditions as the dependent variable, and age group, region, and sex as the independent variables to estimate the percentage with 0, 1, 2, 3, and ≥4 conditions in each age group by race and ethnicity.
      Following recommendations from the National Center for Health Statistics for multiply imputed data analysis,
      • Goodman RA
      • Posner SF
      • Huang ES
      • Parekh AK
      • Koh HK
      Defining and measuring chronic conditions: imperatives for research, policy, program, and practice.
      to estimate the annual low-income prevalence by race and ethnicity we used the mean annual estimate obtained by separate logistic regressions using a similar approach as above, but with each of the multiply imputed low-income variables as the dependent variable and an indicator for each year as the independent variables. Similarly, we used the mean prevalence estimate of multimorbidity obtained from separate regressions for each of the income groups.
      Supplementary References
      • 1.
        King G, Tomz M, Wittenberg J. Making the most of statistical analyses: Improving interpretation and presentation. Am J Pol Sci. 2000;44(2):347-361.
      • 2.
        National Center for Health Statistics. Multiple imputation of family income and personal earnings in the National Health Interview Survey: methods and examples. August 2019. Available at: https://nhis.ipums.org/nhis/resources/tecdoc18.pdf. Accessed October 15, 2021.

      References

        • US Department of Health Human Services
        Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions.
        US Department of Health and Human Services, Washington, DC2010
        • Goodman RA
        • Posner SF
        • Huang ES
        • Parekh AK
        • Koh HK
        Defining and measuring chronic conditions: imperatives for research, policy, program, and practice.
        Prev Chronic Dis. 2013; 10: E66
        • King DE
        • Xiang J
        • Pilkerton CS
        Multimorbidity trends in United States adults, 1988-2014.
        J Am Board Fam Med. 2018; 31: 503-513
        • Johnson-Lawrence V
        • Zajacova A
        • Sneed R
        Education, race/ethnicity, and multimorbidity among adults aged 30-64 in the National Health Interview Survey.
        SSM Popul Health. 2017; 3: 366-372
      1. United States Department of Health and Human Services. Task Force on Black and Minority Health. Report of the Secretary's Task Force on Black and Minority Health. Report of the Secretary's Task Force on Black & Minority Health. 1985. Available at: http://resource.nlm.nih.gov/8602912. Accessed March 17, 2022.

        • Blewett LA
        • Rivera Drew JA
        • King ML
        • Williams KCW
        IPUMS health surveys: National Health Interview Survey, version 6.4.
        IPUMS, Minneapolis, MN2019 (Available at: https://doi.org/10.18128/D070.V6.4. Accessed March 17, 2022)
      2. United States Census Bureau. People in families by family structure, age, and sex, iterated by income-to-poverty ratio and race. Available at:https://www.census.gov/data/tables/time-series/demo/income-poverty/cps-pov/pov-02.html#par_textimage_30. Accessed March 17, 2022.

        • Mahajan S
        • Caraballo C
        • Lu Y
        • et al.
        Trends in differences in health status and health care access and affordability by race and ethnicity in the United States, 1999-2018.
        JAMA. 2021; 326: 637-648
        • Drye EE
        • Altaf FK
        • Lipska KJ
        • et al.
        Defining multiple chronic conditions for quality measurement.
        Med Care. 2018; 56: 193-201
      3. Division of Health Interview Statistics National Center for Health Statistics. Multiple imputation of family income and personal earnings in the National Health Interview Survey: methods and examples. August 2019. Available at:https://nhis.ipums.org/nhis/resources/tecdoc18.pdf. Accessed March 17, 2022.

        • Barnett K
        • Mercer SW
        • Norbury M
        • Watt G
        • Wyke S
        • Guthrie B
        Epidemiology of multimorbidity and implications for health care, research, and medical education: a cross-sectional study.
        Lancet. 2012; 380: 37-43
        • Boersma P
        • Black LI
        • Ward BW
        Prevalence of multiple chronic conditions among US adults, 2018.
        Prev Chronic Dis. 2020; 17: E106
        • Ward BW
        • Schiller JS
        • Goodman RA
        Peer reviewed: multiple chronic conditions among us adults: a 2012 update.
        Prev Chronic Dis. 2014; 11: E62
        • Ward BW
        • Schiller JS
        Prevalence of multiple chronic conditions among US adults: estimates from the National Health Interview Survey, 2010.
        Prev Chronic Dis. 2013; 10: E65
      4. Centers for Disease Control and Prevention. National Health Interview Survey, 1997-2018. Survey description document. Available at:https://www.cdc.gov/nchs/nhis/1997-2018.htm. Accessed March 17, 2022.

      5. Centers for Medicare & Medicaid Services. National Health Expenditure Data, Historical. Available at:https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsHistorical. Accessed March 17, 2022.

      6. Kamal R, McDermott D, Cox C. Peterson Center on Healthcare-Kaiser Family Foundation Health System Tracker. How has U.S. spending on healthcare changed over time? Available at:https://www.healthsystemtracker.org/chart-collection/u-s-spending-healthcare-changed-time/. Accessed March 17, 2022.

        • Koh HK
        • Graham G
        • Glied SA
        Reducing racial and ethnic disparities: the action plan from the Department of Health and Human Services.
        Health Aff (Millwood). 2011; 30: 1822-1829
        • Zhang Y
        • Misra R
        • Sambamoorthi U
        Prevalence of multimorbidity among Asian Indian, Chinese, and Non-Hispanic White Adults in the United States.
        Int J Environ Res Public Health. 2020; 17: 3336
        • Lee S
        • Martinez G
        • Ma GX
        • et al.
        Barriers to health care access in 13 Asian American communities.
        Am J Health Behav. 2010; 34: 21-30
        • Bailey ZD
        • Krieger N
        • Agénor M
        • Graves J
        • Linos N
        • Bassett MT
        Structural racism and health inequities in the USA: evidence and interventions.
        Lancet. 2017; 389: 1453-1463
        • Bor J
        • Cohen GH
        • Galea S
        Population health in an era of rising income inequality: USA, 1980-2015.
        Lancet. 2017; 389: 1475-1490
      7. Bhutta N, Chang AC, Dettling LJ, Hsu JW. FEDS Notes: Disparities in wealth by race and ethnicity in the 2019 Survey of Consumer Finances. Washington: Board of Governors of the Federal Reserve System, September 28, 2020. Available at:https://www.federalreserve.gov/econres/notes/feds-notes/disparities-in-wealth-by-race-and-ethnicity-in-the-2019-survey-of-consumer-finances-20200928.htm. Accessed March 17, 2022.

        • Arokiasamy P
        • Uttamacharya U
        • Jain K
        • et al.
        The impact of multimorbidity on adult physical and mental health in low- and middle-income countries: what does the study on global ageing and adult health (SAGE) reveal?.
        BMC Med. 2015; 13: 178
        • Ryan A
        • Wallace E
        • O'Hara P
        • Smith SM
        Multimorbidity and functional decline in community-dwelling adults: a systematic review.
        Health Qual Life Outcomes. 2015; 13: 168
        • Nunes BP
        • Flores TR
        • Mielke GI
        • Thumé E
        • Facchini LA
        Multimorbidity and mortality in older adults: A systematic review and meta-analysis.
        Arch Gerontol Geriatr. 2016; 67: 130-138
        • Johnson PJ
        • Mentzer KM
        • Jou J
        • Upchurch DM
        Unmet healthcare needs among midlife adults with mental distress and multiple chronic conditions.
        Aging Ment Health. 2022; 26: 775-783
        • Wang L
        • Si L
        • Cocker F
        • Palmer AJ
        • Sanderson K
        A systematic review of cost-of-illness studies of multimorbidity.
        Appl Health Econ Health Policy. 2018; 16: 15-29
        • Mays VM
        • Cochran SD
        • Barnes NW
        Race, race-based discrimination, and health outcomes among African Americans.
        Annu Rev Psychol. 2007; 58: 201-225
        • Oh H
        • Glass J
        • Narita Z
        • Koyanagi A
        • Sinha S
        • Jacob L
        Discrimination and multimorbidity among black Americans: findings from the National Survey of American Life.
        J Racial Ethn Health Disparities. 2021; 8: 210-219