Acculturation and the use of complementary and alternative medicine.

Abstract

The use of complementary and alternative medicine (CAM) has been growing substantially in the US in recent years. Such a growth in CAM use coincides with an ongoing increase in the proportion of the foreign-born population in the US. The main objective of this study is to examine the relation between acculturation and the use of CAM therapies among immigrants. Data from a CAM supplement to the 2002 National Health Interview Survey were analyzed to estimate the effects of acculturation on the likelihood of using different CAM therapies over the past 12 months prior to the survey. The results suggest that the level of acculturation-as measured by nativity/length of stay in the US and language of interview-is strongly associated with CAM use. As immigrants stay longer in the US or as their use of English becomes more proficient, the likelihood that they use CAM therapies increases as well, and it gradually approaches the level of CAM use by native-born Americans. Moreover, this relation between acculturation and CAM use generally persists even after the effects of socioeconomic status, health insurance coverage, self-reported health status, and other demographic variables have all been taken into consideration. The substantially lower rates of CAM use by recent immigrants and its possible causes warrant further research.

Full text

Page 1

Social Science & Medicine 66 (2008) 439–453
Acculturation and the use of complementary
and alternative medicine
Dejun Su?, Lifeng Li, Jose ´ A. Paga ´ n
The University of Texas-Pan American, Edinburg, TX, USA
Available online 17 October 2007
Abstract
The use of complementary and alternative medicine (CAM) has been growing substantially in the US in recent years.
Such a growth in CAM use coincides with an ongoing increase in the proportion of the foreign-born population in the US.
The main objective of this study is to examine the relation between acculturation and the use of CAM therapies among
immigrants. Data from a CAM supplement to the 2002 National Health Interview Survey were analyzed to estimate the
effects of acculturation on the likelihood of using different CAM therapies over the past 12 months prior to the survey. The
results suggest that the level of acculturation—as measured by nativity/length of stay in the US and language of
interview—is strongly associated with CAM use. As immigrants stay longer in the US or as their use of English becomes
more proficient, the likelihood that they use CAM therapies increases as well, and it gradually approaches the level of
CAM use by native-born Americans. Moreover, this relation between acculturation and CAM use generally persists even
after the effects of socioeconomic status, health insurance coverage, self-reported health status, and other demographic
variables have all been taken into consideration. The substantially lower rates of CAM use by recent immigrants and its
possible causes warrant further research.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: USA; Acculturation; Complementary and alternative medicine (CAM); Immigrants
Introduction
The use of complementary and alternative med-
icine (CAM) has increased substantially in the
United States over the last two decades (Barnes,
Powell-Griner, McFann, & Nahin, 2004). Survey
data indicate that the proportion of adults using
CAM practices, therapies, and products within the
last year increased from 34% in 1990 to 62% in
2002 (Pagan & Pauly, 2005). The annual number of
visits to CAM providers is now higher than the
number of visits to primary care physicians, and
yearly out-of-pocket expenditures on CAM services
exceed $27 billion (The Institute of Medicine, 2005).
The growth in CAM use is taking place at the same
time that the size of the foreign-born population is
increasing in the US. The relative size of the US
immigrant population has grown from 8% in 1990 to
11.7% in 2003 (Antecol & Bedard, 2006; US Census
Bureau, 2004). By 2045, as many as 13.3% of the US
population would have been born abroad (US Census
Bureau, 2000). Due to their lower socioeconomic
status and other barriers (e.g. language proficiency
ARTICLE IN PRESS
www.elsevier.com/locate/socscimed
0277-9536/$-see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.socscimed.2007.08.023
?Corresponding author. Tel.: +19563808798;
fax: +19563185303.
E-mail addresses: dsu@utpa.edu (D. Su), lli@utpa.edu
(L. Li), jpagan@husky.panam.edu (J.A. Paga ´ n).

Page 2

and health literacy) in access to medical services,
many immigrants resort to CAM for health care.
A recent study on the socio-cultural factors associated
with health care utilization among Latino immigrants
reported that they tend to use CAM first, and then
seek conventional medical help if CAM is not
effective (Garces, Scarinci, & Harrison, 2006).
The increasing share of the foreign-born popula-
tion in the US highlights the need for assessing the
role of acculturation in the use of CAM. Previous
studies have established that acculturation—as
measured by nativity/length of residence in the US
or English proficiency—is important in explaining
disparities in health outcomes and access to health
care between immigrants and native-born Amer-
icans as well as within different immigrant groups
(Abraido-Lanza, Chao, & Florez, 2005; Antecol &
Bedard, 2006; Evenson, Sarmiento, & Ayala, 2004;
Laroche, 2000; Leclere, Jensen, & Biddlecom, 1994;
Shah, Zhu, Wu, & Potter, 2006; Zambrana, Breen,
Fox, & Gutierrez-Mohamed, 1999). A converging
finding from these studies is that when immigrants
become more acculturated to the US, their health
behaviors as well as their access to health care tend
to gradually resemble those of native-born Amer-
icans. However, to what extent this finding holds in
the case of CAM use is still unclear. While previous
studies have revealed differential patterns of CAM
use across demographic and socioeconomic groups,
little is known about the role of acculturation in
CAM use (Barnes et al., 2004).
The objective of this study is to examine the linkages
between acculturation and CAM use among immi-
grants in the US. Specifically we assess how accultura-
tion is related to the odds of using at least one CAM
modality as well as the odds of using specific CAM
modalities in the past 12 months. We hypothesize that
as immigrants live longer in the US, they are more
likely to resemble native-born Americans in terms of
CAM use. Another related hypothesis is that immi-
grants who show a higher level of English proficiency
are more like native-born Americans in terms of CAM
use than those with a lower level of English proficiency.
These two hypotheses will be tested with and without
adjusting for the other relevant predictors of CAM use.
Methods
Data and sample
The data used in this study come from the 2002
National Health Interview Survey (NHIS). The
NHIS is a nationally representative, cross-sectional
survey of the civilian non-institutionalized US
population. It employs a multistage sampling design
and oversamples both African American and
Hispanic populations.Besides
demographics, health status, health behaviors, and
health care access and utilization, the 2002 NHIS
included a CAM supplement that collected data on
31,044 adults 18 years of age or older.
information on
Measures
CAM use
In the 2002 NHIS respondents were asked
whether during the past 12 months they had used
or seen a practitioner for CAM therapies. Several
studies reported 27 types of CAM therapies based
on information from the 2002 NHIS (e.g. Barnes
et al., 2004; Saydah & Eberhardt, 2006). These
CAM therapies included 17 major CAM types and
10 subtypes. In our study, we only focused on the 17
major CAM therapies without differentiating the
subtypes within each major CAM therapy. Based on
information on use of these 17 CAM therapies, we
created a dummy variable to measure whether
during the past 12 months a respondent had used
at least 1 of these 17 CAM therapies.
In terms of the use of specific CAM therapies, we
analyzed the 10 most common CAM therapies
which were used by at least 1% of the sample. Ten
dummy variables were created to denote whether a
respondent ever used each of the 10 therapies over
the past 12 months: prayer, herbal treatments,
relaxation techniques, chiropractic care, yoga-tai-
chi-qigong, massage, special diets, megavitamins,
homeopathy, and acupuncture. A major considera-
tion for analyzing the 10 most common CAM
therapies is that the number of respondents who
reported use of the other seven therapies is too small
(each less than 1% of the sample) to allow for a
reliable assessment of the association between
acculturation and CAM use. Our focus on the 10
most common CAM therapies in the 2002 NHIS is
different from several previous studies that have
analyzed CAM use in terms of CAM domains (e.g.
Mehta, Phillips, Davis, & McCarthy, 2007; Saydah
& Eberhardt, 2006; Upchurch et al., 2007). While
aggregating each single type of CAM into several
major CAM domains can simplify the presentation
of the results, it also prevents us from relating
acculturation to a specific type of CAM. In this
study, we opted for an analysis of the use of specific
ARTICLE IN PRESS
D. Su et al. / Social Science & Medicine 66 (2008) 439–453
440

Page 3

CAM therapies rather than CAM domains since the
former better utilized the richness of the NHIS data
and provided more detailed information. Using a
few generalized CAM domains may lead to
aggregation bias if there is substantial variability
in the use of specific CAM therapies across nativity
groups.
Acculturation
In this study, we used nativity/length of residence
in the US, and language of interview as indicators of
the level of acculturation. These measures have been
widely used in previous research as indicators of
acculturation, particularly in studies examining the
relationship between acculturation and health be-
haviors (Abraido-Lanza et al., 2005; Abraido-
Lanza, Chao, & Gates, 2005; Leclere et al., 1994;
Singh & Siahpush, 2002; Zambrana et al., 1999). In
the 2002 NHIS, respondents were asked their place
of birth and, for those who were foreign-born, how
long they had been living in the US. Based on these
two variables, we created a new categorical variable
‘‘nativity’’—1: native-born, 2: foreign-born and has
been in the US for more than 15 years, 3: foreign-
born and has been in the US for 10–15 years, 4:
foreign-born and has been in the US for 5–10 years,
and 5: foreign-born and has been in the US for less
than 5 years.
Information on the language of interview for the
2002 NHIS sample households was reported by the
interviewers at the end of household interviews, and
was coded as English only, English and Spanish,
Spanish only, and other (less than 1% of the
sample). For example, ‘‘English only’’ means that
each selected subject in a sample household was
interviewed only in English. Bilingual interviewers
or interpreters were recruited to interview respon-
dents who preferred the use of a language other
than English. For respondents from households that
were interviewed in both English and Spanish, the
2002 NHIS provides no information on which
household members were interviewed in English
and which ones were interviewed in Spanish.
However, it is reasonable to assume that the level
of acculturation for individuals from these house-
holds falls on average between the level of
acculturation for those who were interviewed only
in English and that for those who were interviewed
only in Spanish. Thus, without considering the
households interviewed in ‘other’ languages, a
plausible rank of the average level of acculturation
based on language of interview should be something
like this: respondents from households that were
interviewed only in English had the highest level of
acculturation, followed by those from households
interviewed in both English and Spanish, and then
by those from households interviewed exclusively in
Spanish.
Sociodemographic factors
Besides information on CAM use and accultura-
tion, the 2002 NHIS also provides detailed informa-
tion on the sociodemographic profiles of the
respondents and their households. Along with
acculturation measures, our analysis also incorpo-
rated the following explanatory variables to predict
CAM use: age, gender, race/ethnicity, educational
attainment, marital status, total family income,
health insurance coverage, employment status,
self-reported health status, history of chronic
disease, family size, and region of residence.
Analysis
We first estimated a set of logistic regressions to
examine the relationship between nativity/length of
residence in the US and use of at least one CAM
modality or one of the specific CAM modalities
over the past 12 months. Then we estimated these
regressions by incorporating other relevant predic-
tors of CAM use in an effort to evaluate to what
extent these added explanatory variables could alter
the effects of nativity/length of residence on CAM
use. Finally, we replicated these estimated regres-
sion models by replacing nativity/length of resi-
dence in the US with the language in which the
household interview was conducted, which allowed
us to assess the sensitivity in the relationship
between acculturation and CAM use for different
acculturation measures. The statistical analysis took
into account the complex survey design of the 2002
NHIS (stratification, clustering, and oversampling
in a multistage sampling design) by employing the
svy-based commands in Stata (StataCorp, 2005).
Results
Description of the key variables
Table 1 presents the mean or percentage distribu-
tion for all the variables used in this study. Almost
62% of respondents reported that they had used at
least 1 of the 17 CAM modalities in the past 12
months. In terms of the use of specific CAM
ARTICLE IN PRESS
D. Su et al. / Social Science & Medicine 66 (2008) 439–453
441

Page 4

modalities, prayer is the most common, with 44.2%
of the sample practicing it, followed by the use of
herbal remedies, relaxation techniques, and chir-
opractic care (19.6%, 15.3%, and 7.7%, respec-
tively).
Eighty-six percent of the sample respondents were
born in the US and 14% were born abroad. Among
the foreign-born, 7.1% had lived in the US for 15
years or more, 2.3% between 10 and 15 years, 2.0%
between 5 and 10 years, and 2.2% less than 5 years.
About 95% of the interviews were conducted in
English, 2.8% in Spanish, 1.6% in English and
Spanish, and 0.4% in languages other than English
or Spanish.
Nativity/length of stay in the US and CAM use
Table 2 presents both unadjusted and adjusted
odds ratios from logistic regressions of CAM use by
nativity. The odds ratios shown in Table 2 reflect
the gap in the odds of CAM use between the native-
born (the reference group) and other nativity
ARTICLE IN PRESS
D. Su et al. / Social Science & Medicine 66 (2008) 439–453
Table 1
Descriptive statistics for all the variables used in the analysis (N ¼ 31,044)
VariablesMean or %S.E. VariablesMean or %S.E.
CAM use (%)
At least one CAM
Prayer for health reasons
Herbal medicine
Relaxation techniques
Chiropractic care
Yoga-taichi-qigong
Massage
Special diets
Megavitamins
Homeopathy
Acupuncture
Nativity/length of stay in the US (%)
US born
Foreign born and stay
X15 years
10–15 years
5–10 years
o5 years
Language of interview (%)
English only
English and Spanish
Spanish only
Other
Age (%)
18–24 years
25–34 years
35–44 years
45–54 years
55–64 years
65 years or older
Gender (%)
Male
Female
Region (%)
Northeast
Midwest
South
West
Race/ethnicity (%)
Non-Hispanic white
African American
Hispanic
Asian
Other
Educational attainment (%)
oHigh school
High school graduate
Some college
College graduate or higher
Marital status (%)
Married
Divorced
Separated
61.8
44.2
19.6
15.3
7.7
6.1
5.1
3.7
2.9
1.8
1.1
0.4
0.4
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0.1
73.1
11.5
11.5
3.5
0.3
0.5
0.3
0.3
0.2
o0.1
16.6
28.9
29.8
24.7
0.3
0.4
0.4
0.4
57.5
10.4
2.5
0.4
0.2
0.1
85.9
7.1
0.3
0.2
2.3
2.0
2.2
0.1
0.1
0.1
Single/never married
Widowed
Family income (%)
o$ 20,000
$ 20,000–$ 44,999
$ 45,000–$ 74,999
$ 75,000 or over
Insurance status (%)
Uninsured
Insured
Employment status (%)
Not employed
Employed
Self-reported health status (%)
Excellent
Very good
Good
Fair
Poor
History of chronic disease (%)
No
Yes
Family size (weighted mean)
23.4
6.2
0.4
0.1
22.4
28.5
23.6
25.5
0.4
0.4
0.3
0.5
95.2
1.6
2.8
0.4
0.2
0.1
0.1
0.1
15.5
84.5
0.3
0.3 13.6
19.0
22.0
19.1
11.9
14.5
0.4
0.3
0.3
0.3
0.3
0.3
34.2
65.8
0.4
0.4
31.0
31.7
25.1
9.1
3.1
0.4
0.3
0.3
0.2
0.1
48.4
51.6
0.4
0.4
18.8
24.1
37.9
19.1
0.4
0.5
0.6
0.4
55.4
44.6
3.0
0.4
0.4
o0.1
Source: National Health Interview Survey (2002).
442

Page 5

ARTICLE IN PRESS
Table 2
Odds ratios of CAM use by nativity/length of residence in the US: United States, 2002 NHIS
TherapyUnadjusted odds ratioa
Adjusted odds ratiob
Native Foreign bornNativeForeign born
X15 years 10–15 years5–10 years
o5 years
X15 years10–15 years5–10 years
o5 years
At least one CAMc
Reference1.02
(0.91, 1.13)
1.05
(0.95, 1.17)
1.26***
(1.12, 1.42)
0.72***
(0.60, 0.86)
0.50***
(0.38, 0.64)
0.90
(0.70, 1.17)
0.74**
(0.57, 0.97)
0.99
(0.76, 1.29)
0.59**
(0.37, 0.94)
0.82
(0.57, 1.17)
1.28
(0.84, 1.95)
0.80**
(0.66, 0.97)
0.93
(0.77, 1.12)
0.83*
(0.67, 1.03)
0.53***
(0.37, 0.75)
0.48***
(0.30, 0.75)
0.77
(0.51, 1.17)
0.77
(0.52, 1.16)
0.90
(0.56, 1.44)
0.33***
(0.15, 0.71)
1.12
(0.61, 2.06)
1.50
(0.72, 3.12)
0.72***
(0.60, 0.86)
0.84*
(0.70, 1.00)
0.95
(0.74, 1.23)
0.47***
(0.33, 0.68)
0.21***
(0.10, 0.42)
0.84
(0.56, 1.26)
0.28***
(0.14, 0.57)
0.88
(0.52, 1.51)
0.25**
(0.07, 0.85)
0.79
(0.38, 1.66)
1.29
(0.53, 3.13)
0.63***
(0.53, 0.75)
0.75***
(0.64, 0.90)
0.82*
(0.66, 1.02)
0.68***
(0.51, 0.91)
0.08***
(0.03, 0.21)
0.65**
(0.44, 0.94)
0.26***
(0.14, 0.46)
0.59**
(0.37, 0.93)
0.22**
(0.06, 0.77)
1.01
(0.49, 2.05)
0.62
(0.28, 1.37)
Reference1.10
(0.96, 1.26)
1.08
(0.94, 1.23)
1.25***
(1.07, 1.47)
0.75***
(0.60, 0.93)
0.69**
(0.52, 0.93)
1.05
(0.74, 1.50)
0.94
(0.66, 1.33)
1.21
(0.88, 1.65)
0.76
(0.45, 1.30)
0.86
(0.54, 1.39)
1.00
(0.58, 1.75)
0.83*
(0.66, 1.03)
1.01
(0.82, 1.24)
0.73**
(0.56, 0.96)
0.58***
(0.38, 0.88)
0.63*
(0.38, 1.06)
0.78
(0.49, 1.25)
1.00
(0.62, 1.61)
1.16
(0.68, 1.97)
0.67
(0.29, 1.53)
1.04
(0.50, 2.19)
1.02
(0.46, 2.28)
0.82
(0.65, 1.04)
0.94
(0.74, 1.18)
1.04
(0.78, 1.40)
0.52***
(0.34, 0.78)
0.37***
(0.18, 0.76)
0.78
(0.47, 1.30)
0.36***
(0.17, 0.77)
1.34
(0.74, 2.41)
0.04***
(0.01, 0.31)
0.72
(0.29, 1.81)
1.32
(0.47, 3.73)
0.76**
(0.60, 0.94)
0.85
(0.69, 1.06)
0.94
(0.72, 1.23)
0.82
(0.59, 1.13)
0.13***
(0.04, 0.40)
0.58**
(0.36, 0.95)
0.37***
(0.19, 0.72)
0.65
(0.34, 1.26)
0.31
(0.05, 1.93)
1.12
(0.49, 2.55)
0.65
(0.27, 1.59)
Prayer for health reasonsReference Reference
Herbal medicineReferenceReference
Relaxation techniquesReferenceReference
Chiropractic careReferenceReference
Yoga-taichi-qigong ReferenceReference
MassageReferenceReference
Special dietsReferenceReference
MegavitaminsReferenceReference
Homeopathy ReferenceReference
AcupunctureReference Reference
Source: National Health Interview Survey (2002).
*po0.1.
**po0.05.
***po0.01.
aThe estimation of unadjusted odds ratios was based on bivariate logistic regression models.
bAdjusted odds ratios were estimated after controlling for the effects of age, gender, race/ethnicity, education attainment, total family income, marital status, self-reported health
status, region of residence in the US, health insurance coverage, employment status, history of chronic disease, and family size.
cBased on 17 different types of CAM therapies.
D. Su et al. / Social Science & Medicine 66 (2008) 439–453
443