This study aids understanding of the natural history of cardiovascular disease. Over a 20-year follow-up, a significant increase in the prevalence of some CVRF (excessive weight, hypercholesterolemia, hypertension, and alcohol consumption) was found despite a significant reduction in the prevalence of sedentary lifestyles. There was no significant change in tobacco smoking. These findings are very similar to the findings of an earlier study of the physicians included in the present study when analyzed 15 years after the first analysis , although the current study included a broader group of health professional and not only the physicians, over a 5 years longer period of time.
Health education is an essential tool in the prevention and control of chronic non-communicable diseases [17–19]. Monitoring the evolution of some risk factors for CVD in individuals with health-related degrees could, theoretically, determine the real impact of health education as a measure of health promotion because this population has formal training in healthcare and therefore has deeper knowledge of the risks and effects of risk factors for CVD and the risk behaviors associated with these diseases.
An increased prevalence of overweight over time has been reported by Brazilian and international studies analyzing the evolution of the BMI in specific populations [20–23]. Although the population examined in this study had received formal education in health-related fields, the evolution the BMI was similar to that of the general population, with a significant increase in the prevalence of excessive weight over a 20-year follow-up. However, when the prevalence of excessive weight in the study population was compared with data from individuals belonging to the same age group and living in Brazilian state capitals and the Federal District, the study population presented lower rates (32,1 × 55,9 %), even when the study population was compared only with the general population with higher education levels (>12 years of education)  (32,1 × 48,4 %).
Despite a significant rise in hypercholesterolemia over the study period, the prevalence was much lower in the study population than in the general population. A study in nine Brazilian state capitals in individuals with a mean age of 35.4 years found total cholesterol levels >200 mg/dL in 38 % of males and in 42 % of females , compared with 24.2 % of males and females in the study population. Another large study of more than 81,000 individuals with a mean age of approximately 43 years from 13 cities in different regions of Brazil also found a higher prevalence of hypercholesterolemia (40 %)  compared with the present study.
Comparing the number of hypertensive individuals in the study population in both evaluation periods with the population of a large city in the same region  divided by age group showed that study participants had lower rates of hypertension in both phases, even though there was an increase in the prevalence of hypertension between the first and second phases. These data corroborate the relationship between hypertension and low education levels  as the study population had a higher level of education than that of the general population of the city.
Alcohol consumption increased significantly in study subjects between the first and second study phases, and the prevalence was similar to that of the population of the Brazilian state capitals. There was a trend to an association between increased alcohol consumption and higher educational levels .
There was a significant reduction in sedentary lifestyles from 50.2 % in the first phase to 38.1 % in the second phase. In the general population of nine Brazilian capitals, the prevalence of sedentary lifestyles ranged from 28.2 to 54.5 %; however, unlike in the study subjects, the general Brazilian population tended to be more active between 15 and 24 years of age . Our results showed an inverse relationship between educational level and the prevalence of sedentary lifestyles, similar to the results of a Brazilian telephone-based survey of risk and protective factors for chronic diseases (VIGITEL) .
We found that even among the CVRF that increased significantly in the study period, the prevalence rates were lower than those observed in the general population. These data are in agreement with findings from the Nurses Health Study II  and the Physicians Health Study I , in which the prevalence of CVRF was significantly-lower among American nurses and physicians, respectively, compared with the general population [31–33].
The VIGITEL study  showed a prevalence of smoking of 12.1 %. Smoking was more prevalent in individuals with lower educational levels and in those aged between 25 and 65 years. The prevalence rates of smoking of 4.6 and 3.7 % found in the first and second phases of our study, respectively, show that the health professionals analyzed study smoke less than the general population. We also found a trend to an association between less smoking and higher educational levels, similar to Brazilian  and international  data.
We found a similar increase in excessive weight, hypercholesterolemia, and hypertension between men and women between the first and second study phases, but no difference in smoking. There was a significant reduction in sedentary lifestyle, and an increase in alcohol consumption in women but not in men. The results on sedentary lifestyle diverge from international  and Brazilian  data, which show higher rates of sedentary behavior among older women. The trend to increased alcohol consumption in women as they age is also not in agreement with other reports [24, 36].
Analyses of BP, cholesterol, blood glucose, and BMI values showed a significant increase in all of these variables over 20 years. The values of DBP, SBP, blood glucose, cholesterol, and BMI found in 1993 correlated with those obtained in 2013. Thus, individuals with higher levels in 1993 also had higher levels in 2013. These data may be useful for strategies for the early detection of modifiable risk factors and effective modifying interventions, as suggested by a study of CVRF in schoolchildren  and a report by the Bogalusa Heart Study .
One limitation of the study is that not all individuals assessed in 1993 were located in 2013. This may be due to the long time interval, the lack of updated and integrated records from Brazilian health regulators, and the mobility of the study population, with professionals working in regions distant from the study region. This latter factor is especially important because Brazil is a country of continental dimensions, with vast differences in regional development. Even so, the reassessment of > 75 % of the initial group makes the sample representative and supports the conclusions presented.
Another limitation was the use of different methods for cholesterol and blood glucose analysis in the two study phases. However, various reports confirm the correlation between the values obtained using these different methods, suggesting they have no impact on data analysis [12–15]. Likewise, the use of different devices to measure BP was not considered relevant because a validated semi-automated device  and a standardized measure technique  were used in both phases of the study.
Additional studies are needed to determine the possible impact of higher education on health care as a protective factor against chronic degenerative diseases, specifically CVD. This was not an initial goal of this study, which only aimed to assess the evolution of some CVRF over time. A comparison with reported data [24, 28] adjusted for age and educational level indicates a positive difference in health professionals compared with the general population; however, only a study with an appropriate design could objectively answer this question. On the other hand, despite the decrease in sedentary lifestyle in this population there was an increase in the prevalence of hypertension, hypercholesterolemia and excessive weight, as well as an increase of the mean values of all the studied variables, suggesting that aging itself might be the most important aspect rellated to the onset of cardiovascular risk factors.