Lim et al.

BMC Public Health (2024) 24:1153 BMC Public Health

https://doi.org/10.1186/s12889-024-18613-8

RESEARCH Open Access

Social determinants of health and emergency

department visits among older adults
with multimorbidity: insight from 2010
to 2018 National Health Interview Survey
Arum Lim1*, Chitchanok Benjasirisan1, Xiaoyue Liu2, Oluwabunmi Ogungbe1, Cheryl Dennison Himmelfarb1,
Patricia Davidson3 and Binu Koirala1

Abstract
Background Multimorbidity is prevalent among older adults and is associated with adverse health outcomes,
including high emergency department (ED) utilization. Social determinants of health (SDoH) are associated with
many health outcomes, but the association between SDoH and ED visits among older adults with multimorbidity has
received limited attention. This study aimed to examine the association between SDoH and ED visits among older
adults with multimorbidity.
Methods A cross-sectional analysis was conducted among 28,917 adults aged 50 years and older from the 2010
to 2018 National Health Interview Survey. Multimorbidity was defined as the presence of two or more self-reported
diseases among 10 common chronic conditions, including diabetes, hypertension, asthma, stroke, cancer, arthritis,
chronic obstructive pulmonary disease, and heart, kidney, and liver diseases. The SDoH assessed included race/
ethnicity, education level, poverty income ratio, marital status, employment status, insurance status, region of
residence, and having a usual place for medical care. Logistic regression models were used to examine the association
between SDoH and one or more ED visits.
Results Participants’ mean (± SD) age was 68.04 (± 10.66) years, and 56.82% were female. After adjusting for age, sex,
and the number of chronic conditions in the logistic regression model, high school or less education (adjusted odds
ratio [AOR]: 1.10, 95% confidence interval [CI]: 1.02–1.19), poverty income ratio below the federal poverty level (AOR:
1.44, 95% CI: 1.31–1.59), unmarried (AOR: 1.19, 95% CI: 1.11–1.28), unemployed status (AOR: 1.33, 95% CI: 1.23–1.44),
and having a usual place for medical care (AOR: 1.46, 95% CI 1.18–1.80) was significantly associated with having one
or more ED visits. Non-Hispanic Black individuals had higher odds (AOR: 1.28, 95% CI: 1.19–1.38), while non-Hispanic
Asian individuals had lower odds (AOR: 0.71, 95% CI: 0.59–0.86) of one or more ED visits than non-Hispanic White
individuals.

*Correspondence:
Arum Lim
alim19@jh.edu
Full list of author information is available at the end of the article

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data.
Lim et al. BMC Public Health (2024) 24:1153 Page 2 of 10

Conclusion SDoH factors are associated with ED visits among older adults with multimorbidity. Systematic
multidisciplinary team approaches are needed to address social disparities affecting not only multimorbidity
prevalence but also health-seeking behaviors and emergent healthcare access.
Keywords Emergency department, Multimorbidity, NHIS, Social determinants of health

Background multimorbidity may disproportionally occur due to social

Multimorbidity is defined as the co-existence of two or disparities. Multimorbidity was related to low educa-
more chronic conditions [1]. The number of people liv- tional attainment [14] and was more prevalent among
ing with multimorbidity is dramatically increasing world- the Black population and less prevalent among Asian and
wide with the growing aging population and improved Hispanic populations compared to the White population
diagnostic capabilities [2, 3]. According to the pooled in the United States (US) [15]. Moreover, emerging evi-
data from a meta-analysis of studies published between dence has suggested that socioeconomic disadvantages
2000 and 2021, the prevalence of multimorbidity was worsen the burden of multimorbidity in older adults [16].
37.2% globally and 43.1% in North America [3]. Since Particularly, low income is consistently associated with
chronic diseases are usually accompanied by aging, 51% not only a higher prevalence of multimorbidity but also
of adults aged 60 years and older had multimorbidity in worse patient-reported health outcomes in older multi-
the global population [3]. Multimorbidity is also a prob- morbid patients [15, 17]. In this context, education, race/
lem for middle-aged adults, as multimorbidity stiffly ethnicity, and income can be tied up as social determi-
increases after age 50 [4], and 47% of adults 50 years and nants of health (SDoH). In general, SDoH consists of five
older have multimorbidity [3]. In support of this, recent domains: economic stability, education access and qual-
studies have extended their focus to individuals aged 50 ity, healthcare access and quality, neighborhood and built
years and older to investigate multimorbidity and chronic environment, and social and community context [18].
disease burden [5, 6]. Multimorbidity has become a sig- Although gender, education, and health system were the
nificant health issue because of the increasing complexity most frequently investigated as SDoH in older multimor-
of healthcare needs [7]. For example, people with mul- bid populations, limited attention is paid to race/ethnic-
timorbidity need a multidisciplinary approach to deci- ity, socioeconomic status, and political context in the
sion-making for the treatment and management of each current literature [16]. These indicate that investigating
condition and may have to deal with polypharmacy and individuals living with multimorbidity and their SDoH
communication with multiple health providers [1]. Thus, factors associated with healthcare access, particularly
healthcare providers and researchers have been paying access to emergency services, is necessary to understand
attention to prioritizing the complex needs of care for which social context is related to managing multiple
individuals with multimorbidity [1, 7]. chronic conditions leading to ED visits.
Multimorbidity is associated with adverse health out- Generally, ED visits can be considered as health care
comes, such as increased hospital utilization, major needs caused by sudden symptoms, deterioration, or
health decline, and mortality [8–10]. People with multi- injuries. The frequent complaints leading older people to
morbidity frequently contact general practitioners and visit EDs are shortness of breath, chest pain, and lower
visit emergency departments (ED) due to their complex extremity pain/injury, and in approximately 75% of ED
care needs [11, 12]. A study analyzing large electronic visits, older adults were triaged as urgent/emergent [13].
health record data in the Netherlands reported that 11% However, disparities in ED care access and triage pro-
of individuals with multimorbidity had ≥ 12 general prac- cesses exist based on race/ethnicity and health insurance
titioner contacts, and 12% had ED visits in a year [11]. status [19]. There is a need to identify and address dis-
However, the group of patients who frequently contact parities in emergency healthcare access in older people,
general practice would be distinct from those who visit which may produce disproportionated health outcomes
ED. In the previous study, only 29% of people with fre- [19]. Therefore, this study aims to describe the associa-
quent general practice contacts had ED visits [11]. This tion between SDoH and ED visits among older adults
implies that people who do not frequently visit general with multimorbidity.
practice may be more likely to visit ED. In addition, gen-
erally, the intensity of ED resource utilization increased Methods
with age [13], and people who visited EDs had more Study design and data source
chronic conditions and more prescribed medications The study employed a cross-sectional approach to
than the entire multimorbid group [11]. examine the data from 2010 to 2018 National Health
Despite the consistently increased prevalence of mul- Interview Survey (NHIS) conducted by the National
timorbidity for all racial/ethnic groups, the risk for Center for Health Statistics (NCHS) [20]. The NHIS is a
Lim et al. BMC Public Health (2024) 24:1153 Page 3 of 10

cross-sectional population-based survey for non-institu- residence, and having a usual place to go for medical care
tionalized civilians aged 18 years or older US adults. The when sick. Some variables were defined as dichotomous:
data was gathered by face-to-face interviewing with one marital status (currently married/not married, includ-
randomly selected adult per household for the Sample ing never married, divorced, widowed, or separated);
Adult Module. The interview questions covered health- employment status (employed/unemployed), insur-
care services, behaviors, and health status. Detailed infor- ance status (insured/uninsured), and have a usual place
mation regarding the design and methodology of NHIS to go for medical care when sick–a proxy for healthcare
is published elsewhere [20, 21]. This study restricted the access (yes/no). Race/ethnicity was categorized as (non-
study period from 2010 to 2018 to avoid the potential Hispanic White, non-Hispanic Black, non-Hispanic
confounding effects of the COVID-19 pandemic that Asian, and Hispanic). Educational status was categorized
occurred in December 2019 on SDoH and ED visits [22]. by ≤ high school, some college, and ≥ bachelor’s degree.
This study was exempt from institutional review board The poverty income ratio (PIR) was used as a proxy of
review because it used publicly available de-identified financial status. The midpoint of an individual’s family
data published by the NCHS. income was divided by the poverty threshold for the year.
The variable was then categorized as < 1, between 1 and
Sample 1.99, and ≥ 2. A PIR less than one means that the indi-
Inclusion/exclusion criteria vidual income is below the federal poverty level, a PIR
Individuals aged 50 years and older, those who had more between 1 and 1.99 indicates the income is between 100%
than two chronic conditions defined as multimorbid- and 199% of the poverty level, and a PIR greater than two
ity, and those with available ED visit data were included means that the income is more than 200% of the poverty
in this study. The number of chronic conditions was level. The region of residence had four categories: north-
obtained from self-reported disease diagnoses. A total of east, midwest, south, and west. Perceived health status
10 chronic conditions were selected to define multimor- was categorized on a five-scale from “excellent” to “poor.”
bidity, which were collected throughout the 2010–2018
study period. The selected chronic conditions are parts of Covariates
conditions defined based on the National Quality Forum We included three covariates: age in years (measured
Multiple Chronic Conditions framework and aligned as a continuous variable, further categorized as 50–64
with conditions in a study that used this framework to years or 65 and older years), sex (categorized as male or
define multimorbidity [23, 24]. Having chronic condi- female), and the number of chronic conditions. Based
tions was identified by the questions asking if the respon- on the rationale that having two or more chronic condi-
dents had ever been told by a healthcare professional that tions defines multimorbidity, and having three or more
they had diabetes, hypertension, asthma, stroke, cancer, is considered complex multimorbidity [4], the number
arthritis, chronic obstructive pulmonary disease (emphy- of chronic conditions was categorized into 2, 3–4, or ≥ 5
sema or chronic bronchitis), or heart disease (coronary conditions. The category will help stratify the partici-
artery disease, myocardial infarction, angina, or other pants by the severity of multimorbid conditions [25, 26].
heart conditions), or had been told in the past 12 months
that they had weak/failing kidneys or any liver condition. Statistical analysis
Therefore, participants could have between two and ten This study merged NHIS data from 2010 to 2018 and
chronic conditions. applied sampling weights according to NCHS guidelines
[27]. Sociodemographic characteristics were presented
Measurements using descriptive statistics, mean and standard deviation,
Emergency department visits and percentage. Differences and associations in charac-
The study outcome was one or more ED visits in the teristics between respondents with more than 1 ED visit
previous 12 months. Respondents were asked, “During and those without ED visits were examined by survey-
the past 12 months, how many times have you gone to a weighted t-tests for continuous variables and chi-square
hospital emergency room about your own health?” This tests for categorical variables.
includes emergency room visits that resulted in hospital This study used survey-weighted multivariable logis-
admission. The responses were dichotomized as having tic regression to test the association between SDoH
either one or more instances or none. and ED visits within the previous 12 months in people
with multimorbidity. Model 1 included multiple SDoH
Social determinants of health factors, including race/ethnicity, education, income,
The SDoH variables included in this study were race/eth- employment, insurance, marital status, and region of
nicity, marital status, employment and educational status, residence, to show the association of each SDoH factor
poverty income ratio, health insurance status, region of with ED visits, adjusting for all other SDoH effects. Age
Lim et al. BMC Public Health (2024) 24:1153 Page 4 of 10

and sex variables were added in Model 2, and the num- age and sex in Model 2, the associations of SDoH with
ber of chronic conditions in Model 3. Cases with at least ED visits were still preserved. Model 3 adjusted for the
one missing data in any variable were deleted from the number of chronic conditions, which was a proxy of the
analysis, which may cause bias if the missing is not ran- severity of diseases, and all associations were still sig-
dom [28]. Statistical significance was set as a two-sided nificant as Model 1 and 2. People who were not married
α < 0.05. All statistical analyses were conducted using the (Adjusted Odd Ratio [AOR]: 1.19, 95% Confidence Inter-
Stata© SE statistical software. val [CI]: 1.11–1.28), non-Hispanic Black people (AOR:
1.28, 95% CI: 1.19–1.38), had high school education or
Results less (AOR: 1.10, 95% CI: 1.02–1.19), had lower PIR (AOR:
Sample characteristics 1.44, 95% CI: 1.31–1.59), were unemployed (AOR: 1.33,
A total of 28,917 respondents living with two or more of 95% CI: 1.23–1.44), and had a usual place for medical
the 10 chronic conditions were included in the analysis. care (AOR: 1.46, 95% CI: 1.18–1.80) were more likely to
Among them, 68% (n = 19,661) had no ED visit, while the visit ED at least once in the prior 12 months, compared
remaining 32% (n = 9,256) had at least one ED visit in the to their reference groups. The adjusted findings are pre-
previous 12 months. The participants’ mean age (± SD) sented in Table 2.
was 68 (± 10.7). The ED visits group had more female
participants (58.5%) and adults who were not married Discussion
(64.7%) than the no ED visit group (56.0% and 56.0%, This study presented multiple SDoH factors associated
respectively). The ED visits group had more non-His- with ED visits among older people with multimorbid-
panic Black people (15.8% vs. 11.4%) and Hispanic par- ity. Particularly, people who were non-Hispanic Black
ticipants (8.2% vs. 7.2%) than the no ED visit group and people, not married, had poor financial conditions, and
were more likely to have a high school or lower education lower education levels showed higher odds of ED visits.
(53.6% vs. 46.8%) and be unemployed (77.1% vs. 67.9%). This study demonstrated racial/ethnic disparity in ED
Moreover, they were more likely to have low PIR (22.3% visits among older adults with multimorbidity. In this
vs. 13.5%) and more likely to have a usual place for medi- study, non-Hispanic Black people were more likely to
cal care (97.8% vs. 97.0%) than their no ED visit coun- have at least one ED visit than other racial/ethnic popula-
terparts. Perceived health status was poorer (19.2% vs. tions, while non-Hispanic Asian people were less likely to
7.5%) in the ED visit group. Health insurance status did do so. Since multimorbidity was more prevalent among
not differ between the two groups. In terms of chronic Black people and less prevalent among Asian people [15],
conditions, the most frequently reported condition was this study result implied that race/ethnicity potentially
hypertension (83.0%), but there was no significant dif- deepened existing multimorbidity disparities through
ference between the two groups. In both groups, heart emergent healthcare access disparities. Similarly, the
disease (52.0% vs. 39.1%, p <.001) and diabetes (38.9% correlation between race/ethnicity and ED visits may
vs. 35.8%, p <.001) followed, and there were significant indicate existing disparities in multimorbidity status. A
differences in the prevalence of the chronic conditions study pointed out that Black individuals had a similar
between the ED visit group and no ED visit group. COPD prevalence of multimorbidity as other groups who were
was the fourth most prevalent chronic condition in the 5–10 years older, and there was no significant change in
ED visit group, with a higher prevalence than in the no multimorbidity prevalence between the Black and White
ED visit group (37.4% vs. 24.9%, p <.001). However, can- populations from 1999 to 2018 [15]. Factors contributing
cer was the fourth most prevalent chronic condition to this may include the accumulated effect of the health
in the no ED visit group, which was the only disease experiences with chronic conditions in early life, produc-
with a significantly higher prevalence in the no ED visit ing a gap in older age and leading to higher odds of ED
group than the ED visit group (32.1% vs. 30.4%, p <.001). visits. However, since our study results were produced
The average numbers of chronic conditions (± SD) were after other SDoH were adjusted, such as education levels
3.3 ± 1.31 in the ED visit group and 2.7 ± 0.99 in the no ED and financial status, it needs to be investigated in further
visit group, and the proportions of having more than five research to explore the other possible reasons for racial/
conditions were 9.7% and 6.3%, respectively. The charac- ethnic disparity in ED visits. After exploring the mecha-
teristics of the study population can be found in Table 1. nisms of deepening health disparities in the treatment
continuum, it is necessary to eliminate the disparities
Social determinants of health on ED visits led by early-onset chronic conditions and care processes
The variables representing SDoH (marital status, race/ through healthcare intervention and policy.
ethnicity, education, financial status, region of residence, The findings reported that people who were not mar-
and usual healthcare access) were included in Model 1 ried showed higher odds of ED visits than those who
without adjusting for other covariates. After adjusting for were married. Since this study merged responses
Lim et al. BMC Public Health (2024) 24:1153 Page 5 of 10

Table 1 Sociodemographic characteristics of older adults with multimorbidity by the number of ED visits (n = 28,917)
Characteristics Total No ED visits ≥ 1 ED visits P
Weighted, n 9 589 647 6 583 984 3 005 662
Unweighted, n 28 917 19 661 9 256
Age (years, mean ± SD) 68.04 ± 10.66 68.09 ± 9.88 67.94 ± 10.80 0.339
50–64 40.03 39.01 42.25 < 0.001
≥65 59.97 60.99 57.75
Sex (%) 0.001
Female 56.82 56.04 58.53
Male 43.18 43.96 41.47
Marital status (%) < 0.001
Not married 58.70 55.96 64.69
Race/ethnicity (%) <0.001
Non-Hispanic White 76.44 77.90 73.25
Hispanic 7.50 7.16 8.22
Non-Hispanic Black 12.79 11.42 15.81
Non-Hispanic Asian 2.41 2.73 1.69
Other races 0.87 0.79 1.03
Education (%) < 0.001
≥ Bachelor’s degree 22.18 24.00 18.19
Some college 28.89 29.18 28.25
≤ High school 48.93 46.82 53.55
Poverty-income ratio (PIR)*(%) < 0.001
PIR ≥ 2.00 59.08 63.20 50.05
PIR 1-1.99 24.71 23.34 27.70
PIR < 1 16.21 13.46 22.25
Employment status (%) < 0.001
Unemployed 70.79 67.93 77.08
Health insurance status (%) 0.934
Uninsured 4.31 4.30 4.33
Have usual place for medical care (%) 97.22 96.95 97.82 < 0.001
Region of residence (%)
Northeast 17.20 17.1 17.44 0.007
Midwest 24.36 23.85 25.50
South 39.39 39.43 39.30
West 19.04 19.63 17.76
Perceived health status (%) < 0.001
Excellent 6.81 8.27 3.61
Very Good 21.39 24.77 14.00
Good 35.27 37.25 30.94
Fair 25.35 22.20 32.24
Poor 11.18 7.51 19.21
Chronic conditions (%)
Diabetes 36.76 35.79 38.89 < 0.001
Hypertension 82.96 83.05 82.75 0.568
Asthma 25.36 23.95 28.42 < 0.001
Stroke 12.55 10.23 17.63 < 0.001
Cancer 31.54 32.05 30.44 0.022
Arthritis 16.88 14.47 22.15 < 0.001
COPD† 28.8 24.86 37.44 < 0.001
Heart disease‡ 43.13 39.10 51.95 < 0.001
Weak/failing kidneys in 12 m 8.22 6.21 12.64 < 0.001
Any liver conditions 3.86 3.24 5.19 < 0.001
Number of chronic conditions 2.90 ± 1.18 2.73 ± 0.99 3.27 ± 1.31 < 0.001
(mean ± SD)
Lim et al. BMC Public Health (2024) 24:1153 Page 6 of 10

Table 1 (continued)
Characteristics Total No ED visits ≥ 1 ED visits P
2 48.27 54.30 48.27 <0.001
3–4 42.09 39.37 42.09
≥5 9.65 6.33 9.65
SD, standard deviation; ED, emergency department; COPD, chronic obstructive pulmonary disease Weighted sample demographic and characteristics were
presented and used for inferential statistics. *PIR < 1 = below poverty level; PIR 1-1.99 = between 100-199% above poverty level; PIR ≥ 2 = 200% or above poverty level.
†
COPD is defined as ever being told they had COPD, emphysema, or chronic bronchitis. ‡Heart disease is defined as ever being told they had coronary heart disease,
heart attack, or other heart condition/disease

indicating not married, such as divorced, separated, and used the variable ‘having a usual place for medical care’
bereaved, as unmarried participants, people categorized as a proxy of health care access, one of the SDoH factors.
as not married may include those living alone. Thus, they Since it is assumed that people more likely to visit EDs
might lack caregivers, which increases their self-care bur- would have worse health conditions, they need to receive
den, as well as available resources, such as health insur- regular follow-ups to assess and manage their health con-
ance, given that married people are more likely to have ditions. It is reported that multimorbid people are likely
private insurance than unmarried people [29]. More- to spend more on healthcare costs, consequently mak-
over, married people are more likely to have social sup- ing them more vulnerable to cost-related non-adherence
port than those who are not given that marital status is to recommended treatment, resulting from financial
often used as a proxy for informal social support [30, 31]. strain [38, 39]. A study found that more than one-third
This finding aligns with previous studies that older adults of participants living with multimorbidity had not sought
living alone had higher odds of ED admission [32, 33], medical care or purchased medication due to cost [40].
and people with multimorbidity who live alone had sig- Non-adherence to recommended general healthcare vis-
nificantly longer inpatient days after ED admissions than its and medication may lead to worsened symptoms in
those without multimorbidity [32]. These findings may multimorbidity populations, which may result in higher
be supported by the fact that multimorbid people have odds of ED visits. In this context, their financial burden
more care needs due to the complexity of care, as well as should be assessed and managed to prevent non-adher-
greater disease and symptom burdens [34, 35]. This study ence to treatments and management of their multiple
also showed that the lowest education level was associ- chronic conditions, leading to unplanned ED visits due to
ated with higher odds of ED visits than the highest. This sudden deterioration. What is apparent is, however, that
can be related to the gap between high healthcare needs EDs are commonly the safety net of society [41]. A study
and capacity for self-management, given that education reported that the most represented reasons for referral to
is associated with the activation of self-management in social work in ED were financial concerns and resource
patients with multimorbidity [36, 37]. Moreover, lower counseling [41]. This indicates that EDs may play a role
education level was also related to the greater impact of as the safety net to prevent deepening the disparities in
multimorbidity on activities of daily living and mental SDoH among people with multimorbidity.
health, which may affect self-care [35]. Therefore, since Lastly, the COVID-19 pandemic has tremendously
people with multimorbidity have higher needs for self- influenced not only people’s SDoH, including employ-
care, SDoH factors related to self-care, such as marital ment status and socioeconomic level [42], but also
status and education levels, may explain the higher odds ED visits, such as the number of ED visits and hospital
of ED visits in unmarried and lowest education-level par- admission rate from ED [43]. Although this study does
ticipants. Based on this finding, improving self-care and not explain the impact of the COVID-19 pandemic on
health literacy and implementing social support models the association between SDoH and ED visits, we pro-
in older multimorbid populations may prevent worsening pose future studies that examine changes in the con-
health conditions, reducing ED visits. text of SDoH and emergency healthcare access pre- and
The study demonstrated that people who are unem- post-pandemic.
ployed and have lower PIR levels have higher odds of
ED visits. However, these findings need to be cautiously Limitations
interpreted due to the cross-sectional study design. It This study has significance, given that it used large-scale,
could be explained that older people living with multi- nationally representative data to strengthen generaliz-
morbidity who visit EDs at least once in the previous 12 ability. Moreover, to our knowledge, this is the first study
months are more likely to lose or quit their jobs or have to examine the association between SDoH and ED visits
not gotten a chance to be hired due to their poor health in older multimorbid populations. However, this study
conditions. It also influences poverty levels, making acknowledges the following limitations. First, multimor-
lower PIR associated with ED visits. Moreover, this study bidity criteria did not include mental health problems,
Lim et al. BMC Public Health (2024) 24:1153 Page 7 of 10

Table 2 Logistic regression analyses of the associations between social determinants of health and having ≥ 1 ED visit in the prior 12
months among older adults with multimorbidity (N = 28,917)
Social determinants of health Model 1a Model 2b Model 3c
AOR (95% CI) AOR (95% CI) AOR (95% CI)
Age (years)
50–64 - Ref Ref
≥ 65 - 0.76 (0.71–0.82) * 0.77 (0.71–0.82) *
Sex
Female - Ref Ref
Male - 1.01 (0.95–1.07) 1.01 (0.95–1.08)
Marital status
Currently married Ref Ref Ref
Not married 1.20 (1.12–1.28) * 1.22 (1.14–1.31) * 1.19 (1.11–1.28) *
Race/ethnicity
Non-Hispanic White Ref Ref Ref
Hispanic 1.07 (0.97–1.18) 1.06 (0.96–1.17) 1.10 (1.00–1.22)
Non-Hispanic Black 1.28 (1.19–1.38) * 1.25 (1.16–1.35) * 1.28 (1.19–1.38) *
Non-Hispanic Asian 0.66 (0.55–0.79) * 0.67 (0.56–0.80) * 0.71 (0.59–0.86) *
Non-Hispanic Other 1.28 (0.93–1.76) 1.23 (0.89–1.69) 1.08 (0.78–1.50)
Education
≥ Bachelor’s degree Ref Ref Ref
Some college 1.11 (1.02–1.20) * 1.09 (1.00–1.18) * 1.05 (0.97–1.14)
≤ High school 1.14 (1.06–1.23) * 1.15 (1.06–1.24) * 1.10 (1.02–1.19) *
Poverty income ratio (PIR)†
PIR ≥ 2.00 Ref Ref Ref
PIR 1-1.99 1.28 (1.18–1.39) * 1.26 (1.16–1.37) * 1.19 (1.10–1.30) *
PIR < 1 1.70 (1.55–1.86) * 1.59 (1.44–1.75) * 1.44 (1.31–1.59) *
Employment status
Employed Ref Ref Ref
Unemployed 1.35 (1.26–1.45) * 1.52 (1.41–1.64) * 1.33 (1.23–1.44) *
Health insurance status
Insured Ref Ref Ref
Not insured 1.02 (0.87–1.18) 0.92 (0.79–1.07) 1.00 (0.85–1.16)
Have a usual place for medical care
No Ref Ref Ref
Yes 1.54 (1.26–1.89) * 1.57 (1.28–1.93) * 1.46 (1.18–1.80) *
Region
Northeast Ref Ref Ref
Midwest 1.08 (0.98–1.19) 1.08 (0.97–1.19) 1.06 (0.96–1.17)
South 0.95 (0.87–1.03) 0.94 (0.86–1.03) 0.91 (0.83–1.00)
West 0.95 (0.85–1.05) 0.95 (0.85–1.05) 0.94 (0.84–1.05)
Number of chronic conditions
2 - - Ref
3–4 - - 1.76 (1.65–1.88) *
≥5 - - 3.56 (3.23–3.92) *
a. Model 1: Adjusted for other social determinants of health b. Model 2: Adjusted for other social determinants of health, age, and sex. c. Model 3: Adjusted for other
social determinants of health, age, sex, and number of chronic conditions *Denotes statistical significance (P <.05) ED, emergency department; AOR, adjusted odds
ratio; CI, confidence interval; PIR < 1 = below poverty level; †PIR 1-1.99 = between 100-199% above poverty level; PIR ≥ 2 = 200% or above poverty level. Results are
weighted

including substance use. If this study included mental SDoH and ED visits, producing a confounding effect.
health problems as multimorbidity criteria, the preva- Thus, it may be beneficial to include mental health prob-
lence of multimorbidity would increase, which may influ- lems in regression models or multimorbid criteria for
ence the results. Moreover, since people with mental future studies to test whether they influence the associa-
health problems are more likely to have multimorbid- tion between SDoH and ED visits. In addition to mental
ity [44, 45], mental health may be associated with both health problems, some other possible chronic diseases
Lim et al. BMC Public Health (2024) 24:1153 Page 8 of 10

that are common in middle and older age groups should priority globally. There is increasing attention toward
also be comprehensively considered in further stud- studies focusing on etiology, epidemiology, and risk fac-
ies. Second, the study outcome (ED visits) and inclu- tors [1], yet there is still limited evidence to support
sion criteria (multimorbidity) were self-reported, which effective healthcare interventions [4]. As there is a need
may yield recall bias and information bias. A more sys- for increased awareness of multimorbidity, innovation,
tematic way to collect clinical data, such as data extrac- and optimization of the use of existing resources, under-
tion of health care utilization and disease diagnosis standing existing disparities of emergent care needs and
codes from electronic health records, may reduce the vulnerable groups can help determine which factors or
risk of bias in further studies. Third, the cross-sectional combinations of factors are most important to target.
approach in this study could not test a causal relationship The findings of this study underscore the importance of
between SDoH and ED visits. Thus, longitudinal studies not only addressing early-life disparities contributing to
are needed to examine whether SDoH affects ED visits developing multimorbidity but also the SDoH that influ-
to rule out reverse causality. In addition, this study did ences health status and emergent care needs. Particularly,
not cover all domains of the SDoH definition (e.g., neigh- increased health screening and assessment in primary
borhood/built environment and social/community) and care settings is needed for racial/ethnic minority popula-
adjusted for other SDoH to examine each SDoH effect on tions who have the disadvantage of emergent care access.
ED visits. Addressing all SDoH domains inclusively and Moreover, unemployed status and worsened financial
considering the intersectionality of SDoH would be ben- burden, which may hinder treatment adherence, should
eficial in examining the additive effects of SDoH on ED be addressed in the context of the treatment continuum
visits. among multimorbid people to prevent unplanned wors-
Lastly, the number of chronic conditions was adjusted ening symptoms and hospitalization. Lastly, the self-care
in the final regression model to account for the potential burden and need for social support in older multimorbid
confounding effect of the severity of overall chronic con- groups need to be paid more attention to mitigate the
ditions on the association between SDoH and ED visits. SDoH effect on emergent healthcare access.
Usually, the Charlson Comorbidity Index (CCI) or the
number of diseases is used as a proxy to adjust for the Conclusions
severity of overall chronic conditions [46, 47]. However, In conclusion, this study demonstrated that SDoH are
both CCI and the number of comorbidities may not be associated with increased ED visits among older adults
perfectly fitted with this study as a covariate since CCI living with multimorbidity. Systematic multidisciplinary
was developed as a predictor of 1-year mortality and bur- team approaches are needed to address social dispari-
den of disease [48], and the number of diseases cannot ties affecting multimorbidity prevalence, health-seeking
account for how comorbidities interact [46], although it is behaviors, and emergent healthcare access. Therefore,
assumed that increasing the number of diseases may lead researchers, healthcare practitioners, and policymakers
to increased overall severity. Unfortunately, the NHIS should pay attention to addressing the social disparities
dataset in this study did not cover all the diagnoses to by improving the management of chronic health condi-
calculate CCI, so this study included the number of dis- tions and promoting health equity.
eases in Model 3 as a proxy of the severity of the overall
Abbreviations
condition. However, it should be interpreted cautiously AOR Adjusted Odds Ratio
regarding the confounding effect of the severity of condi- CCI Charlson Comorbidity Index
tions in case it does not reflect the severity of the health CI Confidence Interval
ED Emergency Department
condition very well. A few studies have tried to develop NCHS National Center for Health Statistics
proper tools to measure the severity of multimorbid con- NHIS National Health Interview Survey
ditions, such as the multimorbidity interaction severity PIR Poverty Income Ratio
SD Standard Deviation
index [46]. However, this preliminary tool still needs to SDoH Social Determinants of Health
be verified for its reliability and validity in multiple pop- US United States
ulations [46]. Thus, a proper measure for the severity of
Acknowledgements
multimorbid conditions is necessary to be developed to Not applicable.
examine the association between SDoH and ED visits in
older multimorbid populations more precisely. Author contributions
AL analyzed and interpreted the data, prepared a first draft of the manuscript,
and revised the manuscript. CB, XL, and OO analyzed and interpreted the data
Implications and revised the manuscript. CH and PD supervised the process and revised
Multimorbidity is increasing, and individuals with mul- the manuscript. BK conceptualized and designed the study, interpreted the
data, and revised the manuscript. All authors read and approved the final
timorbidity are high utilizers of health care. Preven- manuscript.
tion and management of multimorbidity is now a key
Lim et al. BMC Public Health (2024) 24:1153 Page 9 of 10

Funding 10. Palladino R, Tayu Lee J, Ashworth M, Triassi M, Millett C. Associations between
Dr. Koirala is supported by the NINR P30 NR18093, the Building multimorbidity, healthcare utilisation and health status: evidence from 16
Interdisciplinary Research Careers in Women’s Health (BIRCWH, K12HD085845; European countries. Age Ageing. 2016;45(3):431–5.
PI: Daniel Ernest Ford) program, and the Johns Hopkins School of Nursing 11. Heins M, Korevaar J, Schellevis F, Rijken M. Identifying multimorbid patients
Discovery and Innovation Award. The content is solely the responsibility of the with high care needs - a study based on electronic medical record data. Eur J
authors and does not necessarily represent the official views of the supporting Gen Pract. 2020;26(1):189–95.
agencies. 12. de Souto Barreto P, Lapeyre-Mestre M, Vellas B, Rolland Y. Multimorbidity type,
hospitalizations and emergency department visits among nursing home
Data availability residents: a preliminary study. J Nutr Health Aging. 2014;18(7):705–9.
The datasets generated and/or analyzed during the current study are available 13. Latham LP, Ackroyd-Stolarz S. Emergency department utilization by older
in the NHIS repository: https://www.cdc.gov/nchs/nhis/data-questionnaires- adults: a descriptive study. Can Geriatr J. 2014;17(4):118–25.
documentation.htm. 14. Schiøtz ML, Stockmarr A, Høst D, Glümer C, Frølich A. Social disparities in the
prevalence of multimorbidity– a register-based population study. BMC Public
Health. 2017;17(1):422.
Declarations 15. Caraballo C, Herrin J, Mahajan S, Massey D, Lu Y, Ndumele CD, Drye EE, Krum-
holz HM. Temporal trends in racial and ethnic disparities in multimorbidity
Ethics approval and consent to participate prevalence in the United States, 1999–2018. Am J Med. 2022;135(9):1083–92.
This study was exempt from institutional review board review because it used e1014.
de-identified publicly available data published by the NCHS. 16. Northwood M, Ploeg J, Markle-Reid M, Sherifali D. Integrative review of the
social determinants of health in older adults with multimorbidity. J Adv Nurs.
Consent for publication 2018;74(1):45–60.
Not applicable. 17. von dem Knesebeck O, Bickel H, Fuchs A, Gensichen J, Höfels S, Riedel-Heller
SG, König H-H, Mergenthal K, Schön G, Wegscheider K. Social inequalities in
Competing interests patient-reported outcomes among older multimorbid patients–results of the
The authors declare that they have no competing interests. MultiCare cohort study. Int J Equity Health. 2015;14:1–11.
18. Healthy People. 2030 [https://health.gov/healthypeople/
Disclosure objectives-and-data/social-determinants-health].
The abstract has been submitted to the 2023 Gerontological Society of 19. Sangal RB, Su H, Khidir H, Parwani V, Liebhardt B, Pinker EJ, Meng L, Venkatesh
America Annual Meeting. AK, Ulrich A. Sociodemographic Disparities in Queue Jumping for Emergency
Department Care. JAMA Netw Open. 2023;6(7):e2326338–2326338.
Author details 20. National Center for Health Statistics. NHIS - Data, Questionnaires and Related
1
Johns Hopkins University School of Nursing, 525 N. Wolfe St, Baltimore, Documentation. 2020.
MD, USA 21. Parsons V, Moriarity C, Jonas K, Moore T, Davis K, Tompkins L. Design and
2
University of New York Rory Meyers College of Nursing, 433 1st Ave, NY, Estimation for the National Health Interview Survey, 2006–2015. Vital Health
USA Stat Ser 2 Data Evaluation Methods Res 2014(165):1–53.
3
University of Wollongong, Northfields Ave, Wollongong, NSW, Australia 22. Hartnett KP. Impact of the COVID-19 pandemic on emergency department
visits—United States, January 1, 2019–May 30, 2020. MMWR Morbidity and
Received: 16 November 2023 / Accepted: 15 April 2024 mortality weekly report 2020, 69.
23. Caraballo C, Herrin J, Mahajan S, Massey D, Lu Y, Ndumele CD, Drye EE,
Krumholz HM. Temporal trends in racial and ethnic disparities in Mul-
timorbidity Prevalence in the United States, 1999–2018. Am J Med.
2022;135(9):1083–e10921014.
References 24. Drye EE, Altaf FK, Lipska KJ, Spatz ES, Montague JA, Bao H, Parzynski CS, Ross
1. Navickas R, Petric VK, Feigl AB, Seychell M. Multimorbidity: what do we know? JS, Bernheim SM, Krumholz HM. Defining multiple chronic conditions for
What should we do? J Comorb. 2016;6(1):4–11. quality measurement. Med Care. 2018;56(2):193–201.
2. Mortality G. Causes of death collaborators global, regional, and national life 25. Huang Y-q, Gou R, Diao Y-s, Yin Q-h, Fan W-x, Liang Y-p, Chen Y, Wu M,
expectancy, all-cause and cause-specific mortality for 249 causes of death, Zang L, Li L. Charlson comorbidity index helps predict the risk of mortal-
1980–2015: a systematic analysis for the global burden of Disease Study ity for patients with type 2 diabetic nephropathy. J Zhejiang Univ Sci B.
2015. Lancet. 2016;388:1459–544. 2014;15:58–66.
3. Chowdhury SR, Chandra Das D, Sunna TC, Beyene J, Hossain A. Global and 26. Lim A, Choi M, Jang Y, Lee H. Preoperative frailty based on laboratory data
regional prevalence of multimorbidity in the adult population in community and postoperative health outcomes in patients undergoing coronary artery
settings: a systematic review and meta-analysis. eClinicalMedicine 2023, 57. bypass graft surgery. Heart Lung. 2022;56:1–7.
4. Skou ST, Mair FS, Fortin M, Guthrie B, Nunes BP, Miranda JJ, Boyd CM, Pati 27. Ingram DD, Malec DJ, Makuc DM, Kruszon-Moran D, Gindi RM, Albert M,
S, Mtenga S, Smith SM. Multimorbidity. Nat Reviews Disease Primers. Beresovsky V, Hamilton BE, Holmes J, Schiller JS. National center for health
2022;8(1):48. statistics guidelines for analysis of trends. 2018.
5. Lee JE, Haynes E, DeSanto-Madeya S, Kim YM. Social Determinants of Health 28. Kang H. The prevention and handling of the missing data. Korean J Anesthe-
and Multimorbidity among adults 50 years and older in the United States. siol. 2013;64(5):402–6.
Nurs Res. 2024;73(2):126–37. 29. Pandey KR, Yang F, Cagney KA, Smieliauskas F, Meltzer DO, Ruhnke GW. The
6. Ansah JP, Chiu CT. Projecting the chronic disease burden among the adult impact of marital status on health care utilization among Medicare beneficia-
population in the United States using a multi-state population model. Front ries. Med (Baltim). 2019;98(12):e14871.
Public Health. 2022;10:1082183. 30. Kruse JA, Low LK, Seng JS. Validation of alternative indicators of social sup-
7. Koirala B, Peeler A, Dennison Himmelfarb C, Davidson PM. Living with mul- port in perinatal outcomes research using quality of the partner relationship.
tiple chronic conditions: how we achieve holistic care and optimize health J Adv Nurs. 2013;69(7):1562–73.
outcomes. J Adv Nurs. 2023;79(2):e7–9. 31. Vlachantoni A. Unmet need for social care among older people. Ageing Soc.
8. Thanakiattiwibun C, Siriussawakul A, Virotjarumart T, Maneeon S, Tantai N, 2019;39(4):657–84.
Srinonprasert V, Chaiwat O, Sriswasdi P. Multimorbidity, healthcare utilization, 32. Barrenetxea J, Tan KB, Tong R, Chua K, Feng Q, Koh W-P, Chen C. Emergency
and quality of life for older patients undergoing surgery: a prospective study. hospital admissions among older adults living alone in the community. BMC
Med (Baltim). 2023;102(13):e33389. Health Serv Res. 2021;21(1):1192.
9. Koroukian SM, Warner DF, Owusu C, Given CW. Multimorbidity redefined: 33. Lloyd T, Crellin E, Brine RJ, Shen JY, Wolters AT. Association between house-
prospective health outcomes and the cumulative effect of co-occurring hold context and emergency hospital use in older people: a retrospective
conditions. Prev Chronic Dis 2015, 12. cohort study on indicators for people living alone or living with somebody
Lim et al. BMC Public Health (2024) 24:1153 Page 10 of 10

with frailty, developed from routine healthcare data in England. BMJ Open. 42. Paremoer L, Nandi S, Serag H, Baum F. Covid-19 pandemic and the social
2022;12(5):e059371. determinants of health. BMJ. 2021;372:n129.
34. Tora Grauers W, Volkert S, Dagny Ros N, Dorte J, Ann Dorrit G, Susanne 43. Molina M, Evans J, Montoy JC, Cawley C, Graham-Squire D, Perez K, Raven
R, Rasmus K-R. Symptom burden in multimorbidity: a population-based M, Kanzaria HK. Analysis of Emergency Department Encounters among
combined questionnaire and registry study from Denmark. BMJ Open. High Users of Health Care and Social Service systems before and during the
2021;11(4):e041877. COVID-19 pandemic. JAMA Netw Open. 2022;5(10):e2239076–2239076.
35. Chen YH, Karimi M, Rutten-van Mölken M. The disease burden of 44. Barnett K, Mercer SW, Norbury M, Watt G, Wyke S, Guthrie B. Epidemiology
multimorbidity and its interaction with educational level. PLoS ONE. of multimorbidity and implications for health care, research, and medical
2020;15(12):e0243275. education: a cross-sectional study. Lancet. 2012;380(9836):37–43.
36. Bos-Touwen I, Schuurmans M, Monninkhof EM, Korpershoek Y, Spruit- 45. Melis R, Marengoni A, Angleman S, Fratiglioni L. Incidence and predictors of
Bentvelzen L, Ertugrul-van der Graaf I, de Wit N, Trappenburg J. Patient and multimorbidity in the elderly: a population-based longitudinal study. PLoS
Disease Characteristics Associated with activation for self-management in ONE. 2014;9(7):e103120.
patients with diabetes, Chronic Obstructive Pulmonary Disease, Chronic 46. Gassmann D, Cheetham M, Siebenhuener K, Holzer BM, Meindl-Fridez C,
Heart failure and chronic renal disease: a cross-sectional survey study. PLoS Hildenbrand FF, Virgini V, Martin M, Battegay E. The multimorbidity interaction
ONE. 2015;10(5):e0126400. severity index (MISI): a proof of concept study. Medicine 2017, 96(8).
37. Schmaderer MS, Zimmerman L, Hertzog M, Pozehl B, Paulman A. Correlates 47. Charlson ME, Carrozzino D, Guidi J, Patierno C. Charlson Comorbidity
of patient activation and Acute Care utilization among Multimorbid patients. Index: a critical review of Clinimetric Properties. Psychother Psychosom.
West J Nurs Res. 2016;38(10):1335–53. 2022;91(1):8–35.
38. Laba T-L, Cheng L, Kolhatkar A, Law MR. Cost-related nonadherence to medi- 48. Roffman C, Buchanan J, Allison G. Charlson comorbidities index. J Physio-
cines in people with multiple chronic conditions. Res Social Administrative therapy 2016, 62(3).
Pharm. 2020;16(3):415–21.
39. Larkin J, Foley L, Smith SM, Harrington P, Clyne B. The experience of financial
burden for people with multimorbidity: a systematic review of qualitative Publisher’s Note
research. Health Expect. 2021;24(2):282–95. Springer Nature remains neutral with regard to jurisdictional claims in
40. James L, Louise F, Shane T, Tony H, Barbara C, Patricia H, Susan MS. How do published maps and institutional affiliations.
people with Multimorbidity Prioritise Healthcare when faced with Financial
constraints? A choice experiment. medRxiv 2023:2023.2004.2010.23288383.
41. Selby S, Wang D, Murray E, Lang E. Emergency departments as the Health
Safety Nets of Society: a descriptive and Multicenter Analysis of Social Worker
support in the emergency room. Cureus. 2018;10(9):e3247.