Physical Activity among Primary Health Care Physicians and Its Impact on Counseling Practices

Physical inactivity is one of the five leading risk factors for mortality worldwide. It is believed that physical inactivity is responsible for nearly a third of the ischemic heart disease and diabetes burden and a fifth of the breast and colon cancer burden [1]. The WHO regions of the Eastern Mediterranean Region (EMR) and the Americas have the highest prevalence of physical inactivity, with 34.9% and 39.3%, respectively [2]. It is estimated that about one-quarter (23.3%) of adults globally are physically inactive. In addition, more than three-quarters (76.3%) of adolescents are physically inactive [3]. In Qatar, a population-based survey showed that 45.9% of the Qatari population had a low level of physical activity (PA) [4].

The World Health Organization (WHO) recommendations on PA are to perform 150 min of moderate-intensity exercises or at least 75 min of vigorous-intensity exercise throughout the week [5]; however, evidence suggests that physicians’ PA level is low. A study conducted in Cameroon among primary health care physicians (PHCPs) found that 86% of physicians were not meeting the recommended level of PA [6]. Similarly, a cross-sectional study in Bahrain showed that only 29.6% of physicians exercised =30 min/week, and only 13% reported PA =5 days/week [7]. In Saudi’s western region, a cross-sectional study among PHCPs showed similar results to the study in Bahrain. Out of the 160 physicians studied, 65% were physically inactive, and 35% reported doing some sports activities [8]. However, a recent study conducted in Lebanon showed that 63.7% of PHCPs exercised regularly, but it did not mention whether they were meeting the recommended WHO level of PA [9].

Physicians’ well-being is an essential factor for them to perform at their full capacity. The literature shows that physicians who do not adopt healthy lifestyle behaviors are less likely to counsel their patients on healthy lifestyle behaviors. One study revealed that physically inactive physicians are less likely to counsel patients on activity [10]. Another study among US medical students illustrated that patients’ PA counseling practices were consistently related to the medical students’ PA practices [11]. Moreover, patients of overweight and obese physicians have less trust and adherence to their physicians’ medical advice than physicians of normal weight [12].

The PA among PHCPs and related counseling practice in Qatar remains unclear. This study aims to assess the PA levels and counseling practices among PHCPs in Qatar. Additionally, we aim to provide insights into the impact of PA on PHCPs’ counseling practices, thus, helping create a ground to identify ways to improve the health and well-being of both physicians and their patients. In this study, we hypothesized that physicians who engage in regular PA might have a more positive attitude toward health promotion hence better counseling practices for their patients.

An analytical cross-sectional design was utilized. This study was conducted across all 27 primary health care centers (PHCCs) in Qatar. Physicians running the clinics are either board-certified family physicians (FPs) or general practitioners (GPs). Data collection was carried out between October and December 2020. Approval from the Institutional Review Board was obtained from the PHCC Research and Ethical Committee.

All FPs and GPs working in the listed 27 PHCCs were approached; thus, no sampling technique was applied. The total number of PHCPs was 634, which was acquired from the last annual report produced by the PHCC in 2016 [13], and it was used to calculate the minimum required sample size, but the actual number at the time of data collection was 648 PHCPs. Excluded participants were those who did not consent to participate or who were not practicing at their clinics at the time of data collection (e.g., on annual leave). We calculated the minimum sample size required to obtain sufficient study power using the following formula [14]:

Using precision (d) of 5% and an a error of 0.05 corresponding to a 95% confidence level (1.96), the estimated minimum required sample size based on this calculation was n = 235.

The PHCPs were approached in their clinics by the researchers. The researchers conducted multiple visits to each PHCC in the country to recruit the participants. The PHCPs’ written informed consent was obtained after a summary of the survey was delivered. After a physician finished the survey, they recorded their weight and height.

The items assessing the PA behaviors in the data collection tool were adopted from the WHO stepwise survey questionnaire [15]. To collect data about counseling practices, a modified self-administered questionnaire adapted from a previous study was used [16]. The questionnaire was piloted in two health centers among 35 PHCPs. Based on the given feedback, no significant changes to the tool were applied.

Sociodemographic variables included age, gender, ethnicity, marital status, job title, and years of experience.

The recommended PA by the WHO is 150 min of moderate-intensity PA throughout the week or at least 75 min of vigorous-intensity PA throughout the week [17]. According to WHO, PA is classified into [18]:

• Inactive: defined as doing no or very little PA at work, at home, for transport, or during discretionary time.

• Insufficiently active: defined as doing some PA but less than 150 min of moderate-intensity PA or 60 min of vigorous-intensity PA a week accumulated across work, home, transport, or discretionary domains.

• Sufficiently active: defined as at least 150 min of moderate-intensity PA or 60 min of vigorous-intensity PA a week accumulated across work, home, transport, or discretionary domains, which approximately corresponds to current recommendations in many countries.

For adults, WHO defines the normal weight as having a BMI between 18.5 and 24.9 kg/m² and underweight as having BMI below 18.5 kg/m², while overweight and obesity are defined as having a BMI =25.0 - <30, and =30.0, respectively [19].

Counseling was measured as the percentage of patients counseled regarding PA. The average percentages for counseling practice were transformed into categories using the quartiles.

The database was constructed using the Statistical Package for Social Sciences (SPSS)™ Software Version 25. The analysis included the following steps. First, the descriptive statistics were tabulated in frequencies and percentages for categorical variables or mean ± standard deviation for continuous variables. Little’s Missing Completely at Random (MCAR) test was used to assess if the missing data were MCAR, and after including all study variables, the test concluded that the missing data were MCAR with a nonsignificant p value of 0.43 and a ?² value of 1,092.26. The number of respondents differed between questions; thus, percentages were calculated according to the number of PHCPs responded to that specific question using a pairwise deletion method to deal with missing data. The median and interquartile range (IQR) was used to describe continuous, non-normally distributed variables. To test the association between PA and basic characteristics, a ?² test was used. Where appropriate, when the expected count was below 5 in 20% of the cells, a nonparametric equivalent was used (Fisher’s exact test). All factors that were significant in univariable analysis were further considered in multivariable analysis. To compare ordinal and not normally distributed continuous variables between different subgroups, nonparametric Mann-Whitney and Kruskal-Wallis tests were used. p values of less than 0.05 were considered significant.

Out of 648 physicians registered as PHCPs in the PHCC, 94 were on annual leave or outside the country, and 43 PHCPs were deployed to COVID-19 isolation facilities and not practicing in their primary HCs during the time of data collection. Out of the remaining 511 PHCPs, 306 (59.9%) physicians responded to the survey and consented to participate. Those 306 physicians were distributed across the 27 HCs from central, western, and northern regions.

The number of respondents differed between questions, with the marital status question having the highest 100% (n = 306) response and the nationality question being the lowest with a 71.5% (n = 219) response rate. Percentages were calculated according to the number of PHCPs responded to that specific question. The majority (58.1%) of the respondents were males. The average age was 45.8 ± 7.9 years, with half of the participants in the 41–50 age-groups. The average age for males and females was 47 ± 8 and 44 ± 7 years, respectively. With regard to nationality, participants come from 21 different countries. 4.1% of the physicians were Qatari citizens, and most of the physicians were non-Arab (55.8%), with the UK nationals making up the largest share of the participants (43.8%). 88.9% of participants were currently married, 51.0% held the job title “Family physician consultant,” and 30.2% were GPs. PHCPs had an average of 14 ± 8.3 years of experience, with most of them (82.1%) having 20 or fewer years of experience. The BMI measurements showed that 50.5% of physicians were overweight, and 23.1% were obese, with an average BMI of 27.6 kg/m² (Table 1).

The majority of the physicians (60.5%) did not meet the WHO recommendations for PA. Furthermore, PHCPs reported spending an average of 3.32 ± 2.46 h per day sitting or reclining, excluding working hours. 81.4% of the PHCPs spend 1–5 h, and 18.6% spend between 6 and 11 h sitting or reclining. When asked about walking habits, 204 (67.5%) physicians out of the 302 respondents reported that they walk a minimum of 10 min per day for an average of 4.6 + 1.9 days per week.

Table 2 depicts the univariable analysis for the association between meeting the WHO recommendation for PA and physician characteristics, statistically significant associations were found between being male, non-Arab, and FP consultant and meeting WHO recommendations. Males (50.0%) were more likely to meet the WHO recommendations than females (26.1%); the prevalence of meeting the WHO recommendations for PA was higher among non-Arabs (61.9%) than Arabs (19.1%) and higher among FP consultants (53.5%) than FP specialists (32.7%) and GPs (21%); all p values <0.001.

After performing the logistic regression, adjusting for gender, ethnicity, and job title, we discovered that ethnicity is the only main predictor for meeting PA recommendations. Non-Arabs were 5.55 times more likely to meet PA recommendations (odds ratio: 5.55; 95% confidence interval [CI]: 2.6–11.8) when compared to non-Arabs (Table 3). Out of the 291 physicians who answered the related PA questions, only 128 physicians (44.0%) were sufficiently active based on the WHO PA classification, while 52 (17.9%) and 111 (38.1%) were insufficiently active and inactive, respectively.

In terms of the PA levels and their association with the PHCPs characteristics, male physicians were more likely to be sufficiently active (56.8%), while female physicians were more likely to be inactive (54.6%; p < 0.001). Non-Arabic physicians were more likely to be sufficiently active (66.1%) compared to physicians of an Arabic ethnicity who were more likely to be inactive (58.4%; p < 0.001). Physicians with a job title of FP consultant were more likely to be sufficiently active (56.9%), whereas, for GP and FP specialists, they were more likely to be inactive (55.6%) and (40.8%), respectively (p < 0.001). Also, the length of experience was inversely associated with the PA level. Physicians with ten or fewer years of experience were more likely to be sufficiently active (51.1%). In comparison, physicians who had more than 30 years of experience were more likely to be inactive (60.0%; p < 0.05). Interestingly, the physician’s BMI was not associated with the PA level or meeting the WHO recommendations (Table 4).

The ordinal regression revealed that years of experience and job title were not contributing to the overall PA levels. Gender and ethnicity were the most important predictors for PA levels. Males were more likely to have higher levels of PA than females with an odds ratio of 2.4 (95% CI: 1.2–4.6; p = 0.006). At the same time, non-Arabs were more likely to achieve higher PA levels (odds ratio = 5.7; 95% CI: 2.7–12.0; p < 0.001).

Although 97.2% of physicians believed that it was very to extremely important to counseling their patients on PA, the median percentage of patients counseled about PA during a routine visit was only 60 (IQR: 40–80)%. For that purpose, they spend a median of 3 (IQR: 2–5) minutes per visit counseling the patients. When asked about the percentage of patients that they believe to follow their recommendation, physicians reported that only one-quarter (median = 25 [IQR: 10–50]%) of their patients follow the provided recommendations on PA.

When investigating the relationship between counseling on PA and the physicians’ characteristics, gender was the only characteristic significantly associated with the counseling practices. Female physicians reported counseling a higher percentage of their patients than male physicians (p < 0.001) (Table 5).

The univariable analysis using the Mann-Whitney test showed no association between meeting PA recommendations by the PHCPs and the percentage of patients counseled on PA (p = 0.732). Furthermore, when assessing the relationship between the average percentage of patients counseled on PA, and the different levels of PHCPs PA, the Kruskal-Wallis test showed no significant association (p = 0.752) (Table 6).

Our study found that 60.5% of physicians did not meet the WHO recommendations for PA, and 38.1% were considered inactive. PHCPs spend an average of 3.32 ± 2.46 h per day sitting or reclining, excluding working hours, 81.4% spend 1–5 h, and 18.6% spend between 6 and 11 h. Our findings differ markedly from the Al-Baker et al. [20] 2020 study conducted in Qatar, which showed that 82.5% of PHCPs did not meet the WHO recommendations with a difference of more than 20% from our findings. Although they did not mention when the data were collected in Al-Baker et al. [20] study, they noted that the total number of PHCPs in PHCC was N = 252; therefore, this study must have been carried out before 2011 because the total number of PHCPs was 253 in that year and the PHCPs number in PHCC been increasing steadily since that time [21].

When comparing our results with the general population through the Qatar stepwise survey [4], our data did not differ vastly, where 45.9% of the general population were considered inactive, and females were more likely to engage in low-level PA (54.2% vs. 37.0%). The average time spent reclining by the general population was 2.9 h, which is less than that of our finding, which could be attributed to the age difference. Approximately 70% of the stepwise population were below 44 years of age, whereas more than 70% of the PHCPs were above 41 years of age.

By performing logistic regression for not meeting WHO recommendations among PHCPs, ethnicity (being an Arab) was the only significant factor. It is difficult to compare this finding with other studies assessing ethnicity as a predictor for PA level among physicians due to the scarcity of studies comparing Arabs to non-Arabic physicians. Data from neighboring Arab countries revealed different outcomes to our study. A study from Saudi Arabia reported that only 21.1% of PHCPs were sufficiently active and 78.9% engaged in an insufficient PA [22], and that’s way below our average for physically active (44%) and higher than our average of insufficiently active (56%). Nevertheless, in that study, 60% of the participants were females, while in our study, 58.1% of the participants were males, and this could have been a reason for their low engagement in PA. Although gender was not found to be significant after performing logistic regression in our study, the female gender was a strong predictor for low PA in other studies from the literature [4, 10, 23‒26].

Some studies showed better PA behaviors among health care professionals than ours, but it was observed that younger age was a common factor in these studies. Studies from the USA, Brazil, and Malaysia showed that almost 66%, 53.1%, and 55% of PHCPs were physically active, respectively [24, 27, 28]. However, the average age for the US physicians was 34 years, and 70% of the health care professionals in the Malay and Brazilian studies were below the age of 40. Nevertheless, in our study, there was no significant association between age and PA.

PHCPs play a crucial role in encouraging change in patients’ behavior for a better health outcome by counseling and promoting PA. In our study, 93.1% of physicians counseled on average 55.8% of their patients, and they spent a median of 3 (IQR: 2–5) minutes per visit. This percentage was higher than the findings of another Qatari study by Al-Baker et al. [20] 2020, where only 79.5% of PHCPs advised their patients about PA. Moreover, Al-Baker et al. [20] 2020 revealed that 59.5% of physicians counseled 1–40% and only 4.5% counseled 81–100% of their patients, which was lower than our findings where 56.9% of PHCPs counseled 41–80% of patients, and 13.7% counseled 81–100% of patients. Furthermore, when comparing the time devoted per visit, our findings reported similar time dedicated to that of Al-Baker et al. [20] 2020 study in the above 2 min category (65.3% vs. 64.5%); these findings show that there has been an improvement in the counseling practices over the past 10 years among PHCPs in Qatar.

Few articles from the Arab region addressed the actual counseling practices among physicians. For example, a study conducted in Saudi Arabia revealed that only 61.1% of physicians offered counseling on PA to their patients [10]. Another study conducted in Jordan showed a similar counseling outcome where 95.45% of physicians counseled =50% of their patients, but no physician (0%) counseled =76% of their patients [29]. When asked about the importance of counseling on PA, 64.4% of physicians believed it is very/extremely important, which was less than our findings of 97.4% for the same question.

On the international level, PA counseling practices vary. Between 1999 and 2018, we found six studies from the US that assessed PA counseling. According to these studies, the percentages of physicians who counseled their patients on PA varied between 20% and 69.1% [27, 30‒34]. In Canada and Ireland, physicians offer more counseling on PA than in the USA (85% and 86.4%), respectively [35, 36]. Similar results came from a review of 22 studies conducted in Brazil in 2020 to estimate the prevalence of PA counseling in PHCCs, where the prevalence varied from 20% to 59.4% [37]. Western countries, like the USA, have been promoting PA counseling since the eighties, while Qatar introduced its first national PA guidelines in 2014; however, data have shown that we have achieved a higher rate of increase in PA counseling than the western world [38, 39]. This could be due to the huge investment the government of Qatar had put into the health care system over the past decade.

In our study, females counseled more patients on PA compared to male physicians (61.8% vs. 51.9%) (p value = 0.001). The same observation was reported in Saudi Arabia [40]. This is different from a study from the USA, which showed that gender was not associated with lifestyle counseling [41]. PA counseling was not associated with any other sociodemographic characteristic.

The positive association between physicians’ personal PA and PA counseling is well documented [10, 16, 20, 31, 42‒44]. However, our findings did not agree with that narrative when we investigated the percentages of patients counseled on PA between those who met the WHO recommendations and those who did not (56.7% vs. 55.6%).

This study has multiple strengths. First, it is particularly important to conduct studies of this type in Qatar and especially in PHCC. Qatar is a diverse country with a multicultural society, and approximately 1.5 million of its population depends on PHCC for health services [45]. Second, the PHCPs practicing in PHCC came from different backgrounds, portrayed by having participants from 21 different nationalities in our study. This supports our study’s external validity and makes our results more generalizable. Third, the relatively large sample size in our study and the fact that the sampling frame included all PHCCs in the country reduce the selection bias possibility and support our results’ generalizability.

One of the limitations of our study is being conducted while the health sector was under pressure due to the COVID-19 crisis. Thus, there is a possibility that our study may not reflect the regular practices outside the pandemic. A low response rate (59.9%) and incomplete data were among the limitations of our study. The completeness level varied between the returned questionnaires. The number of respondents differed between questions. However, Little’s MCAR test showed that the missing data were completely at random hence it is unlikely for the missing data to introduce bias in our results. Furthermore, to overcome this issue, percentages were calculated according to the number of PHCPs responded to each specific question using a pairwise deletion method to handle missing data. Additionally, this study may have been the overestimation of reported counseling by PHCPs, given that we relied on self-reported counseling practices rather than asking patients or observing their encounters with physicians. Another potential limitation of this study is that there are variables that could play as confounders or effect modifiers for both PA level and counseling practices, such as the work schedule and physicians’ workload, which were not assessed as part of the participants’ characteristics.

PHCPs demonstrated a high prevalence of physical inactivity. Furthermore, the practice of counseling the public on PA was low. Preventive services targeting physicians to improve their lifestyle behaviors would positively impact their health and the general population’s health. Creating a counseling culture in the health care society is essential. Overall, there is a need to integrate education about PA and into the medical school curriculum and promote PA among physicians.

We would like to acknowledge all the physicians who participated in this study for their time.

This study protocol was reviewed and approved by the Institutional Review Board in PHCC Research and Ethical Committee. Approval reference No: PHCC/DCR/2020/08/094. Informed written consent was acquired from all participants. Confidentiality and the anonymity of the information were ensured.

The authors have no conflicts of interest to declare.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Open Access funding provided by Qatar National Library.

Mohammed Alyafei: conceptualization, methodology, investigation, formal analysis, data curation, project administration, writing – original draft, and writing – review and editing. Mohamad Alchawa: formal analysis, writing – original draft, and writing – review and editing. Abdulaziz Farooq: methodology, formal analysis, and writing – review and editing. Nagah Selim: methodology, supervision, and writing – review and editing. Iheb Bougmiza: methodology, formal analysis, data curation, supervision, and writing – review and editing.

Data are not publicly available due to ethical reasons. Further inquiries can be directed to the corresponding author.