The acquired immunodeficiency syndrome (AIDS) has, in recent decades, taken a worldwide epidemiological priority with an estimation of 34 million people being HIV positive, while 2.5 million have already developed clinical symptoms of the disease [1]. With some regional variability, HIV positive prevalence in Mozambique is about 11.5%, mainly concentrated among the inactive and the reproductive population, which results in a high negative impact on the Mozambican economy [2].

Although prevention is recognized as the most effective solution for reducing the incidence of HIV/AIDS, for those who are already infected, antiretroviral therapy (ART) has become a widespread intervention, and because of its success, has transformed the condition from a fatal to a chronic disease [3, 4]. However, the use of ART produces collateral effects that affect the quality of life of patients, including metabolic, somatic and psychological disorders [5, 6].

Among the strategies to deal with the direct effects of HIV and the secondary effects of ART, exercise has been shown to provide many health benefits, ranging from increased aerobic capacity to mood improvement [7]. It is expected that people with HIV/AIDS can reap some of the benefits of exercise, as the general population [8, 9]. Although research in this specific topic is limited, therapeutic exercise is becoming an alternative therapy, commonly prescribed for people with HIV/AIDS [3, 10, 11].

Research results on the benefits of exercise for individuals with HIV/AIDS and the secondary effects ART’s are conflicting and inconclusive, often due to small sample size and a high percentage of dropout [12]. Due to variability in research methods and the influence of factors such as type, frequency, intensity and duration of exercise, conclusions are less consistent regarding the safety and efficacy of the programs [13]. Associated with this lack of specific findings about exercise and HIV, the recommendations regarding the prescription of exercise for developing and maintaining cardiorespiratory fitness in subjects with HIV/AIDS differs little from that stated for the general populations [11, 14].

A specific problem with exercise programs for subjects with HIV/AIDS is the high dropout rates, as well as the low levels of regularity of exercise participation [15]. In the modern world, the so-called “obesogenic environment” limits exercise adherence, which is aggravated in the population infected with HIV/AIDS due to the disease-related symptoms. This calls for specific methodologies that lead to a better retention and regularity of exercise [16, 17].

Manipulation of the type of exercise could be a key point in reducing exercise dropout, considering that pleasure may increase exercise adherence [18]. Although the enormous variability in exercise mode does not allow for the classification of all structured forms of exercise programs, experimental studies tend to compare formal and playful practices [19, 20]. The first is characterized by a design were the workload is individually determined and controlled accurately on an individual basis [10], whereas the second is performed in groups, prioritizing interaction and pleasure [21, 22]. Research on the exercise benefits using these different types of intervention for the population with HIV/AIDS has not been systematically studied.

The aim of this study was to evaluate and compare the effect of formal exercise versus playful activity on the exercise regularity and the health-related fitness components of HIV-infected individuals who take ART.

Baseline descriptive values for the sample are shown in Table 1.

Mean frequency of participation for the 36 exercise sessions in the two groups was similar and statistically non-significant, either in absolute (p=0.23) or relative (p= 0.57) values (Table 2). In both groups, the participation was very high namely, 92.2% and 90.0% for the FEG and PEG groups, respectively.

Table 3 shows the pre and post intervention results by each group. Mean weight and BMI increased in all groups (1.8 to 2.8%) and was statistically significant for CG (p=0.006). Waist reduced significantly in PEG (-3.7%; p=0.008) while skinfolds reduced only in in CG (p=0.01).

Mean values for health-related physical fitness test variables for each group are shown in Table 4. With the exception of peak VO₂, there were no changes between the pre- and post-intervention measures. In the case of peak VO₂, differences were significant in the exercise groups, indicating a positive effect (FEG=14.8%, p=0.008; PEG=11.2%, p=0.01). Mean values in the control group did not change significantly (0.6%, p=0.85).

Table 5 demonstrates the results of pre- and post-exercise comparison in relation to risk factors for cardiovascular disease, as well as CD4+ count. With the exception of cholesterol, which increased in the exercise groups, no differences were founded in all other variables.

This study aimed to evaluate and compare the effect of two types of exercise interventions on the regularity of exercise participation and health related physical fitness of African HIV-positive women using ART. To accomplish this goal, 53 Mozambican women aged 25-56 years from low socioeconomic areas of Maputo underwent a training program for a period of three months using two different exercise protocols, with a control group. Thus, the study simultaneously examined for exercise retention and the health benefits of exercise.

The first observation was that the frequency of exercise participation was very high, but was not significantly different between exercise groups. Attendance rates reported in the literature were substantially lower [13, 18, 20]. However, it must be taken into account that the study duration was very short (12 weeks) and the specific conditions of the participants were different from the natural environment. Moreover, it being the first time that the participants participated in such a program, they had high expectations related to the effects of the program on their health and well-being. These factors have been suggested has possible limitations when generalizing the results in this type of study [26].

Dropout rates have been described in the literature as a key issue affecting exercise adherence and overall health [17, 27]. The dropout rates disclosed in previous studies vary considerably from 0% [20] to 52% [11]. In the present study, the dropout rate was 7.8% in the formal group, 10% in the playful group and 29% in the control group. Thus, there was an overall exercise adherence of 85%, which may be very high considering the type of population recruited.

An average increase in body mass in all groups was observed. This increase was significant in the FEG and CG groups but not in the PEG. However, after the intervention, the difference in body fat measured by waist and skinfolds were not statistically different in two groups and lower in one. This suggests that the subjects, on average, increased their body mass, but not body fat. Although intervention studies with individuals with HIV and ART show different results, there is a trend towards a consistent loss of body mass.

Roubenoff et al. [28] observed similar results in two different studies. First, in a study done with 10 adult males who performed progressive resistance training with an aerobic component for 16-weeks, three times a week, they recorded a reduction in body fat with no change in muscle mass. In another study carried out with 25 adult men during 8 weeks of high-intensity progressive resistance training, they observed a significant increase in lean body mass and a decrease in body fat [29]. The same observation was described by Grinspoon et al. [30] who found that an intervention of progressive resistance three times a week for twelve weeks provided a significant reduction in BMI, adiposity and waist circumference in HIV individuals.

Some studies applied different types of programs in order to test and compare different exercise methodologies. Terry et al. [31] compared two groups of HIV-infected individuals who performed aerobic exercise and stretching for 12 weeks. Body weight, body fat and waist-hip ratio decreased significantly and similarly in both groups. In a study with HIV+ African women on ART and with manifestations of lipodystrophy [6], a significant reduction in waist circumference and body fat folds was observed after 6 months of cardiorespiratory exercise. The magnitude of fat change was much higher than in weight.

Although some studies do not find any effect of exercise [3], the loss of fat that was not accompanied by a loss in weight observed in this intervention is similar to many of the studies described in the literature. This may be caused by the fact that exercise may simultaneously promote a loss of fat and an increase in muscle mass. This phenomenon is more pronounced when subjects begin exercise from a sedentary condition and with poor physical fitness. Prior to beginning this intervention, the subjects were not involved in any regular exercise program. However, a study with this population, using accelerometers, showed that women of low socioeconomic status were very active given that their activities related to their survival. Although this could be an important factor, there is not enough information to assess the extent to which this may have influenced the results.

Weight gain and fat reduction was expected in the exercise groups but not in the control group. Contact with the research project may have altered the behavior of all involved, including the CG, which may have led them to make some changes in a positive direction. Quite noteworthy was that 37.7% of the subjects were either overweight or obese. However, the percentage of subjects who increased their weight was significantly higher in those with normal or low weight than those overweight or obese. Thus, the increase in weight may be seen as a positive effect. Although it cannot be confirmed, but it may be that the patients were taking medication more regularly and eating better, simply because they had contact with the project, which involved regular consultations, strategic testing and motivational counseling.

Unlike the fat values ​​measured by skinfolds, the average body fat measured by bioimpedance indicated a 58.5% increase, ranging from 40 to 77.8%. This result may be considered contradictory, due to two essential reasons. Firstly, the fact that there are no studies on African women in order to confirm the algorithms that estimate fat mass for the research technique used. The few studies available suggest ethnic differences as the reason for incorrect estimation of fat mass, but do not present solutions [8]. Secondly, due to operational shortcomings, the measurement of bioimpedance in our study did not observe some procedures described in the protocol, such as a fasted state and monitoring of the hormonal cycle of the participants. These two aspects may have interfered with the validity and reliability of the values between the two measurements for each subject.

In the context of physical fitness, the most significant result was observed at V̇O₂ peak. As the exercise groups improved and the CG did not, it may be suggested that there was an effect of exercise on this component regardless of the type of activity. In general, studies show that individuals with HIV may increase their maximum V̇O₂ with exercise, similar to the general population [20, 27, 28, 32].

Stringer et al. [33] observed that 6 weeks of a moderately intense exercise program, practiced three times a week for 1 hour, significantly improved V̇O_2max and lactic threshold, apparently without any deleterious effect on the immune system. Macarthur et al. [34] verified in 25 HIV-infected individuals who performed a 24-week physical training program, a significant increase in maximal oxygen uptake (V̇O_2max) occurred, as well as a positive physiological response to submaximal exercise. In Africa, Mutimura et al. [35] found that after 6 months of training, V̇O_2max increased significantly. The same results were observed in several studies with different experimental designs, which allows for the generalization that aerobic exercise improves cardiorespiratory fitness in individuals with HIV [27, 28, 29].

The results of the tests for strength and endurance did not indicate any change in performance after intervention. Opposite to our findings, change in muscle strength and endurance is described in the literature by several authors. Roubenoff et al. [36] found that after 16 weeks of progressive resistance training, performed three times week, increased strength significantly. The same increase was observed in a study in Caucasian males diagnosed as seropositive for HIV-1 virus, with an exercise regimen that included 35 minutes of strength training and flexibility for 12 weeks [37]. Replication of these results is observed in other studies in different age groups [26, 38].

In view of these results, the absence of an increase in strength observed in this study was not expected. Whereas FEG was subjected to endurance training with controlled loads, including intensity and volume, significant changes were expected. The measure of strength by no specific tests, like Handgrip and Leg Curl, or a low load of the program was perhaps not enough to produce changes and could account for the lack of any observed improvement.

The change in metabolic indicators with exercise is controversial and not linear [27]. The studies vary in methodology with regard to intensity, type and duration of training, as well as evaluation methods, making it difficult to generalize the conclusions [32]. In this study, there were no changes in metabolic risk factors with the exception of cholesterol, which showed increases in all groups, including the controls. A literature review found conflicting results on this topic. For example, triglycerides decreased in several studies [4, 11, 12], but increased in others [3, 19, 39, 40].

In the case of cholesterol, our results are opposite to other studies of patients with HIV [5, 8, 15, 19, 25, 41]. No explanation could be found for the increased cholesterol levels. However, it must be noted that all the subjects remained within the limits considered acceptable, which means that the changes do not have clinical significance.

Most intervention studies with exercise have found no changes in the levels of CD4 +, the only immunological indicator used in this study [6, 7, 21]. Thus, the lack of change is not surprising. The concern that exercise could affect the immune system negatively does not seem justified. The depletion of the immune system with exercise has been observed previously, but only in athletes who perform high-intensity workloads [42].

The present results contribute to the findings that clinically-controlled exercise for HIV patients on antiretroviral therapy is safe and beneficial for cardiovascular health, regardless of the type of exercise. The absence of changes in several key indicators is not a specific hallmark of this specific population, since this has also been observed in studies with healthy subjects. The search for effective responses requires further studies which manipulate the intensity, volume, frequency and duration of exercise.

This study had some methodological limitations. First, the lack of evaluation of quality of life and well-being of the subjects before and after the intervention. The absence of a validated instrument to measure quality of life in African women is the reason this variable was not measured. Also, the decision about what type of exercise promotes effective changes in wellness. Lastly, control of exercise intensity in the playful group was not possible for practical reasons.

In conclusion, this study showed that there is no physiological advantage from either one of the two exercise methods used, and both types are safe with no apparent contraindications for HIV patients on ART.