Journal of Physical Activity and Health, 2013, 10, 222-231
© 2013 Human Kinetics, Inc.
Official Journal of ISPAH
www.JPAH-Journal.com
ORIGINAL RESEARCH
Age Bias in Physicians’ Recommendations
for Physical Activity: A Behavioral Model
of Healthcare Utilization for Adults With Arthritis
Shamly Austin, Haiyan Qu, and Richard M. Shewchuk
Objective: To examine whether age bias exists in physicians’ recommendations for physical activity among
individuals with arthritis. Methods: A cross-sectional sample with 33,071 U.S. adults, 45 years or older with
physician-diagnosed arthritis was obtained from 2007 Behavioral Risk Factor Surveillance System Survey.
We used logistic regression to examine physicians’ recommendations for physical activity as a function of
age controlling for gender, race, education, marital status, employment, income, health insurance, personal
physician, emotional support, body mass index, activity limitations, health status, and comorbidities. Results:
Majority of individuals were females (65%), White (85%), had annual household income < $50,000 (67%),
and with comorbidities (86%). Respondents were approximately equal across age groups: middle-aged group
(53%) and older group (47%). About 36% were obese and 44% had activity limitations, and 44% did not
receive any physicians’ recommendations for physical activity. Results from logistic regression indicated older
adults (≥ 65 years old) were less likely (OR = 0.87; 95% CI, 0.82–0.92) to receive physicians’ recommendations for physical activity compared with the middle-aged group (45–64 years old). Conclusions: This study
indicates that although the benefits associated with the physical activity is well recognized, there is age bias
in physicians’ recommendations for physical activity.
Keywords: physician-diagnosed arthritis, guidelines, older adults, middle-aged adults
According to the 2008 Physical Activity Guidelines
Advisory Committee Report, the Musculoskeletal Health
Subcommittee of the Department of Health and Human
Services recommends that physicians should counsel
individuals with arthritis to pursue physical activities
that are low-to-moderate intensity, not painful, and do
not have a high risk of joint injury. A recommendation of
30 minutes of physical activity for 5 days a week is considered appropriate for most individuals with arthritis.1
Even though physical activity is considered beneficial for reducing pain and improving the health-related
quality of life for individuals with arthritis, there is a high
prevalence of physical inactivity among this population.2
One of the reasons for the high prevalence of physical
inactivity among this population could be lack of physicians’ recommendations for physical activity. Physicians’
recommendations act as a catalyst for changes in health
behavior related to physical activity.3,4 A number of studies in the arthritis literature evaluated individual characteristics associated with physicians’ recommendations for
physical activity.5–7 These studies were consistent in their
finding that females, obese, those with some or higher
college education, and those with activity limitations were
Austin is with the Dept of Critical Care Medicine, University
of Pittsburgh, Pittsburgh, PA. Qu and Shewchuk are with the
Dept of Health Services Administration, University of Alabama
at Birmingham.
222
more likely to receive the recommendations. However,
results on whether older adults received physicians’ recommendations were inconsistent. Some studies reported
that older adults were more likely to receive physicians’
recommendations compared with middle-aged adults.5–7
However, a few other studies in the arthritis literature
reported that physicians were less likely to recommend
physical activity to older adults.8–10 Thus, it was inconclusive whether physicians were recommending physical
activity to older adults with arthritis.
The Behavioral Model of Health Services Utilization
has been widely used to predict health services utilization.11–13 According to the model, we assumed that the
receipt of physicians’ recommendations for individuals with arthritis would be predicted by predisposing,
enabling and need-related factors (Figure 1). We used
the model to explain the inequitable receipt of physical
activity recommendations by older adults compared
with middle-aged adults. Physicians tend to recommend
physical activity to some of their patients. They recommend physical activity to individuals who are more likely
to follow their recommendations. Generally, old age is
accompanied by comorbidities, activity limitations, and
poor social/emotional support.14 Hence, physicians may
think that older adults might not follow their recommendations for physical activity5 and may be less likely to
recommend physical activity to older adults compared
with middle-aged adults. The prevalence of arthritis
Physicians’ Recommendations for Physical Activity 223
Figure 1 — Conceptual framework based on the behavioral model of healthcare utilization depicting association between individuals’ ages and physicians’ recommendations for physical activity.
increases with age.15 This study included only adults
aged 45 years or older. Therefore, the objective of the
study was to examine whether older adults (≥ 65 years
old) with arthritis were less likely to receive physicians’
recommendations for physical activity compared with
middle-aged adults (45–64 years old) after controlling
for need-related, enabling, and other predisposing factors.
Methods
Study Population and Sample
The study had a retrospective cross-sectional design
based on the 2007 Behavioral Risk Factor Surveillance
System (BRFSS). The BRFSS is a random-digit dial
telephone survey administered by the Centers for Disease
Control and Prevention (CDC) for the noninstitutionalized U.S. civilian adult population who are 18 years of
age or older.16 A detailed description of the survey is
documented elsewhere.16 The items from the BRFSS
questionnaire were shown to be valid and reliable.17–21
The arthritis management module was administered
by 19 states in 2007. These 19 states were Alabama,
Alaska, Arizona, California, Connecticut, Florida,
Georgia, Illinois, Indiana, Iowa, Kentucky, Minnesota,
New Mexico, New York, Pennsylvania, Rhode Island,
South Carolina, Virginia, and West Virginia.22 In 2007
BRFSS data, 49,706 respondents reported themselves
having physician-diagnosed arthritis. The final sample
size included 33,071 respondents who had complete
information on the key variables in analysis model
(Figure 2). We did not find individuals in our study to be
Figure 2 — Flowchart for inclusion of individuals in the study.
224
Physicians’ Recommendations for Physical Activity 225
significantly different from those who were excluded on
physicians’ recommendations for physical activity (Φ <
0.01, P = .86), age (Φ = 0.01, P = .10), and comorbidities (Φ = 0.007, P = .36). However, these 2 groups were
different in the proportion of gender (Φ = 0.01, P <
.05), the excluded group (n = 16,635) had more females
relative to the included group. The Greek letter Φ is the
phi-correlation coefficient used to assess the strength of
association between nominal variables.
The bivariate association between individuals’ ages
and physicians’ recommendations for physical activity
was analyzed using chi-square. Because the dependent
variable was a binary variable, multivariate logistic
regression was used to analyze the association between
individuals’ ages and physicians’ recommendations for
physical activity controlling for need-related, enabling,
and predisposing factors. The results were interpreted
as odds ratios. The Hosmer-Lemeshow test was used to
assess the goodness of fit of the model.
Variables and Measurements
The dependent variable was physicians’ recommendations for physical activity. This variable was measured
using the item “has a doctor or other health professional
ever suggested physical activity or exercise to help your
arthritis or joint symptoms.” The response to this item
was dichotomous (yes/no). The independent variable was
age of the individual with arthritis. Age was categorized
as middle-aged adults and older adults.
The control variables in the model were: 1) predisposing factors: gender, race (White vs. non-White);
education (≤ high school vs. > high school); marital status
(married vs. unmarried); employment status (employed
vs. unemployed); 2) enabling factors: annual household
income (< $50,000 vs. ≥ $50,000); had health insurance (yes vs. no); had personal physician (yes vs. no);
had emotional support (yes vs. no); and 3) need-related
factors: body mass index (BMI; nonobese—BMI < 30
vs. obese—BMI ≥ 30); had activity limitations (yes
vs. no); health status (good vs. poor); and presence of
comorbidities (yes vs. no). We computed comorbidities
by summing up the responses to diabetes, hypertension,
high cholesterol, myocardial infarction, angina/coronary heart disease, stroke, asthma, and depression. All
affirmative responses for the above medical conditions
were coded as “1,” and negative responses were coded
as “0,” such that comorbidities ranged from 0–8 when
summed up. Further, individuals with 1–8 comorbidities
were categorized as those who had comorbidities (yes)
and individuals with 0 comorbidities were categorized as
those with no comorbidities (no). Kaplan and colleagues
adopted similar classification in a study on characteristics
of physically inactive older adults with arthritis.23
Data Analysis
The study was approved by the Institutional Review
Board of the University of Alabama at Birmingham
(protocol # N090121006). For data management and
analyses, we used the Statistical Package for Social
Sciences (SPSS) Version 17.0. To analyze the sample
characteristics and associations between dependent and
independent variables, univariate and bivariate statistical tests were used. Multicollinearity was not observed
between the covariates (Table 1). The minimum and
maximum correlations observed among the covariates
were –0.20 and 0.39, respectively.
Results
Table 2 depicts the characteristics of the respondents
in the study. Of these 33,071 respondents, 44% did not
receive any physicians’ recommendations for physical
activity, 47% were older adults, 65% were females, 15%
were non-Whites, 67.4% had an annual income less than
$50,000, 83% had personal physician, 36% were obese,
44% had activity limitations, and 86% had comorbidities.
The respondents in older and middle-aged groups were
significantly different in education (P = .01), employment
(P = .04), and activity limitations (P = .01; Table 2).
Results from the chi-square analysis indicated an
association between physicians’ recommendations for
physical activity and age (Φ = 0.05, P < .001). Results
from the multivariate logistic regression were shown in
Table 3. Older respondents were less likely (OR = 0.87;
95% CI, 0.82–0.92) to receive physicians’ recommendations for physical activity compared with the middle-aged
ones after controlling for need-related, enabling, and other
predisposing factors. Among the predisposing factors, all
variables (gender, race, education, and marital status)
except employment status were significant predictors of
physicians’ recommendations. The strongest predictors
were gender and race. Females were more likely (OR =
1.42; 95% CI, 1.35–1.49) to receive physicians’ recommendations compared with males. Non-Whites were
more likely (OR = 1.32; 95% CI, 1.24–1.41) to receive
physicians’ recommendations for physical activity compared with Whites. All of the enabling factors (income,
health insurance, personal physician, and emotional
support) were significant predictors of physicians’ recommendations. The strongest predictors were income and
having a personal physician. Respondents with an annual
household income of < $50,000 were less likely (OR=
0.89; 95% CI, 0.84–0.95) to receive physicians’ recommendations for physical activity compared with those
with annual incomes ≥ $50,000. Compared with those
who had a personal physician, respondents who did not
have one were less likely (OR = 0.88; 95% CI, 0.83–0.93)
to receive recommendations for physical activity. Except
for health status, all other need-related factors (BMI,
activity limitations, and comorbidities) were significant
predictors of physicians’ recommendations. BMI was
the strongest predictor among the need-related factors.
Obese respondents were more likely (OR = 1.72; 95% CI,
226
Gender
Race
Education
Marital Status
Employment
Income
Health Insurance
Personal Physician
Emotional Support
BMI
Activity Limitation
Health Status
Comorbidities
2
3
4
5
6
7
8
9
10
11
12
13
14
0.05*
0.00
–0.04*
–0.12*
–0.04*
–0.02*
–0.20*
0.20*
0.44*
0.12*
0.09*
–0.09*
0.01
1
1
a
Phi correlation coefficient.
* Correlations were significant at 0.05 level.
Age
1
Variables
0.01*
0.01*
0.01*
0.01*
0.00
–0.04*
0.00
0.13*
0.03*
0.20*
0.05*
0.03*
1
2
0.04*
0.11*
0.00
0.08*
0.13*
0.02*
0.08*
0.12*
0.00
0.09*
0.07*
1
3
0.06*
0.20*
0.04*
0.06*
0.12*
0.02*
0.07*
0.34*
0.15*
0.07*
1
4
0.05*
0.11*
0.09*
0.00
0.15*
0.02*
0.04*
0.35*
0.10*
1
5
0.13*
0.23*
0.20*
–0.02*
0.06*
0.02*
–0.05*
0.31*
1
6
Table 1 Correlation Matrixa for the Independent Variables (n = 33,071)
0.11*
0.26*
0.16*
0.05*
0.18*
0.05*
0.12*
1
7
–0.01
0.08*
0.03*
0.03*
0.09*
0.11*
1
8
–0.04*
0.09*
0.06*
0.00
0.05*
1
9
0.08*
0.19*
0.12*
0.04*
1
10
0.13*
0.13*
0.13*
1
11
0.14*
0.39*
1
12
0.19*
1
13
1
14
Table 2 Variables, Measurements, Frequencies, and P-value
Variables and measurement
Physicians’ recommendations
No
Yes
Independent variables
Predisposing factors
Age
Older adults (≥ 65 years)
Middle-aged adults (45–64 years)
Gender
Female
Male
Race
Non-White
White
Education
≤ HS
> HS
Marital status
Unmarried
Married
Employment
Unemployed
Employed
Enabling factors
Income
< $50,000
≥ $50,000
Health insurance
No
Yes
Personal physician
No
Yes
Emotional support
No
Yes
Need-related factors
BMI
Obese
Nonobese
Activity limitation
Yes
No
Health status
Poor
Good
Comorbidities
Yes
No
Total
(n = 33,071)
(%)
≥ 65 yrs
(n = 15,464)
(%)
45–64 yrs
(n = 17,607)
(%)
P-value
43.9
56.1
46.6
53.4
41.4
58.6
0.49
46.8
53.2
–
–
–
–
–
–
65.0
35.0
65.5
34.5
64.6
35.4
0.55
15.2
84.8
11.7
88.3
18.3
81.7
0.70
47.1
52.9
51.9
48.1
42.8
57.2
0.01*
48.0
52.0
54.6
45.4
42.3
57.7
0.37
66.1
33.9
88.4
11.6
46.5
53.5
0.04*
67.4
32.6
77.4
22.6
58.7
41.3
7.1
92.9
1.4
98.6
12.1
87.9
0.06
16.7
83.3
15.6
84.4
17.6
82.4
0.05
24.6
75.4
22.4
77.6
26.5
73.5
0.91
35.5
64.5
29.0
71.0
41.2
58.8
0.26
44.2
55.8
42.0
58.0
46.3
53.7
0.01*
33.6
66.4
33.8
66.2
33.5
66.5
0.07
86.1
13.9
88.3
11.7
84.2
15.8
0.08
0.41
* P < .05.
227
228 Austin, Qu, and Shewchuk
Table 3 Association Between Individuals’ Ages and Physicians’
Recommendations for Physical Activity Controlling for Covariates
Among Individuals With Arthritis (n = 33,071)
Variable (reference category)
OR (95% CI)
≥ 65 yrs
0.87 (0.82–0.92)**
Gender (Male)
Female
1.42 (1.35–1.49)**
Race (White)
Non-White
1.32 (1.24–1.41)**
Education (> HS)
≤ HS
0.83 (0.80–0.88)**
Marital status (married)
Unmarried
0.92 (0.88–0.97)**
Employment (employed)
Unemployed
Age (45–64 yrs)
Control variables
Predisposing factors
1.03 (0.97–1.09)
Enabling factors
Income (≥ $50,000)
< $50,000
0.89 (0.84–0.95)**
Health insurance (yes)
No
0.83 (0.75–0.90)**
Personal physician (yes)
No
0.88 (0.83–0.93)**
Emotional support (yes)
No
0.87 (0.83–0.92)**
BMI (nonobese)
Obese
1.72(1.64–1.81)**
Activity limitation (no)
Yes
1.28 (1.22–1.35)**
Health status (good)
Poor
1.00 (0.94–1.05)
Comorbidities (no)
Yes
1.33 (1.24–1.42)**
Need-related factors
Abbreviations: OR, Odds Ratio; CI, Confidence Interval.
** P < .001.
1.64–1.81) to receive physicians’ recommendations for
physical activity compared with nonobese respondents.
The Hosmer-Lemeshow goodness of fit test (χ2 =5.85, P
= .66) indicated the model to be a good fit.
Discussion
This study found that older adults were less likely to
receive physicians’ recommendations for physical activity compared with the adults in the middle-aged group
after controlling for need-related, enabling, and other
predisposing factors.
According to the physical activity guidelines, it is
not harmful for individuals with arthritis to participate in
low-to-moderate levels of physical activity.1,24 Further,
there is evidence that individuals with arthritis had relief
from pain, improved physical function, and delayed
onset of disability by engaging in low-to-moderate
level physical activity of approximately 150 minutes
per week, which is 5 times per week for 30 minutes
per session.1 Individuals who engage in any amount of
physical activity gain some health benefits.25,26 Older
respondents with comorbidities who could not meet
the 150 minutes of physical activity a week are advised
to be physically active as their conditions and abilities
allow.26 In addition, according to the American College
of Rheumatologists, the nonpharmacological therapy section of the clinical practice guideline (CPG) recommends
physical activity for individuals with osteoarthritis and
maintains that pharmacological therapy is most effective
when combined with the nonpharmacological therapy.27
Despite the guidelines and evidence in literature, the
study found that older adults with arthritis were less
likely to receive physicians’ recommendations for physical activity compared with the middle-aged adults. This
finding was similar to studies that examined the factors
associated with physicians’ recommendations.5,28 A
review of Cabana’s (1999) work helps us to understand
the reasons why physicians do not adhere to CPG and
why they hesitate to recommend physical activity to older
adults. Cabana and colleagues categorized the barriers to
physicians’ adherence to CPG into knowledge-attitude–
and behavior-related barrier groups. Knowledge-related
barriers among physicians include the lack of awareness
and familiarity due to a large volume of information in
the guidelines, lack of time to go through them, and
inaccessibility of guidelines. Attitude-related barriers
among physicians include the lack of agreement with
guidelines, outcome expectancy, self-efficacy, and inertia
of previous practice. Physicians’ lack of agreement may
Physicians’ Recommendations for Physical Activity 229
be in the interpretation of evidence, applicability to the
patient, cost benefits, lack of confidence in the guideline
developer, rigidity in application, or not practical. Lack
of outcome expectancy occurs when a physician believes
that performance of the guideline recommendation will
not provide expected outcomes. Lack of self-efficacy
occurs when a physician feels that s/he cannot perform
the guidelines or it is simply due to lack of motivation or
inertia from previous practice. In addition, the external
barriers to physicians’ adherence to the guideline were
patient preferences with guideline recommendations or
guideline characteristics, such as the presence of contradictory statements, or environmental factors, such as
lack of time, resources, reimbursements, organizational
constraints, and perceived increase in malpractice liability.29 These barriers may play a role in physicians not
recommending physical activity to older adults.
The additional results from our study reinforced
the mounting evidence in literature that females,5–7
non-White,5,28 unemployed,30 obese,31 individuals with
activity limitations6 and comorbidities28 were more likely
to receive physicians’ recommendations for physical
activity; whereas, adults who were unmarried,7,28 with
low education,5,7,28 low income,31 without health insurance,7,28 without personal physician,32 and without emotional support30 were less likely to receive physicians’
recommendations for physical activity. We observed
that physicians tend to recommend physical activity to
respondents with poor medical/health conditions (obese,
activity limitations, comorbidities). Further, the 2009
Current Population Survey indicated that majority of
older respondents had low income than their middle-aged
counterparts.33 Hence, middle-aged respondents may
represent as better clients to physicians due to reimbursement issues and these individuals may receive more time
and better quality consultation especially for lifestyle
related behavior changes compared with older adults.
Previous research has shown that individuals who did not
have health insurance and personal physician were less
likely to get regular and better care.7,28 Because of the
lack of accessibility to health care, these individuals may
be less likely to receive recommendations for physical
activity from physicians. Fontaine and colleagues suggested that physicians recommended physical activity to
only those individuals whom they thought might follow
their recommendations.5 Individuals with some sort of
social or emotional support were reported to engage in
lifestyle behavior changes than those who did not have
one.34 Therefore, one of the reasons why physicians did
not recommend physical activity to unmarried individuals and those without any emotional support may be they
believed that these individuals might not follow their
recommendations.
The study findings are important for individuals with
arthritis especially for older adults, healthcare providers,
researchers, and health policy makers. This study indicates that despite the physical activity guideline recommendations, there is age bias in physicians’ recommendations for physical activity. Hence, the findings imply
an issue of poor quality of care provided. We recognize
that the retrospective cross-sectional design of our study
can assume associations alone and there are a number
of other factors that contribute to quality of care besides
physicians’ recommendations for physical activity. However, results from our study indicate that there is room
for improvement related to physical activity counseling
by physicians in this population.
Some policy-level interventions at the organizational
level to facilitate physicians’ recommendations for physical activity are training physicians to comply with CPGs,
emphasis on written physician order for physical activity
rather than verbal instructions, provision of incentives
to physicians who recommend physical activity to their
patients, and identify physicians’ barriers in adherence to
CPGs. At the population level, interventions may include
availability of low-cost, centrally located facilities for
exercise and providing physical activity/exercise classes
and programs focused on older adults.
The study results should be considered in the light
of some limitations. The cross-sectional nature of the
study does not allow determining causality. Households without telephones and individuals in institutions and the military were not included in the survey.
In addition, because the responses were self-reports,
social desirability and recall bias in responses were
possible. Respondents’ physician-diagnosed arthritis
were not confirmed through a health professional or
medical records.35 There would be differences in activity limitations among the respondents based on the
type of arthritis they have, but the survey did not have
information on different types of arthritis. In addition,
individuals who reported joint pain/symptoms but
were not diagnosed by a physician for arthritis were
excluded, thus the study may have underestimated
the burden of arthritis.36 The BRFSS item measured
whether the physician recommended physical activity
or not, but it did not measure whether the physician
recommended 30 minutes of physical activity 5 times
a week. Some of the possible confounders for the study
were individuals’ past exercise behavior, number of
visits to the clinic, patient-physician relationship, and
contextual variables, such as recreational facilities
available, crime in the neighborhood, physical activity
habits and physicians’ gender, gender, race or age parity
between the interacting patients and their physicians.
The study examined whether there were any differences
between the included (n = 33,071) and the excluded (n
= 16,635) groups on the variables of age, gender, and
comorbidities to examine any potential bias. We did
not observe significant differences in respondents’ age
and status of comorbidities between these 2 groups.
However, these 2 groups were significantly different
by gender. Further, the major reason for exclusion of
cases from our study was incomplete information on
income. Logistic regression was conducted by including the cases with missing values on income as a third
category. We did not observe any differences between
the model and our study model in terms of magnitude
230 Austin, Qu, and Shewchuk
and direction of associations between individuals’ ages
and physicians’ recommendations for physical activity.
Previous research focused on race and gender bias;
however, age bias received less attention. To our knowledge, this was the first time a study examined age bias in
physicians’ recommendations for physical activity in a
population of adults with arthritis based on a theoretical
framework. The results from this study are generalizable,
since they are based on a nationally representative sample
of US adults administered by the CDC.25,37 Future studies should focus on the association between physicians’
recommendations for physical activity and adherence to
recommended levels of physical activity, barriers encountered by physicians in recommending physical activity
to older adults with arthritis, identify the motivators of
physical activity in older adults with arthritis, and identify the factors associated with long-term involvement
in physical activity in this population. In addition, studies should analyze the association between physicians’
recommendations for physical activity among young-old
(65–74 years), middle-old (75-84 years), and old-old
adults (85 years or older); and examine ageism in physicians’ recommendations for health-promoting behavior.
References
1. Physical Activity Guidelines Advisory Committee. Physical activity guidelines advisory committee report, 2008.
Washington, DC: Department of Health and Human
Services; 2008.
2. Fontaine KR, Heo M, Bathon J. Are US adults with arthritis meeting public health recommendations for physical
activity? Arthritis Rheum. 2004;50:624–628. PubMed
doi:10.1002/art.20057
3. Kreuter MW, Chheda SG. Bull. How does physician advice
influence patient behavior?: evidence for a priming effect.
Arch Fam Med. 2000;9:426–433. PubMed doi:10.1001/
archfami.9.5.426
4. Petrella RJ, Koval JJ, Cunningham DA, Paterson DH. Can
primary care doctors prescribe exercise to improve fitness?
Am J Prev Med. 2003;24:316–322. PubMed doi:10.1016/
S0749-3797(03)00022-9
5. Fontaine KR, Bartlett SJ, Heo M. Are health care professionals advising adults with arthritis to become more
physically active? Arthritis Rheum. 2005;53:279–283.
PubMed doi:10.1002/art.21073
6. Fontaine KR, Haaz S, Bartlett SJ. Are overweight and
obese adults with arthritis being advised to lose weight? J
Clin Rheumatol. 2007;13:12–15. PubMed doi:10.1097/01.
rhu.0000256168.74277.15
7. Honda K. Factors underlying variation in receipt of
physician advice on diet and exercise: applications of the
behavioral model of health care utilization. Am J Health
Promot. 2004;18:370–377. PubMed doi:10.4278/08901171-18.5.370
8. Hirvensalo M, Heikkinen E, Lintunen T, Rantanen T. The
effect of advice by health care professionals on increas-
ing physical activity of older adults. Scand J Med Sci
Sports. 2003;13:231–236. PubMed doi:10.1034/j.16000838.2003.00313.x
9. Hirvensalo M, Heikkinen E, Lintunen T, Rantanen T.
Recommendations for and warnings against physical
activity given to older people by health care professionals. Prev Med. 2005;41:342–347. PubMed doi:10.1016/j.
ypmed.2004.11.020
10. Schonberg MA, Marcantonio ER, Wee CC. Receipt
of exercise counseling by older women. J Am Geriatr
Soc. 2006;54:619–626. PubMed doi:10.1111/j.15325415.2006.00679.x
11. Andersen RM. Revisiting the behavioral model and access
to medical care: does it matter? J Health Soc Behav.
1995;36:1–10. PubMed doi:10.2307/2137284
12. Gochman DS. Handbook of health behavior research i:
personal and social determinants. Spring Street, New
York: Springer; 1997.
13. Murray MD, Morrow DG, Weiner M, et al. A conceptual
framework to study medication adherence in older adults.
Am J Geriatr Pharmacother. 2004;2:36–43. PubMed
doi:10.1016/S1543-5946(04)90005-0
14. Alliance for Aging Research. Ageism: How health care
fails the elderly, 2003. www.agingresearch.org, Accessed
November 25, 2006.
15. Hanmer J, Hays RD, Fryback DG. Mode of administration
is important in US national estimates of health-related
quality of life. Med Care. 2007;45:1171–1179. PubMed
doi:10.1097/MLR.0b013e3181354828
16. Centers for Disease Control and Prevention. Behavioral
Risk Factor Surveillance System: Operational and User’s
Guide, Version 3.0. Atlanta, Georgia; 2006.
17. Brownson RC, Jackson-Thompson J, Wilkerson JC, Kiani
F. Reliability of information on chronic disease risk factors
collected in the Missouri Behavioral Risk Factor Surveillance System. Epidemiology. 1994;5:545–549. PubMed
18. Martin LM, Leff M, Calonge N, Garrett C, Nelson
DE. Validation of self-reported chronic conditions and
health services in a managed care population. Am J Prev
Med. 2000;18:215–218. PubMed doi:10.1016/S07493797(99)00158-0
19. Nelson DE, Holtzman D, Bolen J, Stanwyck CA, Mack
KA. Reliability and validity of measures from the
Behavioral Risk Factor Surveillance System (BRFSS).
Soz Praventivmed. 2001;46(Suppl 1):S3–S42. PubMed
doi:10.1007/BF01318790
20. Nelson DE, Powell-Griner E, Town M, Kovar MG. A
comparison of national estimates from the National Health
Interview Survey and the Behavioral Risk Factor Surveillance System. Am J Public Health. 2003;93:1335–1341.
PubMed doi:10.2105/AJPH.93.8.1335
21. Stein AD, Lederman RI, Shea S. The Behavioral Risk
Factor Surveillance System questionnaire: its reliability in
a statewide sample. Am J Public Health. 1993;83:1768–
1772. PubMed doi:10.2105/AJPH.83.12.1768
22. Centers for Disease Control and Prevention. Questionnaires-Modules by category. Behavioral Risk Factor
Surveillance System 2007. http://apps.nccd.cdc.gov/
BRFSSModules/ModByCat.asp?Yr=2007. Accessed June
14, 2011.
Physicians’ Recommendations for Physical Activity 231
23. Kaplan MS, Huguet N, Newsom JT, McFarland BH. Characteristics of physically inactive older adults with arthritis:
results of a population based study. Prev Med. 2003;37:61–
67. PubMed doi:10.1016/S0091-7435(03)00059-8
24. Callahan LF, Mielenz T, Freburger J, et al. A randomized
controlled trial of the people with arthritis can exercise
program: symptoms, function, physical activity, and psychosocial outcomes. Arthritis Rheum. 2008;59:92–101.
PubMed doi:10.1002/art.23239
25. Abell JE, Hootman JM, Zack MM, Moriarity D, Helmick
CG. Physical activity and health related quality of life
among people with arthritis. J Epidemiol Community Health. 2005;59:380–385. PubMed doi:10.1136/
jech.2004.028068
26. US Department of Health and Human Services. 2008
Physical activity guidelines for Americans. Washington,
DC: Office of Disease Prevention and Health Promotion,
DHHS; 2008.
27. American College of Rheumatology Subcommittee on
Osteoarthritis Guidelines. Recommendations for the
medical management of Osteoarthritis of the hip and
knee. Arthritis Rheum. 2000;43:1905–1915. PubMed
doi:10.1002/1529-0131(200009)43:9<1905::AIDANR1>3.0.CO;2-P
28. Glasgow RE, Eakin EG, Fisher EB, Bacak SJ, Brownson
RC. Physician advice and support for physical activity:
results from national survey. Am J Prev Med. 2001;21:189–
196. PubMed doi:10.1016/S0749-3797(01)00350-6
29. Cabana MD, Rand CS, Powe NR, et al. Why don’t physicians follow clinical practice guidelines? A framework
for improvement. JAMA. 1999;282:1458–1465. PubMed
doi:10.1001/jama.282.15.1458
30. Loureiro ML, Nayga RM, Jr. Obesity, weight loss, and
physician’s advice. Soc Sci Med. 2006;62:2458–2468.
PubMed doi:10.1016/j.socscimed.2005.11.011
31. Wee CC, McCarthy EP, Davis RB, Phillips RS. Physician
counseling about exercise. JAMA. 1999;282:1583–1588.
PubMed doi:10.1001/jama.282.16.1583
32. Bull FC, Jamrozik K. Advice on exercise from a family
physician can help sedentary patients to become active.
Am J Prev Med. 1998;15:85–94. PubMed doi:10.1016/
S0749-3797(98)00040-3
33. Current Population Survey. Annual social and economic
(ASEC) supplement,2009. http://www.census.gov/hhes/
www/cpstables/032010/perinc/new09_000.htm. Accessed
June 14, 2011.
34. Dishman RK, Sallis JF, Orenstein DR. The determinants
of physical activity and exercise. Public Health Rep.
1985;100:158–171. PubMed
35. Freelove-Charton J, Bowles HR, Hooker S. Health related
quality of life by level of physical activity in arthritic older
adults with and without activity limitations. J Phys Act
Health. 2007;4:481–491. PubMed
36. Fontaine KR, Haaz S, Heo M. Projected prevalence of
US adults with self-reported doctor diagnosed arthritis,
2005-2050. Clin Rheum. 2007;26:772–774. doi:10.1007/
s10067-007-0556-7
37. Centers for Disease Control and Prevention. Measuring
healthy days: population assessment of health related
quality of life. Atlanta, Georgia: Centers for Diseases
Control and Prevention; 2000.