A cross-cultural study of physical activity and sedentariness in youth from Mozambique and Portugal

Aim: The purpose of this study was to analyse the effects of age, sex, body mass index (BMI) and country of origin (Mozambique versus Portugal) on physical activity and sedentariness. Methods: Sample size was comprised of 593 adolescents from Mozambique (n=258) and Portugal (n=335), aged 10-15 years old. Height and weight were measured, and BMI was computed; furthermore, subjects were classified in two groups (normal weight and overweight/obese). The mean time (min/day) spent in moderate-to-vigorous physical activity (MVPA), light physical activity (LPA), and sedentary activity were assessed using the GT3X Actigraph accelerometer over at least 4 days (1 weekend day) with a minimum of 10 hours/day of monitoring. Multiple regression analysis was done in SPSS 20. Results: A low percentage of Mozambican (3.9%) and Portuguese (4.3%) youth complied with the MVPA recommendations (60min∙day-1). However, Portuguese boys spend more time in MVPA. Girls and those with higher BMI spend more time in sedentary activities and less time in MVPA and LPA; older adolescents spend more time in sedentary activities and less time in LPA. There is a significant effect of the country of origin on MVPA, favouring Portuguese youth. Conclusions: Results suggest country differences in MVPA between Mozambican and Portuguese youth – the Mozambicans are less involved in MVPA than the Portuguese, probably reinforcing the undesirable effect of the urbanization process in developing nations.


Introduction
The health benefits of moderate-to-vigorous physical activity (MVPA), especially among youth, are well known 1,2 .It is also acknowledged that, around the world, the urbanization process affects children and adolescents' daily lives in many different ways [3][4][5] .Further, this impact is more evident in developing countries, causing drastic changes in lifestyles 6 , leading to decreasing levels in habitual physical activity (PA), increasing levels in sedentariness 6 , and as a consequence, an escalation in the prevalence of overweight and obesity 7,8 .
Despite both developed and developing countries being affected by the urbanization process, they are in different stages of this process 9 .Developed nations are in the so-called "last stage" of nutritional transition 9 , with a high prevalence of sedentariness and obesity, but with people becoming more aware of the benefits of balanced diets and physically active lifestyles, and that changing diets and PA habits will, in all likelihood, prevent or delay co-morbidities related to physical inactivity and obesity.Developing nations, such as Sub-Saharan countries are, in general, in the fourth stage of this transition 9 , characterized by people becoming more affluent with the adoption of a westernized way of life, with the expected increments of overweight/obesity prevalence and a more sedentary daily life.As a consequence, there is a trend in developing nations for high prevalence of sedentariness and a relevant decrease in MVPA levels, namely among youth, with similar frequencies of those found in developed nations 10,11 .
PA differences in European countries have been systematically studied.For example, Aibar, Bois, Generelo, Zaragoza Casterad, Paillard 12 reported that Spanish adolescents were more physically active than French adolescents; in addition they reported that only 35.9% of adolescents complied with current PA guidelines (60 min/day of MVPA).Further, Ramirez-Rico, Hilland, Foweather, Fernandez-Garcia, Fairclough 13 reported that English adolescents were less sedentary and more active than their Spanish peers, and the percentage of adolescents achieving the recommended MVPA levels was about 25% overall.
To our knowledge, few studies have investigated differences in PA among children from countries with diverse socio-economic characteristics.In a recent review of worldwide variation in PA levels of adolescents from 105 countries, Hallal et al. 10 reported that more than 80% of youth do not achieve the recommended daily MVPA levels; further, this prevalence tended to be lower in developing nations.In the African context, Peltzer 14 reported that only 14.2% of children from 8 countries were physically active (5 days or more in a week with at least 60 min•day -1 ); additionally there are large differences among countries meeting the PA guidelines, 17.7% in Uganda and 9.0% in Zambia.
Portugal and Mozambique are countries with different social, economic and cultural characteristics.A previous study conducted in 2009 comparing PA levels in 6-to 16-year-old school-age children and adolescents from these two countries 15 reported that rural Mozambican youth engaged in significantly higher PA levels than their Portuguese peers.However, due the urbanization process that has affected the lifestyles of youth from these two countries, this scenario could not be seen in a more recent study 16,17 .Recent data have shown that the proportion of Portuguese children aged 10-11 years complying with the recommendation of at least 60 min•day -1 of MVPA is 36% (51.6% for boys, 22.5% for girls), and that this prevalence decreases with age, with only 4% (7.9% of boys, 1.2% of girls) of adolescents aged 16-17 years attaining the recommended MVPA guidelines 17 .Although there is no nationally representative data for PA in Mozambican youth, there is some evidence of a decline during the last decades 15,16 .Given that the urbanization process may be contributing to a decrease in children and adolescent's PA levels, the purpose of this study was to analyse the effects of age, sex, BMI and country of origin (Mozambique versus Portugal) on levels of PA and sedentariness in youth.Hence, we hypothesize the presence of significant associations of these covariates in youth PA levels and sedentariness in both countries.

Maputo City and Porto Metropolitan areas and sociodemographic characteristics
Maputo and Porto are two different cities from two different countries in cultural, social and economic terms.Briefly, the city of Maputo (the capital) has 1,194,121 inhabitants with a life expectancy of 57.5 years; its population density is of 3,443.6 per km 2 , and the population growth rate is 1.37%.The illiteracy rate and the child mortality are 9.8% 20 and 55.8/1000 births 21 , respectively.In comparison, the Porto metropolitan area has 1,287,256 inhabitants, with a population density of 1,580.3 per km 2 22, 23 and a population growth rate of -0.58%.The life expectancy of 78 years, the child mortality of 3.1/1000 births, and the illiteracy rate of 3.3% makes it as one of the most developed in Portugal 22 .

Anthropometry
Standardized procedures described by Lohman, Roche, Martorell 24 were used in data collection in the two countries.In Portugal, children were measured wearing light clothes, without shoes or socks, whereas in Mozambique children were assessed without shoes and naked in two private rooms, one for each sex (girls were measured by female technicians, and boys were measured by male technicians).Height was measured to the nearest 0.1 cm with the head positioned in the Frankfurt plane with a portable stadiometer (Holtain Ltd, UK); a calibrated portable Seca (Seca, Germany) scale (in Mozambique) and a portable bioelectrical impedance (TANITA BC-418 MA Segmental Body Composition Analyser; Tanita, Corporation, Tokyo, Japan) scale (in Portugal) were used to measure weight.BMI was computed using the standard formula [weight(kg)/ height(m) 2 ], and the International Obesity Task Force cut points 25 were used to classify subjects in two groups (normal weight and overweight/obese)

Physical Activity and Sedentary Time
PA and sedentary time (ST) were monitored by Actigraph GT3X accelerometers (ActiGraph, Pensacola, FL).Children and adolescents wore the accelerometer at their waist on an elasticized belt, placed on the right mid-axillary line, for at least 7 days, including 2 weekend days; all were asked to remove the accelerometers only when sleeping and when performing waterbased activities (such as taking a shower or swimming).To be eligible for this study, youth had at least 4 days (including at least one weekend day) with a minimum of 10 hours of monitoring Physical activity and sedentariness in Mozambique and Portugal per day.All 593 children fulfilled this condition.Non-wear time was defined as any sequence of at least 20 consecutive minutes of zero activity counts 26 .

Statistical Analysis
Initially, basic descriptive statistics and t-test, by sex and country, were computed.Analysis of covariance, adjusted for age and mean wear time of the accelerometer (WTA), was used to compare mean differences in BMI, MVPA, LPA and ST among Portuguese and Mozambican boys and girls.Further, multiple regression was used to investigate the effect of country (Portugal vs Mozambique), sex, age and BMI on MVPA, LPA and ST.All statistical analysis was done on SPSS 20 and the significance level was set at 5%.

Results
Descriptive statistics for BMI, PA and ST are shown in Table 1.On average, Portuguese and Mozambican boys and girls failed to achieve the 60 min•day -1 recommended MVPA levels.Except for MVPA, significant differences were found for LPA and ST -Mozambican children and adolescents have, on average, more LPA and ST; further, Portuguese youth have higher BMI values.Figure 1 presents graphical representation of the mean values of BMI, PA and ST, adjusted for age and WTA (except for BMI that was only adjusted for age).After adjustments, BMI differences remained, but no statistically significant differences were observed for LPA and ST between countries; on the other hand, MVPA differences became significant for boys and for the total sample.
A very low frequency of children and adolescents, from both countries, complied with the recommended MVPA daily levelsonly 4.3% in Mozambique and 3.9% in Portugal.Additionally, a moderate to high frequency of children were classified as being overweight/obese: 18.2% in Mozambique and 32.5% in Portugal.

Discussion
In this paper we found differences in LPA and ST between countries, where Mozambican youth (especially boys) spent more time in these behaviours when compared with their Portuguese peers.Differences in PA between children and adolescents from many nationalities have been reported by Aibar, Bois, Generelo, Zaragoza Casterad, Paillard 12 , Ramirez-Rico, Hilland, Foweather, Fernandez-Garcia, Fairclough 13 , and Verloigne et al. 28 , showing differences not only in distinct PA expressions, but also in sedentariness (both determined by accelerometer).Similarly, using questionnaire data from 8 African countries, Peltzer 14 reported differences in school-children and adolescents' PA.These results highlight the potential of social, cultural and environmental correlates on youth behaviour, namely their PA 29,30 .
Cross-cultural studies contrasting youth PA differences from countries with diverse socio-economic characteristics are scarce.Additionally, if available they frequently report PA levels or sedentariness by country or region, aiming to explore the worldwide patterns in these behaviours 10,31 .A previous study 15 compared PA levels (assessed by accelerometry) between rural Mozambican and urban Portuguese youth, and reported that Mozambicans had significantly higher total PA, showed less PA decline with age, but engaged in fewer minutes at high intensity PA than their Portuguese peers.The results for MVPA from the present study are somewhat similar to these previous results, with Mozambican boys and girls having lower MVPA than their Portuguese peers; however, significant mean differences were observed in LPA or ST.These results can be explained in different ways.Firstly, in the last decades, Mozambique passed through a social and economic transition period, with changes in lifestyle with visible consequences in youth PA and sedentariness, especially those from urban/suburban areas 16 .As such, children and adolescents are more sedentary and spend less time in activities of moderate and high intensity, and this can be a possible explanation for the lower time spent in MVPA found among Mozambican youth as compared to Portuguese.
In general terms, youth from both countries do not comply, on average, with recommended MVPA levels (60 min•d -1 ).This low mean value reflects the percentage of children attaining the recommended MVPA guidelines: 8.3% and 7.7% in Mozambican and Portuguese boys, respectively, and only 1.3% and 1.5% in Mozambican and Portuguese girls, respectively.Although these are very low values, and of great concern, they are in accordance with the current worldwide trend which is characterized by low PA levels, in association with high ST 10 .For example, European youth data showed that only 4.6% of girls, and 16.8% of boys, aged 10-12 years, from 5 countries, reached the daily MVPA guidelines 28 .In the African context, Muthuri, Wachira, Onywera, Tremblay 32 reported that among 9-11 years old Kenyan children, only 12.6% met the MVPA recommendations, while Peltzer 14 concluded that, among adolescents from eight African countries (Botswana, Kenya, Namibia, Senegal, Swaziland, Uganda, Zambia, Zimbabwe), only 14.2% spend at least 60 min•d -1 in MVPA on 5 or more days per week.
Results of the multiple regression analysis showed significant effects of sex, age, BMI, WTA and country on PA and sedentariness.There are consistent lines of evidence showing sex and age differences in PA, where girls and older children and adolescents tend to be less active and more sedentary [33][34][35] .Our results reinforce this trend, since girls spend less time in MVPA and LPA, as well as more time in sedentary activities; older adolescents are less engaged in LPA but are more sedentary than their younger counterparts.There is evidence that BMI is inversely associated with PA and positively associated with sedentariness [36][37][38] .Among Mozambican and Portuguese adolescents, higher BMI was associated with lower MVPA and LPA, but also with higher time spent in sedentary behaviour.This result is of concern because of the observed increases in the prevalence of overweight/obesity in young population groups during the last decades 7,8 .In addition, children with more WTA had more MVPA, LPA and ST as expected, since the higher the amount of time recorded by the accelerometer, the higher the time spent in different PA phenotypes and sedentary activities will be registered.
The last PA and sedentariness predictor was country of origin, and the results showed that Mozambicans tend to spend significantly less time in MVPA.It is possible that this result is related to lifestyle changes observed in Mozambique during the last decades, mostly due to the urbanization process, which lead to a decrease in leisure time activities linked with increments in electronic entertainment media use among youth as previously suggested 39,40 ; as a result, children became less active and more sedentary 16 ; however, because we do not have any objective information regarding this issue, we can only speculate about this possible association.In addition, there seems to exist a trend among African children and adolescents to be engaged in greater amounts of LPA as compared to MVPA 15,32,41 , although there were no differences in LPA observed in this study.On the other hand, a greater involvement of Portuguese children and adolescents in sports may also explain these results, since their sports practice gives them opportunities for engagement in more energy-demanding activities 15 .
This study is not without limitations.Firstly, it was not considered a proportional sample in each country.Furthermore, the present study did not intend to have representative samples from each city, nor from each country, as it is similar to other well-known studies 42,43 .Secondly, we do not have information regarding the context of the PA and sedentariness (we only focused on objectively determined overall MVPA, LPA and ST), which could provide a better understanding of the crosscountry differences.Thirdly, although cut-off points advocated by Treuth et al. 27 are internationally accepted, they tend to produce lower estimates of PA than other published cut-points, which can partly explain the low percentage of children complying with the recommended 60 minutes of daily MVPA.Fourthly, we did not adjust PA or ST for biological maturation because a large number of subjects did not have information in sitting height (about 20% of the sample).Yet, as suggested by Sherar, Cumming, Eisenmann, Baxter-Jones, Malinal 44 , the association between biological maturity and PA is rather inconsistent.
This study also has several strengths.Firstly, the use of an objective method (accelerometer) to measure PA and ST is a marked strength.Secondly, the use of the same accelerometer protocol in both countries ensures a high degree of robustness in the data.Thirdly, the use of samples from very different countries in socio-economic and cultural terms provides a unique opportunity to study correlates of PA and sedentary behaviour.

Conclusions
In summary, the present study found cross-country differences in MVPA between Mozambican and Portuguese youth.In addition, in both countries, a low percentage of children complied with the recommended daily MVPA levels.In general, girls, older children and those with higher BMI tend to spend less time in MVPA and LPA and more time being sedentary.Mozambicans are less involved in MVPA than Portuguese.Since low levels of MVPA in early life are related to obesity and its comorbidities in adulthood (i.e., they track over time), promoting strategies and programs to enhance active living styles during this period should be primary public health and education targets.

Figure 1 :
Figure 1: Graphical representation of means±standard errors for BMI, PA and ST after adjustment for age and WTA, by country.
those younger than 10 or older than 15 years, were excluded during sample selection and/or data screening.All children involved in the study had a written consent signed by parents or legal guardians.The Mozambican National Bioethics Committee (Mozambique, ethical number: 44/CNBS/05) and University of Porto ethics committee (Portugal, ethical number: 15/CEUP/2012), as well as school authorities, approved each study protocol.

Table 1 :
Descriptive statistics (mean±SD), t and p-values by sex and country.

Table 2 :
Frequency of children attaining recommended MVPA levels, by age, sex, and country.

Table 3 :
Multiple regression analysis results for MVPA, LPA and ST predictors.