PCD-601; No.

of Pages 10
ARTICLE IN PRESS
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx

Contents lists available at ScienceDirect

Primary Care Diabetes

journal homepage: http://www.elsevier.com/locate/pcd

Original research

Correlates of physical activity counseling by health

providers to patients with diabetes and
hypertension attended by the Family Health
Strategy in the state of Pernambuco, Brazil

Jessyka Mary Vasconcelos Barbosa a,∗ , Wayner Vieira de Souza a ,

Renan Williams Marques Ferreira a , Eduardo Maia Freese de Carvalho a ,
Eduarda Angela Pessoa Cesse a , Annick Fontbonne b
a Department of Community Health, Aggeu Magalhaes Research Center, Fiocruz Pernambuco, Av. Prof. Moraes Rego,
S/N, Cidade Universitária, 50670-420 Recife, PE, Brazil
b IRD—Institute of Research for Development, UMR 204 Nutripass, 911 Avenue Agropolis, BP 64501, 34394

Montpellier Cedex 5, France

a r t i c l e i n f o a b s t r a c t

Article history: Aims: The aim of this study was to determine correlates of physical activity (PA) counseling by
Received 11 December 2016 health providers of the Brazilian primary care delivery system, for hypertensive and diabetic
Received in revised form subjects, as well as correlates of actual leisure-time PA of these subjects.
3 April 2017 Methods: This was a cross-sectional study conducted in random samples of 785 hypertensive
Accepted 5 April 2017 and 822 diabetic subjects, in the State of Pernambuco, Brazil. Relationships between PA
Available online xxx counseling and leisure-time PA and explanatory variables were sought through multiple
logistic regressions.
Keywords: Results: PA counseling had been received by 59.4% of the diabetic and 53.0% of the hyper-
Counseling tensive subjects; around 30% of the diabetic and the hypertensive subjects declared having
Physical activity leisure-time PA. After adjustment, factors associated with PA counseling for diabetic subjects
Diabetes mellitus were: female gender, formal schooling, hypertension, obesity; for hypertensive subjects:
High blood pressure being on a weight-loss diet, age between 60 and 74 and over 75. For both subject groups,
Brazil leisure-time PA was more frequent when they lived in a medium-sized municipality, were
female, aged between 60 and 75, and on a weight-loss diet.
Conclusion: PA counseling appeared restricted to subjects with excess weight and/or car-
diovascular risk factors when it should be directed to all subjects with hypertension or
diabetes.
© 2017 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.

Abbreviations: FHS, Family Health Strategy; CHA, community health agents; PA, physical activity; SERVIDIAH, Evaluation of Health SER-
VIces for DIAbetic and Hypertensive Subjects; BP, blood pressure; CVD, cardiovascular disease; VIGITEL, Surveillance System of Protective
and Risk Factors for Chronic Diseases Telephone Survey in Brazil.
∗
Corresponding author at: NESC/CPqAM/Fiocruz, Av. Prof. Moraes Rego, S/N, Cidade Universitária, 50670-420 Recife, PE, Brazil.
E-mail addresses: jessykamary@yahoo.com.br, jessykamary@gmail.com (J.M.V. Barbosa), wayner@cpqam.fiocruz.br (W.V.d. Souza),
renan.williams@hotmail.com (R.W.M. Ferreira), freese@cpqam.fiocruz.br (E.M.F.d. Carvalho), educesse@cpqam.fiocruz.br (E.A.P. Cesse),
annick.fontbonne@idr.fr (A. Fontbonne).
http://dx.doi.org/10.1016/j.pcd.2017.04.001
1751-9918/© 2017 Primary Care Diabetes Europe. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
2 p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx
1. Introduction
The Brazilian national health system has a community-based
approach to providing primary healthcare through an inter-
disciplinary health team, free of charge for all citizens; and,
nowadays, at least 62% of the population is covered by the
Family Health Strategy (FHS) designed to perform several pri-
mary care functions. Patients registered within FHS units are
attended by a health team consisting of a physician, a nurse, a
nurse assistant, and several community health agents—CHA.
They get a monthly home visit by the CHA for family health
orientations and to make appointment with the physician
when required; and health promotion activities are regularly
performed at the FHS unit [1,2].
One important health promotion procedure is physical
activity (PA) counseling [3,4]. It is estimated that 30% of the
adult population worldwide is physically inactive [5] and that
physical inactivity caused more than 5.3 million of the 57 mil-
lion deaths in 2008 [6]. However, a systematic review in the
USA found that around 50% of the physicians provided PA
counseling for the patients [7]. In Brazil, the prevalence of PA
counseling for the adult population is about 20% [8].
Limited data are available in Brazil on PA counseling and
leisure-time PA among hypertensive and diabetic subjects,
although PA importance for blood pressure [9] and glycemic
control [10,11] is well known. Consequently, the aim of this
study was to study PA counseling by FHS health providers
toward hypertensive and diabetic subjects in the State of Per-
nambuco, Brazil; and actual leisure-time PA practice in these
subjects.
2. Methods
The present study was performed using the database of the
SERVIDIAH study (Evaluation of Health SERVIces for DIAbetic
and Hypertensive Subjects) which was a cross-sectional epi-
demiological study conducted between November 2009 and
December 2010, in a representative sample of hypertensive
and diabetic subjects attended by the FHS in the State of Per-
nambuco. The general aim of the SERVIDIAH study was to
assess the implementation and actual results of the Brazilian
plan to improve primary care for hypertension and diabetes
within the FHS [12]. Following are summarized the main
methodological procedures for the present study. Further
details can be found in Fontbonne et al. [13].
2.1. Study sample
The study covered 35 municipalities, of which 16 had less
than 20,000 inhabitants (small-sized), 16 had between 20,000
and 100,000 inhabitants (medium-sized), and three had over
100,000 inhabitants (large-sized). From these 35 municipali-
ties, 208 FHS teams were randomly selected and within each
team a random sample of between three to six hypertensives
and three to six diabetics subjects, in order to ensure an even
balance of subjects by municipality size category, was drawn.
Subjects were considered hypertensive when they were reg-
istered by the CHA as having hypertension but not diabetes,
Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
and diabetic when registered as having type 2 diabetes, with
or without hypertension.
The final sample consisted of 785 hypertensive and 822 dia-
betic subjects aged over 20 years. The general response rate
was 86.7%.
The SERVIDIAH study was approved by the Ethics Commit-
tee of the Aggeu Magalhaes Research Centre, Oswaldo Cruz
Foundation (registration number 43/2008), and the Brazilian
National Commission of Ethics in Research under the registra-
tion number 889/2008. All respondents were informed about
the objectives and procedures of the study, and signed a con-
sent form.
2.2. Data collection
The data were collected through a structured question-
naire by face-to-face interviews, conducted by trained field
researchers. The participants were interviewed in their homes
or in a room in the FHS health unit.
PA counseling was assessed through the following ques-
tion: “Has any health provider of the Family Health Strategy ever
counseled you to modify (improve) your physical activity habits?
(Yes/No)”. Leisure-time PA was self-reported as “none”, “light”
(e.g., walking, bicycling or dancing for more than three hours
per week), “moderate” (e.g., running, practicing gymnastics or
another sport activity for more than three hours per week)
or “vigorous” (sports competition), and analyzed as a Boolean
variable (Yes = any leisure-time PA/No = no leisure-time PA).
Socio-demographic and lifestyle characteristics were self-
reported, and categorized as follows: age (20–59 years/60–74
years/75 and over); sex (Male/Female); formal schooling (Yes/No);
family income (Up to minimum wage/Over minimum wage;
monthly minimum wage at the time of the study was R$545,
US$310); tobacco consumption (Yes, present smoker/No, never or
ex-smoker); alcohol consumption (Yes/No); adhesion to salt-free
(Yes/No), sugar-free (Yes/No), and/or weight-loss diets (Yes/No).
Resting systolic and diastolic blood pressure (BP) was mea-
sured at three different times during the interview with the
subjects well-relaxed for all three measures by an Omron
®
digital wrist BP device. The mean of the three values was
used for analysis. In hypertensive subjects without diabetes,
BP was considered well-controlled when systolic BP was below
140 mmHg and diastolic BP below 90 mmHg; in diabetic sub-
jects, respective thresholds for good control were 130 and
80 mmHg [14].
For diabetic subjects, HbA1c levels were measured in a
capillary blood sample using the in2it point-of-care analyzer
(Bio-Rad Laboratories Inc., Berkeley, USA). Diabetes was con-
sidered well-controlled if HbA1c level was below 7% [15].
All study subjects were weighed (accurate to 0.1 kg) with
a Tanita BC553 electronic scale (Tanita Corp., Tokyo, Japan).
Height was measured with a portable stadiometer (Altur-
aexata, Belo Horizonte, Brazil). Body Mass Index (BMI) was
used to classify the subjects as overweight (between 25 and
29.9 kg/m2 ) or obese (30 kg/m2 or over) [16]. Waist circumfer-
ence (WC) was measured halfway between the lowest rib and
the upper ridge of the iliac crest; men with WC ≥ 102 cm and
women with WC ≥ 88 cm were classified as having high WC
[16].
PCD-601; No. of Pages 10
ARTICLE IN PRESS
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx 3

2.3. Statistical analyses

Table 1 – Descriptive characteristics of the hypertensive
and diabetic subjects attended by the Family Health
Analyses were conducted separately for hypertensive and dia- Strategy, Pernambuco State, Brazil, 2010.
betic subjects using the statistical software Statistical Package
Hypertensive Diabetic
for Social Sciences (SPSS), version 20 (SPSS Inc., Chicago, USA). subjects subjects
Descriptive statistics (frequency, mean and standard devi-
ation) were performed primarily. Then, to identify factors N % N %
independently associated to counseling and actual leisure- Socio-demographic characteristics
time PA, the data were analyzed by logistic regression with Size of municipality
adjustment for all variables. Large 299 38.1% 320 38.8%
Medium 281 35.8% 363 44.1%
Small 205 26.1% 140 17.1%
Age group
3. Results 20–59 354 45.1% 365 44.3%
60–74 297 37.8% 341 41.4%
The characteristics of the subjects are shown in Table 1. Mean 75 and over 134 17.1% 117 14.2%
age for hypertensive subjects was 61.0 ± 13.9 years, and for Sex
Male 224 28.5% 254 30.8%
diabetic subjects 61.1 ± 13.2 years. Around 70% in both sam-
Women 561 71.5% 569 69.2%
ples were female, 40% had no formal schooling, and more
Formal schooling
than 30% had family monthly income below the minimum No 301 38.3% 311 37.8%
wage. Although overweight or obesity was present in around Yes 484 61.7% 512 62.2%
two thirds of the subjects, only 31% of the hypertensive sub- Family monthly income
jects, and 29.4% of the diabetic subjects, had leisure-time PA ≤minimum wage 269 36.5% 246 30.7%
and only 16.2% of the hypertensive and 13.3% of the diabetic >minimum wage 468 63.5% 555 69.3%

subjects declared being on a weight-loss diet. Having had PA Characteristics of the disease
counseling was reported by 53.0% of the hypertensive and With hypertension
59.4% of the diabetic subjects. No – – 188 23.0%
Yes – – 629 77.0%
PA counseling and the actual practice of leisure-time PA
Well-controlled blood pressure
were related. Hypertensive subjects practicing leisure-time PA
No 446 56.9% 606 74.2%
had 61% more chance of having had PA counseling, compared Yes 338 43.1% 211 25.8%
to inactive subjects (95% CI: 1.18–2.21). The respective OR for Well-controlled HbA1c
diabetic subjects was 1.79 (95% CI: 1.30–2.48). No – – 548 69.5%
The unadjusted and adjusted analyses of association Yes – – 240 30.5%
between PA counseling and the independent variables among BMI
Normal 190 25.5% 197 36.5%
the hypertensive subjects are displayed in Table 2. Adjusted
Overweight 267 35.9% 294 39.4%
analyses showed that subjects on weight-loss diet had more
Obesity 287 38.6% 255 34.1%
chance of having received counseling for PA. Conversely, older High waist circumference
subjects both 60–74 and over 74 had less chance of having No 259 34.7% 259 31.4%
received counseling by a health provider in the FHS. Other Yes 488 65.3% 565 68.6%
variables were associated with more frequent PA counseling, Lifestyle
but only in unadjusted analyses: being a woman, having had Physical activity counseling
formal schooling, having well-controlled blood pressure, and No 360 47.0% 324 40.6%
being obese. Yes 406 53.0% 475 59.4%
Table 3 presents the crude and adjusted analyses for PA Leisure-time physical activity
No 539 69.0% 574 70.6%
counseling among diabetic subjects. The diabetic subjects who
Yes 242 31.0% 239 29.4%
had more chance to be counseled were: women, subjects with
Tobacco consumption
formal schooling, with associated hypertension and obese. No 681 86.8% 716 87.0%
Lifestyle characteristics were not associated to PA counseling. Yes 104 13.2% 107 13.0%
Being on a weight-loss diet or having high WC was associated Alcohol consumption
to PA counseling only in the unadjusted analyses. No 635 80.9% 705 85.6%
Table 4 presents the relations to leisure-time PA as depen- Yes 150 19.1% 118 14.4%
On weight-loss diet
dent variable among the hypertensive subjects. For adjusted
No 657 83.8% 705 85.7%
analyses, the variables that maintained association to leisure- Yes 127 16.2% 111 13.5%
time PA were: municipality size (medium-sized: OR: 0.62), sex Adhering to salt-free diet
(female: OR: 0.50), age (60–74 years: OR: 2.15) and being on a No 88 11.4% – –
weight-loss diet (OR: 2.69). No variable related to the clinical Yes 686 88.6% – –
characteristics of hypertension was associated with leisure- Adhering to sugar-free diet
No – – 104 12.7%
time PA, even in the unadjusted analyses.
Yes – – 716 87.3%
Table 5 shows associations with actual leisure-time PA
practice among diabetic subjects. In adjusted analyses, sub-
jects living in a medium sized municipality and women had

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
4 p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx

Table 2 – Crude and adjusted odds ratios for physical activity counseling by health providers to hypertensive subjects
attended by the Family Health Strategy, Pernambuco State, Brazil, 2010.
PA counseling Crude OR (95%CI) Adjusted OR (95%CI)
N (%)
Socio-demographic characteristics
Size of municipality
Large 164 (57.7%) 1.00 1.00
Medium 145 (52.0%) 0.79 (0.57–1.10) 1.02 (0.69–1.52)
Small 97 (47.8%) 0.67* (0.47–0.96) 0.87 (0.56–1.37)
Age group
20–59 212 (60.7%) 1.00 1.00
60–74 149 (51.6%) 0.69* (0.50–0.94) 0.64* (0.44–0.94)
75 and over 45 (35.2%) 0.35** (0.23–0.53) 0.44** (0.25–0.78)
Sex
Male 93 (42.3%) 1.00 1.00
Women 313 (57.3%) 1.83** (1.34–2.52) 1.16 (0.75–1.77)
Formal schooling
No 135 (45.8%) 1.00 1.00
Yes 271 (57.5%) 1.61** (1.20–2.15) 1.24 (0.84–1.81)
Family monthly income
≤minimum wage 141 (53.4%) 1.00 1.00
>minimum wage 238 (52.0%) 0.94 (0.70–1.28) 0.99 (0.70–1.41)

Characteristics of the disease

Duration (years) 10.3 ± 9.3 1.01 (0.99–1.03) 1.02 (1.00–1.04)
Well-controlled blood pressure
No 217 (49.5%) 1.00 1.00
Yes 189 (57.8%) 1.40* (1.05–1.86) 1.17 (0.83–1.65)
BMI
Normal 91 (48.7%) 1.00 1.00
Overweight 129 (49.6%) 1.04 (0.71–1.51) 0.67 (0.41–1.08)
Obesity 170 (60.7%) 1.63* (1.12–2.37) 0.70 (0.39–1.25)
High waist circumference
No 114 (45.4%) 1.00 1.00
Yes 280 (58.6%) 1.70** (1.25–2.31) 1.54 (0.93–2.54)

Lifestyle
Tobacco consumption
No 357 (53.8%) 1.00 1.00
Yes 49 (48.0%) 0.79 (0.52–1.21) 0.96 (0.58–1.58)
Alcohol consumption
No 326 (52.5%) 1.00 1.00
Yes 80 (55.2%) 1.11 (0.78–1.60) 0.79 (0.50–1.25)
On weight-loss diet
No 324 (50.6%) 1.00 1.00
Yes 82 (65.6%) 1.86** (1.25–2.78) 1.90* (1.16–3.11)
Adhering to salt-free diet
No 37 (43.0%) 1.00 1.00
Yes 363 (54.3%) 1.57 (0.99–2.47) 1.46 (0.86–2.48)
∗
p < 0.05.
∗∗
p < 0.01.

less chance to practice leisure-time PA, while being between
60 to 74 years old or consuming alcohol, was associated with
more chance to practice leisure-time PA. Unadjusted analyses
showed also a higher frequency of leisure-time PA practice in
subjects who had formal schooling or who were not abdomi-
nally obese.
4. Discussion
The socio-economic profile of our study samples, with low
family income and almost half with no formal school-
ing, reflects the FHS policy implementation which initially
focussed on poorer than average regions for reducing
inequities in access and utilization of care [1,2].
Having received PA counseling was reported by around
half of the subjects with hypertension or diabetes. This esti-
mate is close to others found in cross-sectional studies carried
out in Brazil [8,17,18]. Only one of them [17] had a repre-
sentative sample of subjects attended by the FHS. It was
conducted in cities with more than 100,000 habitants from the
South and the North-East of Brazil, and included adults (30–64
years) and elderly (65 years or more) subjects. Close to 50%
of the subjects with hypertension or diabetes reported having
received PA counseling, which was higher than in the gen-
eral sample (28.9% for the adults and 38.9% in the elderly).

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx 5

Table 3 – Crude and adjusted odds ratios for physical activity counseling by health providers to diabetic subjects attended
by the Family Health Strategy, Pernambuco State, Brazil, 2010.
PA counseling Crude OR (95%CI) Adjusted OR (95%CI)
N (%)
Socio-demographic characteristics
Size of municipality
Large 199 (65.5%) 1.00 1.00
Medium 192 (54.1%) 0.62* (0.45–0.85) 0.72 (0.48–1.07)
Small 84 (60.0%) 0.79 (0.52–1.20) 1.01 (0.60–1.70)
Age group
20–59 223 (61.9%) 1.00 1.00
60–74 197 (59.2%) 0.89 (0.66–1.21) 0.83 (0.57–1.21)
75 and over 55 (51.9%) 0.67 (0.43–1.03) 0.60 (0.34–1.06)
Sex
Male 123 (51.2%) 1.00 1.00
Women 352 (63.0%) 1.62* (1.20–2.20) 1.74** (1.16–2.65)
Formal schooling
No 153 (50.5%) 1.00 1.00
Yes 322 (64.9%) 1.81* (1.36–2.43) 1.53* (1.05–2.23)
Family monthly income
≤minimum wage 137 (56.6%) 1.00 1.00
>minimum wage 324 (60.6%) 1.18 (0.87–1.61) 1.19 (0.83–1.73)

Characteristics of the disease

Duration (years) 8.7 ± 7.5 1.01 (0.99–1.03) 1.02 (0.99–1.05)
Well-controlled HbA1c
No 315 (59.5%) 1.00 1.00
Yes 137 (58.3%) 0.95 (0.69–1.98) 0.92 (0.64–1.35)
With hypertension
No 83 (44.9%) 1.00 1.00
Yes 389 (63.9%) 2.19* (1.57–3.06) 2.12** (1.40–3.22)
Well-controlled blood pressure
No 359 (60.6%) 1.00 1.00
Yes 112 (55.7%) 0.81 (0.59–1.12) 0.95 (0.62–1.44)
BMI
Normal 98 (50.5%) 1.00 1.00
Overweight 178 (61.4%) 1.55* (1.08–2.24) 1.50 (0.93–2.43)
Obesity 170 (60.7%) 2.19* (1.49–3.24) 2.26** (1.24–4.12)
High waist circumference
No 137 (53.9%) 1.00 1.00
Yes 338 (62.0%) 1.39* (1.03–1.88) 0.81 (0.49–1.35)

Lifestyle
Tobacco consumption
No 421 (60.8%) 1.00 1.00
Yes 53 (50.5%) 0.67 (0.44–1.01) 0.77 (0.46–1.30)
Alcohol consumption
No 399 (58.4%) 1.00 1.00
Yes 76 (65.5%) 1.34 (0.88–2.02) 1.47 (0.87–2.48)
On weight-loss diet
No 393 (57.4%) 1.00 1.00
Yes 79 (71.8%) 1.93* (1.24–3.00) 1.23 (0.72–2.13)
Adhering to sugar-free diet
No 57 (55.9%) 1.00 1.00
Yes 416 (59.9%) 1.19 (0.78–1.80) 1.47 (0.87–2.49)
∗
p < 0.05.
∗∗
p < 0.01.

Similarly, a study on a nationally representative sample of
Brazilian adults (20–59 years), carried out from August 2008
to March 2009, identified a higher prevalence of PA counsel-
ing among subjects reporting to be with hypertension (41.9%)
or diabetes (55.6%) than in the general sample (20%) [8]. In
Pelotas, a city from the South of Brazil, in a representative
sample of the adult population (20–69 years), 56.2% declared
having received PA counseling [18]; there were no separate
estimates for hypertensive or diabetic subjects. By contrast
to these studies, frequency of PA counseling was much higher
in the National Health Survey performed from August 2013 to
February 2014 by the Ministry of Health in partnership with
the Brazilian Institute of Geography and Statistics: it showed
that 80.0% of the hypertensive and 84.4% of the diabetic sub-
jects had received PA counseling in primary care [19]. It could

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
6 p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx

Table 4 – Crude and adjusted odds ratios for leisure-time physical activity practice by hypertensive subjects attended by
the Family Health Strategy, Pernambuco State, Brazil, 2010.
Leisure-time PA Crude OR (95%CI) Adjusted OR (95%CI)
N (%)
Socio-demographic characteristics
Size of municipality
Large 108 (36.2%) 1.00 1.00
Medium 75 (27.0%) 0.65* (0.46–0.93) 0.62* (0.40–0.94)
Small 59 (28.8%) 0.71 (0.48–1.04) 0.89 (0.55–1.42)
Age group
20–59 94 (26.6%) 1.00 1.00
60–74 120 (40.7%) 1.90** (1.36–2.64) 2.15** (1.43–3.25)
75 and over 28 (21.2%) 0.75 (0.46–1.20) 1.34 (0.73–2.48)
Sex
Male 84 (37.8%) 1.00 1.00
Women 158 (28.3%) 0.65** (0.47–0.90) 0.50** (0.32–0.79)
Formal schooling
No 84 (28.3%) 1.00 1.00
Yes 158 (32.6%) 1.23 (0.90–1.69) 1.25 (0.82–1.89)
Family monthly income
≤minimum wage 74 (27.6%) 1.00 1.00
>minimum wage 150 (32.2%) 1.24 (0.89–1.73) 1.09 (0.74–1.60)

Characteristics of the disease

Duration (years) 10.1 ± 9.1 1.01 (0.99–1.02) 1.00 (0.98–1.02)
Well-controlled blood pressure
No 135 (30.4%) 1.00 1.00
Yes 107 (31.8%) 1.07 (0.79–1.45) 1.32 (0.91–1.90)
BMI
Normal 60 (31.7%) 1.00 1.00
Overweight 85 (32.1%) 1.02 (0.68–1.52) 0.70 (0.42–1.15)
Obesity 93 (32.5%) 1.04 (0.70–1.54) 0.60 (0.33–1.11)
High waist circumference
No 84 (32.7%) 1.00 1.00
Yes 153 (31.5%) 0.95 (0.69–1.31) 1.24 (0.72–2.12)

Lifestyle
Tobacco consumption
No 212 (31.2%) 1.00 1.00
Yes 30 (29.4%) 0.92 (0.58–1.45) 0.71 (0.41–1.24)
Alcohol consumption
No 188 (29.7%) 1.00 1.00
Yes 54 (36.2%) 1.34 (0.92–1.95) 1.09 (0.67–1.75)
On weight-loss diet
No 183 (28.0%) 1.00 1.00
Yes 58 (46.0%) 2.20** (1.49–3.24) 2.69** (1.68–4.33)
Adhering to salt-free diet
No 21 (24.1%) 1.00 1.00
Yes 216 (31.6%) 1.45 (0.67–2.44) 1.76 (0.94–3.29)
∗
p < 0.05.
∗∗
p < 0.01.

suggest that the practice of PA counseling by health providers
is growing in Brazil.
Outside Brazil, in developed countries, estimates appear
to be a little higher than in our study. The 2006 US National
Health Interview Survey Outcomes found that PA counsel-
ing for subjects with diabetes or prediabetes, aged 40 years
or older, was 65.5% for women and 62.9% for men [20]. It is
important to highlight that the analysis was performed only in
overweight or obese subjects; and our study showed a higher
frequency of counseling for subjects with overweight and obe-
sity, close to what was found in the US study. In Canada,
the 2009 Survey of Living with Chronic Disease, carried out
through telephone interviews, had a component of analyses
considering the subjects (aged 20 or older) who reported hav-
ing been diagnosed with hypertension by a health provider
[21]. In this subgroup, the frequency of PA counseling was 70%.
Factors that influenced the probability of having received
PA counseling were relatively similar between hypertensive
and diabetic subjects, although some statistical differences
could arise in multivariate analyses. An important factor influ-
encing PA counseling was the existence of overweight or
obesity. Although the relations could be statistically different
between the models and the study groups (hypertensive or
diabetic), markers of overweight/obesity such as BMI, or the
fact of being on a weight-loss diet, increased the probability
of having received PA counseling. This probably reflects the

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx 7

Table 5 – Crude and adjusted odds ratios for leisure-time physical activity practice by diabetic subjects attended by the
Family Health Strategy, Pernambuco State, Brazil, 2010.
Leisure-time PA Crude OR (95%CI) Adjusted OR (95%CI)
N (%)
Socio-demographic characteristics
Size of municipality
Large 113 (36.1%) 1.00 1.00
Medium 86 (24.0%) 0.60** (0.40–0.78) 0.63* (0.42–0.94)
Small 39 (28.1%) 0.69 (0.45–1.07) 0.65 (0.39–1.11)
Age group
20–59 105 (29.1%) 1.00 1.00
60–74 114 (34.0%) 1.26 (0.91–1.73) 1.69** (1.16–2.47)
75 and over 20 (17.2%) 0.51* (0.30–0.86) 0.72 (0.38–1.36)
Sex
Male 95 (37.5%) 1.00 1.00
Women 144 (25.8%) 0.58** (0.42–0.80) 0.59* (0.39–0.91)
Formal schooling
No 69 (22.6%) 1.00 1.00
Yes 170 (33.5%) 1.73** (1.25–2.40) 1.19 (0.80–1.77)
Family monthly income
≤minimum wage 71 (29.3%) 1.00 1.00
>minimum wage 162 (29.5%) 1.02 (0.73–1.42) 1.09 (0.74–1.59)

Characteristics of the disease

Duration (years) 8.7 ± 7.5 0.97** (0.95–0.99) 0.98 (0.95–1.00)
Well-controlled HbA1c
No 152 (28.3%) 1.00 1.00
Yes 77 (32.1%) 1.20 (0.85–1.66) 0.99 (0.68–1.45)
With hypertension
No 57 (30.5%) 1.00 1.00
Yes 181 (29.1%) 0.95 (0.66–1.35) 1.06 (0.68–1.66)
Well-controlled blood pressure
No 176 (29.4%) 1.00 1.00
Yes 62 (29.8%) 1.03 (0.73–1.45) 1.16 (0.76–1.76)
BMI
Normal 58 (30.1%) 1.00 1.00
Overweight 95 (32.6%) 1.14 (0.77–1.70) 1.08 (0.65–1.80)
Obesity 81 (32.1%) 1.11 (0.74–1.67) 0.88 (0.47–1.65)
High waist circumference
No 90 (35.2%) 1.00 1.00
Yes 149 (26.8%) 0.67* (0.48–0.92) 1.04 (0.61–1.76)

Lifestyle
Tobacco consumption
No 209 (29.6%) 1.00 1.00
Yes 30 (28.3%) 0.92 (0.59–1.45) 0.78 (0.45–1.35)
Alcohol consumption
No 188 (27.1%) 1.00 1.00
Yes 51 (43.6%) 2.09* (1.40–3.12) 1.86* (1.13–3.06)
On weight-loss diet
No 193 (27.7%) 1.00 1.00
Yes 44 (40.0%) 1.75** (1.16–2.65) 1.64 (1.00–2.66)
Adhering to sugar-free diet
No 32 (30.8%) 1.00 1.00
Yes 206 (29.2%) 0.91 (0.58–1.42) 0.85 (0.50–1.44)
∗
p < 0.05.
∗∗
p < 0.01.

knowledge by FHS healthcare providers that behavioral coun-
seling on diet and PA has benefits and the greater benefits
appear in the highest risk group for cardiovascular disease
(CVD) [25]. In line with this explanation is our observation that
PA counseling was more frequent in diabetic subjects when
they had associated hypertension, and that, in other studies,
individuals with CVD risk factors were more likely to be coun-
seled [8,23]. However, it has to be reminded that all subjects
with hypertension and/or diabetes do benefit from physical
activity [14,26], and therefore that PA counseling should not
be restricted to those with overweight/obesity or other CVD
risk factors.
PA counseling was less likely to be directed to older sub-
jects, and even less when they were 75 or older, than to adults
below 60 years of age. One explanation could be the lower
health provider confidence in behavior changes by older peo-

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
8 p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx
ple [22,23]. However, other studies have found contrary trends.
In the already-quoted study by Duro et al. [8], an increased lin-
ear trend of counseling according to age was observed, but the
study sample was limited to adults between 20 and 59 years
of age. In other studies carried out in Brazil and in the USA,
older subjects were more likely to get counseling by health
providers than adults [17,23]. The discrepancies between these
studies and our results suggest that more investigations are
needed to improve knowledge about the relation between PA
counseling and age. Other socio-demographic characteristics
influencing PA counseling were sex and education. Women
were more likely to receive PA counseling than men, a finding
already observed in the general population by other studies
[8,17,18,23]. PA counseling was also more frequently directed
to subjects who had formal schooling, possibly indicating
another provider bias, similar to the bias regarding older age,
namely that subjects with no formal schooling would have dif-
ficulties in understanding advice and/or putting it into practice
[24]. Other studies have also shown that PA counseling was
less frequently reported by subjects with lower income-level
or lower formal education level [8,17,18,23].
In our study, actual practice of leisure-time PA was reported
by around one third of the subjects in both the hyperten-
sive and the diabetic groups. This is a little higher than the
23% frequency of leisure-time PA practice found in the Cana-
dian Community Health Survey, in elderly subjects with one
or more chronic diseases [27]. However, in Brazil, the VIGI-
TEL study, an annual population-based inquiry by telephone,
showed that in 2010, 12.7% of adults reported having at least
150 min per week of leisure-time PA [28]. Our much larger
estimate can be attributed to the less stringent definition of
leisure-time PA in our study, as the question about practice of
physical activity did not specify frequency and/or duration of
each bout of PA. There is also the possibility of recall bias, as
PA was all self-reported.
Subjects who declared having leisure-time PA were more
likely to have received PA counseling. However, as the design of
our study was cross-sectional, it is impossible to infer causal-
ity: it may be that PA counseling was an incentive for actual
practice, but the relation may also denote a bias on the part of
health providers, who may be more prone to counsel already
physically-active subjects.
Factors related to leisure-time PA were almost the same in
hypertensive and diabetic subjects. The age-group from 60 to
74 years old was more likely to be physically active than adults
or people 75 years or older. This could indicate a beneficial
effect of increased leisure-time at retirement, at least in the
first “younger” years of it. However, other studies have gener-
ally shown a decrease in leisure-time PA with the increase of
age [28,29]; and as to retirement itself, one study has shown
that it may slightly increase non-sport leisure-time PA, but on
the other hand it considerably decreased work-related PA [30].
Women were less active during leisure-time than men, a find-
ing reported by other studies [5,27–29,31]. However, it has to
be kept in mind that, because of cultural aspects, men tend
to be more active in their leisure-time while women could be
more active in the household [32–34].
Few other characteristics were related with leisure-time PA
practice. Prominently, there were no relations with markers of
overweight/obesity, except the fact of being on a weight-loss
Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
diet. As being on a weight-loss diet was acknowledged by only
one-sixth of our study sample, while two-thirds were over-
weight or obese, it may be hypothesized that the association
identified particularly motivated individuals, who combined
diet to PA in order to lose weight. An association was also
observed between alcohol consumption and leisure-time PA,
particularly strong in diabetic subjects, and with no clear
explanation; it has to be kept in mind, though, that the variable
was defined in a “yes/no” way, gathering in the “no” category
both abstainers and ex-drinkers. The latter may suffer from
ill-health and consequently not engage in physical activity.
A last factor which influenced both PA counseling and
actual leisure-time practice was the size of the municipal-
ity. The tendency was for large-sized municipalities to have
better estimates, on both accounts. Better practice could be
related to environmental characteristics such as availabil-
ity of facilities for PA practice [31], or simply better-suited
and better-kept physical environment—sidewalks, tree cover,
disciplined road traffic, etc. [35]. Better counseling may be
another example of globally better care in large-sized munic-
ipalities, which was already observed in the SERVIDIAH Study
[13,36], and attributed to possible shortcomings in the decen-
tralization process which characterizes the Brazilian health
system, where large-sized municipalities appear more capable
of attracting both financial and qualified human resources.
5. Conclusions
In this study of two large representative samples, with a high
response rate, of hypertensive and diabetic subjects attended
by the FHS in the state of Pernambuco, North-East Brazil, there
is evidence of insufficient practice of leisure-time PA, showing
the need to increase PA counseling by health providers at pri-
mary healthcare level. This counseling should avoid biases,
such as disregarding older or poorly-educated subjects, and
should not be restricted to subjects with overweight-obesity or
other cardiovascular risk factors; all subjects diagnosed with
hypertension or diabetes would be benefited by PA counsel-
ing. Efforts should be made to convey more equity between
municipalities, regardless of their size.
Conflicts of interest
The authors state that they have no conflict of interest.
Funding
The SERVIDIAH study was supported by the National Coun-
cil of Technological and Scientific Development (CNPq), Brazil
[Grant numbers: 403640/2008-3; 490855/2008-3; 576677/2008-
6]; by the Foundation for Science and Technology of the State
of Pernambuco (Facepe), Brazil [Grant number: APQ 1378-
4.00/08]; and by the Institute of Research for Development
(IRD), France.
Acknowledgments
We would like to thank the patients and the providers of the
Family Health Strategy for their participation, and the dedica-
tion of our staff members in the SERVIDIAH study.
PCD-601; No. of Pages 10
ARTICLE IN PRESS
p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx 9

references Brasileiras de Hipertensão, Arq. Bras. Cardiol. 95 (2010) 1–51,

http://dx.doi.org/10.1590/S0066-782X2010001700001.
[15] ADA, Standards of medical care in diabetes, J. Clin. Appl.
Res. Educ. 36 (2016), http://dx.doi.org/10.2337/dc13-S011.
[1] J. Macinko, M.B.M. Harris, Brazil’s family health
[16] WHO, Report of a WHO Consultation on Obesity: Preventing
strategy—delivering community-based primary care in a
and Managing the Global Epidemic, World Health
universal health system, N. Engl. J. Med. 372 (2015)
Organization, Geneva, 1998.
2177–2181, http://dx.doi.org/10.1056/NEJMp1002530.
[17] F.V. Siqueira, M.V. Nahas, L.A. Fachini, D.S. Silveira, R.X.
[2] J. Paim, C. Travassos, C. Almeida, L. Bahia, J. Macinko, The
Piccini, E. Tomasi, E. Thumé, P.C. Hallal, Counseling for
Brazilian health system: history, advances, and challenges,
physical activity as a health education strategy, Cad. Saude
Lancet 377 (2011) 1778–1797,
Publica 25 (2009) 203–213,
http://dx.doi.org/10.1016/S0140-6736(11)60054-8.
http://dx.doi.org/10.1590/S0102-311X2009000100022.
[3] H. Tulloch, M. Fortier, W. Hogg, Physical activity counseling
[18] P.C. Hallal, P.T. Machado, G.F. Del Duca, I.C. Silva, T.C.
in primary care: who has and who should be counseling?
Amorim, T.T. Borges, A.J. Rombaldi, M.R. Azevedo, A.G.
Patient Educ. Couns. 64 (2006) 6–20,
Knuth, Physical activity advice: short report from a
http://dx.doi.org/10.1016/j.pec.2005.10.010.
population-based study in Brazil, J. Phys. Act. Health 7 (2010)
[4] A.R. Gagliardi, G. Faulkner, D. Ciliska, A. Hicks, Factors
352–354.
contributing to the effectiveness of physical activity
[19] C.L. Szwarcwald, P.R.B. de Souza Júnior, G.N. Damacena, W.
counselling in primary care: a realist systematic review,
da Silva de Almeida, D.C. Malta, S.R. Stopa, M.L.F.P. Vieira,
Patient Educ. Couns. 98 (2015) 412–419,
C.A. Pereira, Recomendações e práticas dos
http://dx.doi.org/10.1016/j.pec.2014.11.020.
comportamentos saudáveis entre indivíduos com
[5] P.C. Hallal, L.B. Andersen, F.C. Bull, R. Guthold, W. Haskell, U.
diagnóstico de hipertensão arterial e diabetes no Brasil:
Ekelund, Global physical activity levels: surveillance
Pesquisa Nacional de Saúde (PNS), 2013, Rev. Bras.
progress, pitfalls, and prospects, Lancet 380 (2012) 247–257,
Epidemiol. (2015) 132–145,
http://dx.doi.org/10.1016/S0140-6736(12)60646-1.
http://dx.doi.org/10.1590/1980-5497201500060012.
[6] I.-M. Lee, E.J. Shiroma, F. Lobelo, P. Puska, S.N. Blair, P.T.
[20] R. Dorsey, T. Songer, Lifestyle behaviors and physician
Katzmarzyk, Effect of physical inactivity on major
advice for change among overweight and obese adults with
non-communicable diseases worldwide: an analysis of
prediabetes and diabetes in the United States, 2006, Prev.
burden of disease and life expectancy, Lancet 380 (2012)
Chronic Dis. 8 (2011) A132 http://www.pubmedcentral
219–229, http://dx.doi.org/10.1016/S0140-6736(12)61031-9.
.nih.gov/articlerender.fcgi?artid=3221573&tool=pmcentrez
[7] J.J. VanWormer, N.P. Pronk, G.J. Kroeninger, Clinical
&rendertype=abstract.
counseling for physical activity: translation of a systematic
[21] R.L. Walker, M.E. Gee, C. Bancej, R.P. Nolan, J. Kaczorowski,
review into care recommendations, Diabetes Spectr. 22
M. Joffres, A. Bienek, F. Gwadry-Sridhar, N.R.C. Campbell,
(2009) 48–55, http://dx.doi.org/10.2337/diaspect.22.1.48.
Health behaviour advice from health professionals to
[8] S.M.S. Duro, E. Tomasi, F.V. Siqueira, D.S. Silveira, E. Thume,
Canadian adults with hypertension: results from a national
L.A. Facchini, Adult physical activity counseling by health
survey, Can. J. Cardiol. 27 (2011) 446–454,
professionals in Brazil: a national urban population survey, J.
http://dx.doi.org/10.1016/j.cjca.2011.03.013.
Phys. Act. Health 12 (2015) 1177–1183,
[22] S. McPhail, M. Schippers, An evolving perspective on
http://dx.doi.org/10.1123/jpah.2013-0213.
physical activity counselling by medical professionals, BMC
[9] R.E. Brown, M.C. Riddell, A.K. Macpherson, K.L. Canning, J.L.
Fam. Pract. 13 (2012) 31,
Kuk, The joint association of physical activity,
http://dx.doi.org/10.1186/1471-2296-13-31.
blood-pressure control, and pharmacologic treatment of
[23] J. Ma, G.G. Urizar, T. Alehegn, R.S. Stafford, Diet and physical
hypertension for all-cause mortality risk, Am. J. Hypertens.
activity counseling during ambulatory care visits in the
26 (2013) 1005–1010, http://dx.doi.org/10.1093/ajh/hpt063.
United States, Prev. Med. 39 (2004) 815–822,
[10] A.P. Carson, L.B. Williams, A.N. Hill, Physical activity in
http://dx.doi.org/10.1016/j.ypmed.2004.03.006.
diabetes: is any better than none? J. Diabetes Complications
[24] S. Willems, S. De Maesschalck, M. Deveugele, A. Derese, J. De
28 (2014) 257–258,
Maeseneer, Socio-economic status of the patient and
http://dx.doi.org/10.1016/j.jdiacomp.2013.10.012.
doctor-patient communication: does it make a difference?
[11] S. Kaizu, H. Kishimoto, M. Iwase, H. Fujii, T. Ohkuma, H. Ide,
Patient Educ. Couns. 56 (2005) 139–146,
T. Jodai, Y. Kikuchi, Y. Idewaki, Y. Hirakawa, U. Nakamura, T.
http://dx.doi.org/10.1016/j.pec.2004.02.011.
Kitazono, Impact of leisure-time physical activity on
[25] D. Campo-Outcalt, Diet, exercise, and CVD: when counseling
glycemic control and cardiovascular risk factors in Japanese
makes the most sense, J. Fam. Pract. 63 (2014) 458–460
patients with type 2 diabetes mellitus: the Fukuoka Diabetes
http://ovidsp.ovid.com/ovidweb.cgi?
Registry, PLoS One 9 (2014),
T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=ovftp
http://dx.doi.org/10.1371/journal.pone.0098768.
&AN=00004790-201408000-00006.
[12] Brazilian National Strategy for the Reorganization of Care
[26] Sociedade Brasileira de Diabetes, Diretrizes da Sociedade
for Arterial Hipertension and Diabetes Mellitus, Plano de
Brasileira de Diabetes: 2013–2014, AC Farmaceûtica, São
Reorganização da Atenção à Hipertensão Arterial e ao
Paulo (2014).
Diabetes Mellitus, Rev. Saude Publica 35 (2001) 584–588,
[27] M. Ashe, W. Miller, J. Eng, L. Noreau, Older adults, chronic
http://dx.doi.org/10.1590/S0034-89102001000600014.
disease, and leisure-time physical activity, Gerontology 55
[13] A. Fontbonne, E.Â.P. Cesse, I.M.C. Sousa, W.V. Souza, A.F.B.
(2009) 64–72, http://dx.doi.org/10.1159/000141518.
Bezerra, E.F. Carvalho, Risk factor control in hypertensive
[28] G.I. Mielke, P.C. Hallal, D.C. Malta, I.-M. Lee, Time trends of
and diabetic subjects attended by the Family Health Strategy
physical activity and television viewing time in Brazil:
in the State of Pernambuco, Brazil: the SERVIDIAH study,
2006–2012, Int. J. Behav. Nutr. Phys. Act. 11 (2014) 101,
Cad. Saude Publica 29 (2013) 1195–1204,
http://dx.doi.org/10.1186/s12966-014-0101-4.
http://dx.doi.org/10.1590/S0102-311X2013000600016.
[29] G.F. Del Duca, M.V. Nahas, L.M.T. Garcia, J. Mota, P.C. Hallal,
[14] Sociedade Brasileira de Cardiologia, Sociedade Brasileira de
M.A. Peres, Prevalence and sociodemographic correlates of
Hipertensão, Sociedade Brasileira de Nefrologia, VI Diretrizes
all domains of physical activity in Brazilian adults, Prev.

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001
PCD-601; No. of Pages 10
ARTICLE IN PRESS
10 p r i m a r y c a r e d i a b e t e s x x x ( 2 0 1 7 ) xxx–xxx

Med. 56 (2013) 99–102,
http://dx.doi.org/10.1016/j.ypmed.2012.11.007.
[30] A.S. Slingerland, F.J. Van Lenthe, J.W. Jukema, C.B.M.
Kamphuis, C. Looman, K. Giskes, M. Huisman, K.M.V.
Narayan, J.P. Mackenbach, J. Brug, Aging, retirement, and
changes in physical activity: prospective cohort findings
from the GLOBE study, Am. J. Epidemiol. 165 (2007)
1356–1363, http://dx.doi.org/10.1093/aje/kwm053.
[31] C.S. Gomes, F.P. Matozinhos, L.L. Mendes, M.C. Pessoa, G.
Velasquez-Melendez, Physical and social environment are
associated to leisure time physical activity in adults of a
Brazilian city: a cross-sectional study, PLoS One 11 (2016)
e0150017, http://dx.doi.org/10.1371/journal.pone.0150017.
[32] B.J. Ayotte, J.A. Margrett, J. Hicks-Patrick, Physical activity in
middle-aged and young-old adults: the roles of self-efficacy,
barriers, outcome expectancies, self-regulatory behaviors
and social support, J. Health Psychol. 15 (2010) 173–185,
http://dx.doi.org/10.1177/1359105309342283.
[33] A.J. Oliveira, C.S. Lopes, M. Rostila, G.L. Werneck, R.H. Griep,
A.C.M.P. De Leon, E. Faerstein, Gender differences in social
support and leisure-time physical activity, Rev. Saude
Publica 48 (2014) 602–612,
http://dx.doi.org/10.1590/S0034-8910.2014048005183.
[34] R. Salles-Costa, M.L. Heilborn, E. Faerstein, C.S. Lopes,
Gender and leisure-time physical activity, Cad. Saude
Publica 19 (2003) 325–333,
http://dx.doi.org/10.1590/S0102-311X2003000800014.
[35] R.C. Brownson, C. Hoehner, K. Day, A. Forsyth, J. Sallis,
Measuring the built environment for physical activity: state
of the science, Am. J. Prev. Med. 36 (2009) 53,
http://dx.doi.org/10.1016/j.amepre.2009.01.005.
[36] M.N.S. de C. Barreto, E.Â.P. Cesse, R.F. Lima, M.G. da S.
Marinho, Y. da S. Specht, E.M.F. de Carvalho, A. Fontbonne,
Analysis of access to hypertensive and diabetic drugs in the
Family Health Strategy, State of Pernambuco, Brazil, Rev.
Bras. Epidemiol. 18 (2015) 413–424,
http://dx.doi.org/10.1590/1980-5497201500020010.

Please cite this article in press as: J.M.V. Barbosa, et al., Correlates of physical activity counseling by health providers to patients
with diabetes and hypertension attended by the Family Health Strategy in the state of Pernambuco, Brazil, Prim. Care Diab. (2017),
http://dx.doi.org/10.1016/j.pcd.2017.04.001