Barriers to physical activity among adults in primary healthcare units in the National Health System: a cross-sectional study in Brazil

ABSTRACT BACKGROUND: Many factors may negatively impact physical activity (PA), but studies lack evidence of individual predictors of perceived barriers to PA among adults in primary healthcare units. OBJECTIVE: To analyze associations between sociodemographic characteristics, health conditions, leisure-time physical activity (LTPA), PA counseling and perceived barriers to LTPA among adult patients in primary healthcare units of the National Health System in Brazil. DESIGN AND SETTING: Cross-sectional study on a representative sample of adults in primary healthcare units in São José dos Pinhais, Paraná, Brazil. METHODS: This study was conducted in 2019, among 779 adults (70% women). Barriers to LTPA, sociodemographic characteristics (sex, age, marital status, skin color, education and income), health conditions (body mass index, hypertension, diabetes, dyslipidemia, coronary disease and medications), LTPA level and PA counseling received were measured using validated, standardized procedures. The data were analyzed using chi-square and Mann-Whitney U tests. RESULTS: The most prevalent barriers were “feeling too tired” (53%) and “lack of time” (52%). PA counseling was inversely associated with “lack of time” (45% versus 57%; P < 0.001) but positively associated with “injury or disease” (38% versus 29%; P = 0.008). There was an inverse linear trend between the number of barriers and LTPA (walking and total) (P < 0.001). Most barriers differed in comparisons of sociodemographic characteristics, health conditions, LTPA and counseling (P < 0.05). CONCLUSIONS: The barriers vary according to the individual predictors. Counseling strategies need to be specific for each barrier and may be promising for promoting LTPA within primary healthcare.


INTRODUCTION
Around 28% of the worldwide adult population is not physically active, with higher prevalence among women (32%) and Latin American populations (39%), especially in Brazil (47%). 1 Therefore, physical inactivity has been considered to be a pandemic, with significant impact on public health and elevated economic burden with regard to healthcare, lack of productivity and premature mortality rates due to chronic disease. 2 In Brazil, the National Policy on Health Promotion is an important strategy for confronting physical inactivity that guides physical activity promotion at the community level through primary healthcare. 3 There is strong evidence showing the protective effects of regular physical activity against major chronic diseases and for improvement of quality of life. 2,4 Leisure-time physical activity (LTPA) has higher potential for modification than physical activity at work, for example. Additionally, it is affected by psychological, biological, social and behavioral processes that operate at individual, group and social levels. 5 Many factors can affect choices and opportunities for being physically active during leisure time, 6,7 which demonstrates that LTPA is influenced by factors that go beyond motivation or the population's knowledge of its health benefits. 2 Therefore, the perception of barriers to LTPA needs to be considered with regard to the success of programs to promote physical activity, 8 since those barriers can differ according to sex, age, health conditions, social and cultural characteristics, types of LTPA and lifestyle. 9,10 The barriers most commonly reported among the Brazilian adult population are intrapersonal and related to lack of motivation, time and physical limitations. 9 Overall, these barriers have been identified in specific populations such as students, teachers and police officers, 9 which may not represent the demographic characteristics, health conditions and lifestyles of the adult populations who use the services available in primary care units. 9,10 For example, counseling received by patients from professional teams in primary care units may stimulate LTPA. 11,12 Counseling is a form of orientation based on information, understanding and support that has the aim of facilitating incorporation of new attitudes as behavioral change towards health self-care, such as better dietary habits and smoking cessation, among others. [13][14][15] This is an effective and low-cost initiative for promoting LTPA within primary healthcare, 11,12 and it has also been recognized as a health education action that has the potential to develop individuals' autonomy to face health-related behavioral change, such as increased active commuting and reduction of sedentary behavior. 16,17 Although the main barriers to LTPA in the Brazilian popula-

Sample size
The sample size was estimated from the average number of appointments registered on the website "Transparent Health" during January and February 2019 (N = 34,275). To ensure a representative sample, the proportion of patients receiving PA counseling from healthcare professionals was considered (this proportion was 30%). The confidence level was kept at 95%, the sampling error was 4% and the design effect was 1.5. With these data, the minimum number of participants was estimated as 745.
However, to take into account potential losses and refusals, this estimate was increased by 10%, to yield a total of 820 patients.
We also agreed to recruit at least 100 extra participants (n = 920) to allow future studies to perform multivariate analyses while avoiding estimation errors. The number of users to be recruited was calculated proportionally to the number of appointments in each primary care unit and ranged from 31 to 92 users.

Participant selection
Participants were systematically selected based on their position in the waiting room at the primary care unit, counting from one to five, from left to right, using the entrance door as a reference. The third patient was approached. If this person refused to participate or did not meet the inclusion criteria, the first person to their left was selected.
Only adults (≥ 18 years) were invited to participate in the study.
Those who did not live in the city or were using the unit for the first time, had any physical limitation that would prevent LTPA (e.g. use of a wheelchair or crutches) or had a cognitive limitation that prevented them from understanding the survey (e.g. hearing loss), were excluded (n = 9).

Data collection
Ten trained surveyors administered face-to-face interviews between April and October 2019. These took place either before or after a consultation with a healthcare professional and were conducted in a private, reserved room to avoid external influences. The average duration of the interview was 18 minutes (standard deviation, SD = 5 minutes; range = 9-55 minutes).

Outcome variables: barriers to leisure-time physical activity
The perception of barriers was assessed through an instrument developed for the Brazilian population. 24 Participants answered the question: How often do the following affect your LTPA? Eight individual barriers were evaluated, based on the ones most frequently mentioned in the literature: 9,25 1) lack of money; 2) feeling tired; 3) lack of company; 4) lack of time; 5) having an injury or disease; 6) fear of injury; 7) dislike of physical activity or exercise; and 8) feeling too old. A five-point Likert scale was available for each barrier, with the following options: always, often, sometimes, rarely and never.
For analysis purposes, the options "always", "often" and "sometimes" were grouped and categorized as "yes" (code 1), thus indicating the presence of a barrier. The options "rarely" and "never" were grouped and categorized as "no" (code 0), which consequently represented the lack of a barrier. The total number of barriers was determined as the sum of the eight barriers, ranging from zero to eight.

Sociodemographic characteristics
The patients' sex was noted and age was grouped into three cat-

Statistical analysis
The data were analyzed using descriptive statistics (average, standard deviation and median of the number of barriers) and using the absolute and relative frequency distribution of qualitative variables. The prevalence of each barrier was determined between the categories of predictor variables, and the association was analyzed using chi-square tests (χ 2 ) for heterogeneity and linear trend. The normality of the distribution of the number of barriers was analyzed using the Kolmogorov-Smirnov test.
The data did not present normal distribution (P < 0.001), and the Mann-Whitney U test was used to compare the number of barriers according to physical activity counseling. Analyses were conducted using the SPSS 26 software, and the significance level was kept at 5%.

RESULTS
A total of 935 users were approached, with a refusal rate of 14% At least one barrier to LTPA was reported by 89% of the participants. The average number of barriers was 2.70 ± 0.1 (median = 3), and this was similar between the participants who reported receiving physical activity counseling and those who did not (2.73 ± 0.08 versus 2.70 ± 0.10; P = 0.822, respectively). Feeling tired (53%) and lack of time (52%) were the barriers most reported (Figure 1).
Female sex was associated with the barriers "feeling tired" (56.3% versus 46.8%; P = 0.015), "lack of company" (47.6% versus 31.1%; P < 0.001) and "dislike of exercise" (29.6% versus 18.7%; P = 0.002). Other associations between sociodemographic characteristics, health conditions, LTPA and perception of each barrier are shown in Table 2. Only marital status was not associated with any of the barriers (P > 0.05) ( Table 2).   This study presents a sample with an important sociodemographic characteristic, representing a population that is more vulnerable to physical inactivity in Brazil (e.g. women, low-income individuals, with chronic diseases and who depend on continuous medication). This is relevant since community programs to promote physical activity in Brazil allow free access for the population with these characteristics. 23,31 In this study, "feeling too tired" and "lack of time" were the most prevalent barriers, and this agreed with the results from other studies. 9,20,25,32 These reports may be partially explained by the partici- Intrapersonal, interpersonal and environmental barriers faced by adolescents, adults and older adults have been widely *P value of χ 2 for heterogeneity reported and identified in the literature. 9,19,32 However, these reports cannot be extrapolated for the general population. In fact, among adult patients seen at primary care units, the main barriers are health conditions and lack of appropriate locations for exercising near their homes, among others. 18,19 However, these barriers differ among individuals. For example, a healthy young single male who is university-educated and has a high income, and who plays sports regularly and intends to swim in a gym, will probably have barriers that are different to those of a middle-aged black female who is a day laborer with a low income and is physically inactive and hypertensive, and who intends or needs to participate in guided walking groups offered at the primary care unit near home. 18,19 Physical activity counseling was inversely associated with the "lack of time" barrier but positively associated with "having an injury or disease". The time available for activities is a multifactorial issue that may not be easily changed. 9,32 However, counseling is a strategy for health promotion that involves orientation and support from professionals. This action may contribute to a more positive perception among individuals with regard to organizing their time for LTPA. In this manner, the relevance of counseling in the process of behavioral change is strengthened. [11][12][13] The positive association between counseling and the barrier of "having an injury or disease" may be explained by studies that have shown that there is a higher probability that older adults and those with chronic disease are the ones receiving counseling. 20,33 The number of barriers was associated with the prevalence of < 150 minutes per week of walking and total LTPA. Similar results were found in a population study in Pelotas (RS), another mid-sized city in the southern region of Brazil. 25 These results are essential and reiterate the idea that counseling needs to be directed at the individual level to stimulate LTPA and reduce the perception of barriers. Therefore, it is clear that reducing barriers should be at the center of healthcare professionals' actions towards increasing the level of LTPA in this population.
Multiple sociodemographic characteristics and health conditions presented associations with barriers to LTPA in the expected ways. For example, the perceptions of "feeling too tired", "lack of company" and "dislike of exercise" were also found to be more likely among women in other studies. These perceptions can be explained by double shifts, lack of support from a partner and, consequently, low motivation for activities. 9   Continuous and adequate physical activity counseling could be emphasized for people with chronic disease or fear of illness, and for people who are insufficiently active during their leisure time. 11,22 This could make it easier for primary care unit users to understand that regular physical activity is not a potential cause of injury or pain. 9 The Physical Activity Guidelines for the Brazilian Population can contribute to a better understanding of physical activity among the population and help healthcare professionals regarding the content for counseling. 36 Also, healthcare professionals could encourage users to attend public open spaces (e.g. parks, squares, sports and leisure centers and fitness zones) that are easily accessible and free, for LTPA. 19,31,32 Some limitations of this study need to be considered in order to adequately interpret and extrapolate these results. First, the sample was limited to adult patients at primary care units in the urban area of a mid-sized city in southern Brazil. Second, these units did not have a physical education professional as part of their healthcare teams, which may have affected the perception of barriers. Third, the quantitative approach using a short, standardized survey to evaluate intrapersonal barriers did not allow much depth when exploring interpersonal or environmental matters of relevance to implementation of physical activity promotion programs in the community.
Therefore, it was impossible to capture contextual information that could explain or signify feelings towards barriers. This would only have been possible through a qualitative or mixed-method approach. 10 Lastly, the cross-sectional design limited the capacity to establish causality between predictor variables and the outcomes.

CONCLUSION
The perceived barriers to physical activity that were most often reported were "feeling too tired" and "lack of time". These barriers differed in importance according to sociodemographic characteristics, health conditions and levels of LTPA. Physical inactivity increased with increasing numbers of perceived barriers. Future studies may advance the analyses and explore other interpersonal and environmental aspects of barriers through using mixed methods among patients at primary care units in larger cities or rural areas of smaller municipalities. It is also relevant to assess the effectiveness of counseling actions provided by trained professionals, using an evidence-based technical protocol to reduce the perception of barriers against LTPA.