Page 1 of 21
Archives of Business Review – Vol. 8, No.7
Publication Date: July 25, 2020
DOI: 10.14738/abr.87.8569.
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
1
Associations of Multiple Chronic Health Conditions with Health
Behavior
Chong Hwan Son
Ph.D. Economics, Department of History, Philosophy,
Social Sciences, Hudson Valley Community College,
Troy, New York, U.S.A.
ABSTRACT
This study aimed to scrutinize the association of the number of chronic
health conditions with health behavior. The health behavior was
measured by meeting the 2008 Physical Activity Guidelines (PAGs) for
Americans for five physical activity levels for adults aged 18 years or
older in the United States using data from the 2017 Behavioral Risk
Factor Surveillance System (BRFSS). The empirical results of a
multivariate logistic regression analysis revealed that respondents
living with chronic health conditions were more likely to participate in
aerobic physical activities, but not meeting the PAGs. In the insufficient
physical activity subgroup, all of the predicted odds ratios were greater
than one and increased as the number of chronic health conditions
increased. It implied that the increase in the number of chronic
conditions was positively associated with participating in insufficient
physical activity. Respondents who reported having less than three
chronic health conditions were more likely to meet the aerobic physical
activity guidelines compared with respondents living with three or
more chronic health conditions. Importantly, respondents who
reported having 4 or more chronic health conditions had a higher
likelihood of meeting the recommendations for muscle-strengthening
activity. However, chronic health conditions would significantly
discourage respondents from participating in both aerobic and muscle- strengthening physical activities. In conclusion, this study found that
chronic health conditions played an important role in determining
regular participation in the level of physical activity for individuals
living with chronic health conditions.
Keywords: Physical Activity, Chronic Health Conditions, BRFSS, the 2008
Physical Activity Guidelines.
INTRODUCTION
It is well known that participating in regular physical activity is one of the most important things
individuals can maintain or improve their fitness and physical function, and regular physical
activity can produce a large number of immediate and long-term health benefits, both physically
and mentally (Farren, et al., 2017; Kujala, 2018). Participating in regular physical activity helps
physically to strengthen bones and muscles, control weight, maintain healthier body composition,
increase the resistance to diseases, etc. The mental health benefits of regular physical activity are
improved mental health and enhanced health-related quality of life by increasing enjoyment, self-
Page 2 of 21
URL: http://dx.doi.org/10.14738/abr.86.8569 2
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
esteem, energy, and engagement in social activities (Farren, et al., 2017). These health benefits of
regular physical activity are important pillars for promoting healthy lifestyles in public health
perspectives. For improving public health through participation in regular physical activity,
physical activity recommendations have been promoted by health organizations.
Current physical activity recommendations by the 2008 Physical Activity Guidelines (PAGs) for
Americans (U.S. Department of Health and Human Services, 2008) and the World Health
Organization (WHO) were subject to the evidence relating physical fitness, physical activity, and
energy expenditure to health outcomes. The recommendations provided key guidelines that adults
should do at least 150 minutes of moderate-intensity aerobic physical activity per week, or 75
minutes of vigorous-intensity aerobic physical activity per week, or an equivalent combination of
moderate- and vigorous-intensity aerobic physical activity. The aerobic physical activity should be
performed in episodes of at least 10 minutes duration. Besides, the 2008 PAGs provided additional
guidelines that adults should also engage in muscle-strengthening activities that were of moderate- or high-intensity activities and worked for all major muscle groups at least 2 days a week. Research
has indicated that muscle-strengthening activities also provide additional health benefits, such as
cardiovascular and metabolic benefits independent of aerobic physical activity. Therefore,
regularly participating in both aerobic physical activity and muscle-strengthening physical activity
was encouraged (Carlson, et al., 2015; Farren, et al., 2017).
As an increase in the average life expectancy over the last decades in the United States – the life
expectancy at birth for males (females) had increased from 71.8 (78.8) years in 1990 to 76.1 (81.1)
years in 2016 (National Centers for Health Statistics, 2017), the proportion of the elderly
population has steadily grown in the United States (Horner and Cullen, 2016; Son, 2020). As a
result, the prevalence and burden of individuals living with multiple chronic health conditions have
been anticipated to be an upward trend. The study (Newman, et al., 2019) defined chronic health
conditions as “conditions that last a year or more and require ongoing medical attention and/or
limit activities of daily living.” Chronic health conditions are a long term illness that can reduce an
individual’s quality of life and/or his/her life expectancy. However, studies (Booth, et al., 2012;
Miyawaki, et al., 2017) advocated that the chronic health conditions can be prevented or controlled
through regular participation in physical activity and that being active was an important
component of a health-promoting lifestyle for adults with chronic diseases. For instance, studies
(Rydwik, et al., 2012; Miyawaki, et al., 2017; Hallgren, et al., 2019) found that the likelihood of
having an incidence of major depressive disorder fell as the level of physical activity rose and that
physical activity was associated with a reduction in the risk of chronic health diseases and
conditions. Their findings induced that engaging in any level of physical activity for individuals
without chronic health conditions was beneficial to them to prevent chronic health problems. At
the same time, individuals living with chronic health conditions were expected to participate more
in regular physical activity to treat or manage their chronic health conditions. For example, the
study (Zechner and Gill, 2016) reported that the number of chronic health conditions was
positively associated with the amount of self-reported exercise using 31 common chronic health
conditions.
In the end, individuals living with chronic health problems may go in search of controlling or
minimizing their affliction through a positive form of health behavior (Cockerham, et al., 2014),
such as participating in physical activity. If the individuals desire to participate in regular physical
Page 3 of 21
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
3
activity to improve the transition with chronic health conditions, they could pursue a feasible
intensity level of physical activity based on the degree of their functional limitations: some may be
able to do the more intensive physical activity than others may do. The studies (Bowles, 2012;
Pettee Gabriel, et al., 2012) defined the physical activity as “the behavior that involves human
movement, resulting in physiological attributes including increased energy expenditure and
improved physical fitness.” An important part of the definition was the addressing of physical
activity as a behavior, which brought up a deliberate character of physical activity. It was the
central idea to develop a hypothesis for the current study to investigate the association of the
number of chronic health conditions with health behavior, such as meeting the 2008 PAGs. The
number of chronic health conditions would be served as a proxy for the degree of functional
limitations in the current study. This study conducted the multivariate logistic regression for five
physical activity levels for adults aged 18 years or older in the United States using the 2017
Behavioral Risk Factor Surveillance System (BRFSS) which computed and provided an assessment
of physical activity levels. Respondents from the BRFSS were classified into five mutually exclusive
physical activity levels, such as physical inactivity, insufficient physical activity, meeting the
aerobic PAGs, meeting the muscle-strengthening PAGs, and meeting both the PAGs (Desmond, et
al., 2015; Bennie, et al., 2019a). A detailed description of the BRFSS data and the definition of the
physical activity levels are presented in section 3.
ANALYTICAL FRAMEWORK
The level of physical activity based on the 2008 PAGs was served as a dichotomous dependent
variable. Each respondent was coded as either ‘1’ if ‘meeting the recommended level of physical
activity’ or ‘0’ if ‘not meeting’. Estimates of the probability of meeting the recommended level of
physical activity could be developed as a behavioral equation, such as a general linear probability
equation:
yi =β0 + xikβk + ui (1)
where the subscript i denoted an observation, and yi was the probability of meeting the
recommended level of physical activity at the ith observation. xik was an i x k matrix of explanatory
variables where k was the number of explanatory variables. β0 was a constant term, and βk was a
coefficient matrix of the equation with a dimension of k x1. ui was the error term at each
observation i with the expected value of ui, E(ui) = 0. Thus, the conditional expected value of the
dependent variable was expressed as E(yi|xik) = β0 + xikβk. The predicted values of the dependent
variables should have a certain probability of occurrence that should be between zero and one.
However, the linear probability model had an unboundedness problem that a predicted value could
fall outside of the meaningful range of zero to one. As a result, the predicted probability could be
quite different from the observed value for each observation. The binomial logit model could avoid
the unboundedness problem and have been widely used in research dealing with a binary
dependent variable (Maddala, 1992; Kleinbaum and Klein, 2002; Son, 2014).
In the logistic regression, an unknown probability (y) could be estimated for any given linear
combination of independent variables. The dependent variable in the logistic regression follows
the Bernoulli distribution (Marshall and Olkin, 1985) which is just a case of the binomial
distribution with just one trial: success is ‘1’ and failure is ‘0’. If the probability of success is y, the
probability of failure is (1 – y). Therefore, it is required to link together the independent variables
to essentially the Bernoulli distribution. The link is called a logit. In logistic regression, the
Page 4 of 21
URL: http://dx.doi.org/10.14738/abr.86.8569 4
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
probability is unknown like in the binomial (Bernoulli) distribution problems. As mentioned above,
the goal of using the logistic regression is to estimate the unknown probability (y) for a given linear
combination of independent variables. The estimate of y is y-hat, !". To tie together the linear
combination of independent variables and in an essence of the Bernoulli distribution, it is needed
of a function that linked them together or mapped the linear combination of variables that could
result in any value into the Bernoulli probability distribution with a domain from zero to one. The
natural log of the odds ratio (the logit) is the link function:
ln (odds) = ln# $
%&$
' is expressed as ‘logit (y) or ln (y) – ln (1 – y)’, such as
logit (y) = ln# $
%&$
' (2)
In equation (2), if y = 0, logit (y) would be –∞; if y = 1, logit (y) would be +∞; and if y = 0.5, logit (y)
would be zero. In the logit linked function graph, the probability (y) ranges from 0 to 1 alone the x- axis, but the probability (y) should be on the y-axis. It could be achieved by taking an inverse of the
logit function. To rewrite the logit function:
logit (y) = ln# $
%&$
', and then set ln# $
%&$
' = α
where,
α is some number,
y ranges between 0 and 1.
The inverse of the logit function becomes:
logit-1 (y) = # $
%&$
' or # $
%&$
'= eα
Solve for y,
y = # ()
%*()' (3)
This study assumes that α is a linear combination of independent variables and their coefficients
as α = β0 + xikβk in equation (3). Eventually, the probability becomes the Sigmoid function curve
(Ezeafulukwel, et al., 2018; Ezeafulukwel, et al., 2020), known as the ‘S’ curve in a range of 0 to 1
inclusive alone the y-axis. Thus, equation (3) becomes the logistic function with substituting β0 +
xikβk for α, and then the probability would be a function of independent variables (xs):
P(x) = + (,-./01,1
%*(,-./01,1
2 (4)
In equation (4), when β0 + xikβk approaches to – ∞, P(x) becomes 0:
P(x) = # (3∞
%*(3∞'; # (3∞
%*(3∞' = # %
%*(∞'; and then # %
%*(∞' ≅ 0
In equation (4), when β0 + xikβk approaches to +∞, P(x) becomes 1:
Page 5 of 21
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
5
P(x) = # (.∞
%*(.∞' = 1, and then # (.∞
%*(.∞' ≅ 1
Ultimately, the range of the logistic function in equation (4) would be between 0 and 1, regardless
of the value of α which is a linear combination of independent variables with constant coefficients.
This is the primary reason the logistic model is wildly used in research dealing with the binary
dependent variable.
To solve for logit (y) from equation (2), substitute equation (3) for y in the odds in equation (2),
and then the algebraic procedure is the following:
+ !
1 − !
2 =
:;
1 + :;
1 − :;
1 + :;
=
:;
1 + :;
1 + :;
1 + :; − :;
1 + :;
=
:;
1 + :;
1
1 + :;
= :;
where α = β0 + xikβk.
Thus, the logit(y) in equation (2) becomes:
logit (y) = ln# $
%&$
'; ln# $
%&$
' = =>( :;); =>( :;) = A; and then =>( :;) = BC + DEFBF
Thus,
logit (y) = BC + DEFBF (5)
Equation (5) indicates that the natural logarithm of the odds ratio becomes a linear function of the
independent variables. From equation (2) to equation (5), the above algebraic processes result in
that the inverse of the logit function linked to the estimated regression equation. Thus, the
estimated probability from equation (4) becomes:
!"= + (,G-./01,G1
%*(,G-./01,G1
2 (6)
This estimated probability equation links to the logistic regression. From equation (6), the odds
ratio could be derived. Suppose there is only an independent variable in the model. If the value of
an independent variable, X, changes by one, let y0 be the probability of success when X = 0; y1 be
the probability of success when X =1. If the value of X changes from 0 to 1 or any increment by one,
the odds of success for X = 0 and X = 1 are the following:
the odds of success for X = 0,# $-
%&$-
' = :HG-*IHGJ , and then:HG-*IHGJ = :HG- and the odds of success for
X = 1,# $J
%&$J
' = :HG-*IHGJ, and then :HG-*IHGJ = :HG-*HGJ
Thus, the odds ratio between X = 0 and X = 1 could be derived when the value of an independent
variable ( e.g. X) changes by one (from X = 0 to X = 1).
# KJ
J3KJ
'
# K- J3K-
' = (,G-.,GJ
(,G- ;
(,G-.,GJ
(,G- = (,G-(,GJ
(,G- ;
and then (,G-(,GJ
(,G- = :HGJ
Page 6 of 21
URL: http://dx.doi.org/10.14738/abr.86.8569 6
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
Finally, the odds ratio in the logistic regression is expressed as an exponential of the estimated
coefficient (BL
F) of an independent variable shown above the algebraic procedures. In this study,
odds ratios were obtained from a multivariate logistic regression analysis.
DATA
Data were drawn from the 2017 BRFSS which was a large-scale national health survey. The BRFSS
was a collaborative project of the CDC with all 50 states and the District of Columbia in the United
States. The BRFSS was designed to conduct monthly ongoing health-related telephone surveys to
collect uniform state-specific data on health-related risk behaviors, chronic diseases, chronic
health conditions, access to health care, and use of preventive
Table 1: Definition, Mean, and Standard Deviations of Variablesa
Variables Definition Mean and SD
Physical Activitiesb
Inactive
Respondent answers “no’ to the leisure-time physical activity question or
respondent reports performing an aerobic physical activity bout for less than 10
minutes in duration or respondent reports performing a non-aerobic activity.
22.00 (41.43)
Insufficient Respondent reports greater than 10 minutes and less than or equal to 149 minutes
of aerobic activity per week. 15.70 (36.38)
Meet aerobic
Respondent reports at least 150 minutes per week of moderate-intensity activity,
or at least 75 minutes per week of vigorous-intensity activity, or an equivalent
combination of moderate-intensity and vigorous-intensity (multiplied by 2)
totaling at least 150 minutes per week.
30.82 (46.17)
Meet strengthen Respondent reports participation in muscle-strengthening activities at least 2
times per week. 10.20 (30.26)
Meet both
Respondent reports participation in muscle-strengthening activities at least 2
times per week and at least 150 minutes per week of moderate-intensity activity,
or at least 75 minutes per week of vigorous-intensity activity, or an equivalent
combination of moderate-intensity and vigorous-intensity (multiplied by 2)
totaling at least 150 minutes per week.
21.28 (40.93)
Marital Status
Married Respondent is married or a member of an unmarried couple. 58.65 (49.25)
Divorced Respondent was divorced or separated. 13.86 (34.55)
Widowed Respondent was widowed. 6.58 (24.80)
Never married Respondent is never married. 20.91 (40.67)
Income Level
Income 20 Respondent’s annual household income from all sources is less than $20,000. 15.91 (36.58)
Income 2035 Respondent’s annual household income from all sources is $20,000 to less than
$35,000. 18.33 (38.70)
Income 3550 Respondent’s annual household income from all sources is $35,000 to less than
$50,000. 13.15 (33.80)
Income 5075 Respondent’s annual household income from all sources is $50,000 to less than
$75,000. 15.20 (35.91)
Income 75+ Respondent’s annual household income from all sources is $75,000 or more. 37.40 (48.39)
Race
White, non- Hispanic Respondent is white only, non-Hispanic. 66.70 (47.13)
Black, non- Hispanic Respondent is black only, non-Hispanic. 11.50 (31.91)
Page 7 of 21
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
7
Hispanic Respondent is Hispanic. 14.02 (34.72)
Other, non- Hispanic Respondent is other race, non-white, non-black, and non-Hispanic. 7.78 (26.78)
Age
Age 18-24 Respondent aged 18-24 years. 9.14 (28.82)
Age 25-34 Respondent aged 25-34 years. 16.10 (36.75)
Age 35-49 Respondent aged 35-49 years. 25.30 (43.47)
Age 50-64 Respondent aged 50-64 years. 28.44 (45.11)
Age 65+ Respondent aged 65 years or older. 21.02 (40.75)
Education
Less high school
graduate Respondent never attended school or completed grades 1 through 11. 10.56 (30.73)
High school
graduate Respondent completed grade 12 or GED (high school graduate). 26.10 (43.92)
Some college
graduate Respondent completed college 1 year to 3 years. 32.15 (46.71)
College graduate
or more Respondent completed college 4 years or more (college graduate). 31.20 (46.33)
BMIc
Under weight BMI < 18.50 1.59 (12.51)
Normal weight 18.50 ≤ BMI < 25.00 30.77 (46.16)
Over weight 25.00 ≤ BMI < 30.00 36.06 (48.09)
Obese 30.00 ≤ BMI 31.58 (46.48)
Others
Male Respondent is male. 50.34 (50.00)
Employed Respondent is employed for wages or self-employed. 59.88 (49.01)
Current smoker Dichotomous variable that equals 1 if respondent now smokes cigarettes every
day or some days. 15.72 (36.40)
Current drinkerd
Dichotomous variable that equals 1 if respondent reported having had at least one
drink of any alcoholic beverage such as beer, wine, a malt beverage or liquor in the
past 30 days.
57.70 (49.40)
Chronic health scorese
Score 0 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 0. 30.01 (45.83)
Score 1 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 1. 24.43 (42.97)
Score 2 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 2. 17.73 (38.19)
Score 3 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 3. 11.91 (32.39)
Score 4 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 4. 7.45 (26.25)
Score 5 Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 5. 4.20 (20.06)
Score 6+ Dichotomous variable that equals 1 if the total number of CHCs respondents
reported equals 6+. 4.27 (20.22)
Notes
a. Standard deviations are in parentheses. Means (%) and standard deviations are weighted by the
BRFSS sample weight factor. All means are statistically significant using two tailed t-test with 95%
Page 8 of 21
URL: http://dx.doi.org/10.14738/abr.86.8569 8
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
confidence interval. b. Classify respondents as meeting the 2008 Physical Activity Guidelines for
Americans. c. Body Mass Index (BMI): weight (kg) divided by square of height (meter). d. One drink
is equivalent to a 12-ounce beer, a 5-ounce glass of wine, or a drink with one shot of liquor. e. The
total number of the chronic health conditions for each respondent from 13 chronic health
conditions: hypertension awareness module, cholesterol awareness, and 11 chronic conditions in
the chronic health condition module. CHC denotes chronic health condition.
services related to the leading causes of death and disability from the non-institutionalized adult
population aged 18 years or older residing in the United States. The BRFSS field operations were
managed by the state health department that followed the protocols provided by the states with
technical assistance provided by the CDC. Some states from the beginning stratified population
allowed them to collect region-wise data in their states. The total baseline observations for adults
aged 18 years or older from the 2017 BRFSS were 450, 016. The present study excluded
observations with ‘missing values’, ‘unanswered questions’, ‘don’t know/not sure’, ‘questions not
asked’, and ‘refusals’. 281,913 survey participants (62.65% of the total baseline observations)
answered all the questions on physical activities, socio-demographic variables, health behavior
variables, and chronic health conditions. Descriptions and summary statistics are shown in Table
1.
Physical Activity
The methodology and questionnaires employed to assess physical activity within the BRFSS have
been published elsewhere. The questionnaires have been shown both reliable and valid in
assessing physical activity behavior (Yore, et al., 2007; Loprinzi, et al., 2015). To collect data
regarding the prevalence of participation in physical activity, the BRFSS interviewers asked,
“During the past month, other than your regular job, did you participate in any physical activities
or exercises such as running, calisthenics, golf, gardening, or walking for exercise?” If the
participants answered ‘yes’, they were asked, “What type of physical activity or exercise did you
spend the most time doing during the past month?” From this question, the interviewers used a
coding list of 76 physical activities plus an “other” category to classify the activities. In that coding
protocol, examples of aerobic activities included walking, hiking, biking, swimming, and running,
while non-aerobic activities included gardening, painting, golf, and bowling (Bennie, et al., 2019a;
Bennie, et al., 2019b). For each activity, respondents were asked to respond to the following
questions to measure frequency and duration, respectively: “How many times per week or per
month did you take part in this activity during the past month?”; “And when you took part in this
activity, for how many minutes or hours did you usually keep at it?” After these questions, the
interviewers also asked respondents about the other type of physical activity: “What other types
of physical activity gave you the next most exercise during the past month?” For each activity,
respondents were asked to respond to the following questions to measure frequency and duration,
respectively: “How many times per week or per month did you take part in this activity during the
past month?”; “And when you took part in this activity, for how many minutes or hours did you
usually keep at it?” Physical activity or exercises to strengthen muscles were assessed by asking,
“During the past month, how many times per week or per month did you do physical activities or
exercises to strengthen your muscles?” Respondents were prompted: “Do not count aerobic
activities like walking, running, or bicycling. Count activities using your own body weight like yoga,
sit-ups or push-ups and those using weight machines, free weights, or elastic bands.”
Page 9 of 21
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
9
A metabolic equivalent value was assigned to each activity listed on the BRFSS. Aerobic activity
should be performed in bouts of 10 minutes or longer in its calculations of the total minutes of
moderate- or vigorous-intensity activity. Aerobic activity levels were calculated for each
respondent using the 2008 PAGs. Respondents who reported no leisure-time activity were
classified as physically inactive, and respondents who did not have any bouts of aerobic activity
longer than 10 minutes were classified as also physically inactive. Respondents who reported
bouts of aerobic activity but did not meet the guideline threshold were classified as being
insufficiently active. Respondents who reported more than 150 minutes of moderate-intensity
activities, or more than 75 minutes of vigorous activity, or an equivalent combination were
classified as being sufficiently active as meeting the recommendations for aerobic physical activity.
The BRFSS also assessed meeting the recommendations for muscle-strengthening activity.
Respondents who reported participating in muscle-strengthening activities at least two times per
week were classified as meeting the muscle-strengthening guidelines. As a consequence, five
mutually exclusive subgroups of the physical activity levels in this study were used: physically
inactive, insufficient physical activity, meeting the aerobic physical activity guidelines, meeting the
muscle-strengthening activity guidelines, and meeting both the guidelines (Carlson, et al., 2015;
Desmond, et al., 2015; Boyer, et al., 2018; Bennie, et al., 2019a; Bennie, et al., 2019b). Table 1 shows
the classification of physical activity levels.
Explanatory Variables
Explanatory variables were socio-demographic variables: marital status, income level,
race/ethnicity, age, education level, gender, and employed; health behavioral variables: Body Mass
Index (BMI), current smoker, and current drinker; and chronic health conditions. These socio- demographic and health behavior variables were widely used in the research in physical activity
and chronic health problems (Miyawaki, et al., 2017; Boyer, et al., 2018; Bennie, et al., 2019b; Li, et
al., 2019).
Socio-demographic variables
Respondents were asked about marital status: married, divorced, widowed, separated, never
married, and a member of an unmarried couple. The six categories of marital status were then
reclassified into four marital subgroups: married (married or a member of an unmarried couple),
divorced (divorced or separated), widowed, and never married. After regrouping, 58.65%, 13.86%,
6.58%, and 20.91% of survey participants were married, divorced, widowed, and never married,
respectively. The BRFSS asked participants for their annual household income from all sources.
The participants could report one of the eight categories of annual household income levels from
less than $10,000 to $75,000 or more. This study re-categorized them into five income levels.
15.91% of participants reported their annual household income of less than $20,000, and 18.33%,
13.15%, 15.20%, and 37.40% of participants reported $20,000 ≤ income < $35,000, $35,000≤
income < $50,000, $50,000 ≤ income < $75,000, and $75,000 and over, respectively. Participants
were also asked to give the races. The participants were classified into four racial categories: White,
but not Hispanic; Black, but not Hispanic; Hispanic; and other races. 66.70% of the respondents
were White, non-Hispanic, 11.50% of respondents were Black, non-Hispanic, 14.02% of
respondents were Hispanic, and 7.78% of respondents were other races. Age was categorized into
five mutually exclusive categories: 18-24, 25-34, 35-49, 50-64, and 65 and older. Education was
divided into four categories: 10.56% of respondents completed less than high school, and 26.10%,
Page 16 of 21
URL: http://dx.doi.org/10.14738/abr.86.8569 16
Son, C. H. (2020). Associations of Multiple Chronic Health Conditions with Health Behavior. Archives of Business Research, 8(7). 1-21.
Table 5: Odds Ratios by Physical Activity Levela
Inactive
(95% CI)
Insufficient
(95% CI)
Meet Aerobic
(95% CI)
Meet Strengthen
(95% CI)
Meet Both
(95% CI)
Marital Status
Married 1 1 1 1 1
Divorced 0.97 (0.95-1.00) 0.93 (0.90-0.96) 0.90 (0.88-0.92) 1.22 (1.17-1.27) 1.16 (1.13-1.20)
Widowed 1.05 (1.02-1.08) 0.85 (0.82-0.89) 0.97 (0.95-1.00) 1.13 (1.07-1.19) 1.01 (0.97-1.05)
Never married 0.93 (0.90-0.96) 0.95 (0.91-0.98) 0.93 (0.90-0.96) 1.15 (1.10-1.20) 1.15 (1.11-1.19)
Income Level
Income 20 1 1 1 1 1
Income 2035 0.94 (0.91-0.97) 0.96 (0.92-0.99) 1.09 (1.06-1.12) 0.98 (0.93-1.03) 1.08 (1.04-1.12)
Income 3550 0.84 (0.81-0.87) 0.91 (0.87-0.95) 1.15 (1.11-1.18) 0.98 (0.93-1.04) 1.23 (1.18-1.28)
Income 5075 0.75 (0.73-0.78) 0.90 (0.86-0.93) 1.17 (1.13-1.21) 0.99 (0.94-1.05) 1.35 (1.30-1.41)
Income 75+ 0.59 (0.57-0.61) 0.84 (0.81-0.88) 1.18 (1.14-1.22) 1.01 (0.96-1.06) 1.63 (1.57-1.69)
Race
White, non-Hispanic 1 1 1 1 1
Black, non-Hispanic 1.08 (1.05-1.12) 1.16 (1.12-1.21) 0.71 (0.69-0.74) 1.23 (1.18-1.29) 1.13 (1.09-1.18)
Hispanic 1.06 (1.02-1.10) 1.02 (0.98-1.07) 0.85 (0.82-0.88) 1.07 (1.02-1.13) 1.17 (1.13-1.22)
Other, non-Hispanic 0.90 (0.86-0.93) 1.01 (0.97-1.06) 0.93 (0.90-0.96) 1.13 (1.07-1.19) 1.14 (1.10-1.19)
Age
Age 18-24 1 1 1 1 1
Age 25-34 1.45 (1.35-1.55) 1.08 (1.02-1.16) 1.21 (1.15-1.28) 0.81 (0.76-0.86) 0.72 (0.68-0.76)
Age 35-49 1.78 (1.66-1.90) 1.07 (1.01-1.14) 1.44 (1.37-1.52) 0.54 (0.51-0.57) 0.65 (0.62-0.69)
Age 50-64 1.93 (1.81-2.06) 0.90 (0.85-0.96) 1.73 (1.65-1.83) 0.39 (0.37-0.42) 0.64 (0.61-0.67)
Age 65+ 2.16 (2.02-2.31) 0.64 (0.60-0.68) 1.99 (1.88-2.09) 0.32 (0.30-0.34) 0.67 (0.64-0.71)
Education
Less high school 1 1 1 1 1
High school 0.84 (0.81-0.88) 0.97 (0.93-1.03) 1.11 (1.07-1.16) 1.10 (1.03-1.18) 1.30 (1.22-1.38)
Some college 0.62 (0.60-0.65) 0.99 (0.94-1.05) 1.17 (1.12-1.22) 1.28 (1.19-1.38) 1.71 (1.61-1.82)
College graduate 0.42 (0.40-0.43) 1.03 (0.98-1.09) 1.18 (1.32-1.23) 1.40 (1.30-1.50) 2.15 (2.03-2.28)
BMIb
Normal weight 1 1 1 1 1
Under weight 1.44 (1.33-1.55) 1.07 (0.97-1.18) 0.91 (0.85-0.98) 0.90 (0.80-1.01) 0.84 (0.78-0.92)
Over weight 1.18 (1.15-1.21) 1.21 (1.18-1.25) 1.07 (1.04-1.09) 0.94 (0.91-0.97) 0.74 (0.73-0.76)
Obese 1.79 (1.75-1.83) 1.45 (1.41-1.50) 0.92 (0.90-0.94) 0.83 (0.80-0.86) 0.48 (0.47-0.50)
Others
Male 0.92 (0.90-0.94) 0.79 (0.78-0.81) 0.97 (0.95-0.99) 1.29 (1.25-1.32) 1.25 (1.23-1.28)
Employed 1.23 (1.21-1.26) 1.25 (1.22-1.28) 0.86 (0.85-0.88) 1.03 (1.00-1.07) 0.84 (0.82-0.86)
Current smoker 1.61 (1.57-1.65) 0.93 (0.90-0.95) 1.00 (0.97-1.02) 0.78 (0.75-0.81) 0.66 (0.64-0.68)
Current drinkerc 0.71 (0.69-0.72) 1.01 (0.99-1.03) 1.12 (1.10-1.14) 1.04 (1.01-1.07) 1.21 (1.18-1.23)
Chronic health scoresd
Score 0 1 1 1 1 1
Score 1 1.06 (1.03-1.10) 1.09 (1.05-1.12) 1.03 (1.01-1.06) 0.95 (0.92-0.99) 0.91 (0.88-0.93)
Score 2 1.21 (1.17-1.25) 1.15 (1.11-1.19) 1.00 (0.98-1.03) 0.91 (0.87-0.95) 0.83 (0.80-0.85)
Score 3 1.35 (1.31-1.40) 1.16 (1.11-1.20) 0.96 (0.93-0.99) 0.96 (0.92-1.01) 0.74 (0.72-0.77)
Score 4 1.54 (1.48-1.60) 1.23 (1.18-1.29) 0.86 (0.83-0.89) 1.02 (0.96-1.08) 0.66 (0.63-0.69)
Score 5 1.76 (1.69-1.84) 1.23 (1.17-1.30) 0.78 (0.75-0.82) 1.03 (0.96-1.10) 0.59 (0.56-0.62)
Score 6+ 2.07 (1.98-2.16) 1.18 (1.11-1.24) 0.63 (0.60-0.66) 1.20 (1.12-1.29) 0.55 (0.52-0.58)