Intensity matters: impact of physical activity energy expenditure at moderate and vigorous intensity on total and abdominal obesity in children

Physical activity (PA) guidelines advocate that children should accumulate at least 60 min of moderate-to-vigorous PA daily. Still, it is not clear how body fat may differ if the same dose of PA is accumulated at different intensities. We aimed to determine the independent associations of energy expenditure (EE) at moderate (MPA) and vigorous (VPA) PA intensity on total and abdominal fat in children and if these associations were moderated by cardiorespiratory fitness (CRF). A total of 326 children (girls = 171, boys = 151) aged 10–12 years had PA assessed with accelerometers. Total fat mass index (FMI) and abdominal FMI were assessed with DXA. CRF was assessed by a cycle ergometer test. Linear regression models were used to model the outcomes with the inclusion of an interaction term to test for moderation effects. An inverse association was found between VPA EE and FMI (β = −0.013, p < 0.001) and abdominal FMI (β = −0.0014, p < 0.001) independent of MPA EE. In contrast, MPA EE was not related to adiposity independent of VPA EE (p > 0.05). The relationships between the PA intensities and FMI and abdominal FMI were moderated by CRF. MPA EE was positively associated with adiposity in children with high levels of CRF, whereas VPA EE remained inversely associated with adiposity regardless of CRF level, although the strength of the association was lower in those with higher CRF levels. PA programs should provide opportunities for children to perform VPA in order to achieve healthier body fat profiles and avoid excess adiposity.


INTRODUCTION
Childhood overweight and obesity are known to track into adulthood and are associated with the early development of several cardiometabolic complications, such as that of type 2 diabetes mellitus, hypertension, and dyslipidemia [1][2][3]. Physical activity (PA) has been recommended as an important component for the prevention and treatment of obesity and its associated chronic diseases [4]. Current guidelines recommend that youth engage in at least an average of 60 min per day of moderate-to-vigorous intensity physical activity (MVPA) [5]. However, for a given dose of MVPA, the proportion of time spent in moderate PA (MPA) and vigorous PA (VPA) in children and adolescents fulfilling the World Health Organization (WHO) PA recommendations may differ, raising the question if those accumulating higher volumes of VPA may have superior health benefits for reduced adiposity when compared to the same volume of MPA. This issue has been addressed in the literature, where reviews of observational investigations in children reported superior benefits of VPA over MPA for adiposityrelated outcomes [6,7]. These findings are likely due to the increased overall energy expenditure (EE) resulting from VPA compared to a similar volume of MPA, which may contribute to a greater impact on one of the components of the energy balance equation, thus, favoring a decrease in adiposity. Nevertheless, the debate on the impact of VPA vs MPA on adiposity remains far from being consensual, with the latest WHO PA guidelines highlighting the need to clarify the uncertainty surrounding this issue [5]. In fact, none of the previous observational investigations in children [6,7] compared the effects of VPA and MPA on adiposity when matched for EE. For instance, in adults, it has been shown that for a given EE (i.e., 500 METs min/week), VPA had superior benefits for improving metabolic syndrome when compared to MPA [8].
We aimed to determine the independent associations of EE dose at the moderate and vigorous intensity on total and abdominal fat in children. Given that higher fitness may predispose children towards having lower body fat [9], we also assessed whether CRF influenced the relationships between PA EE at different intensities and adiposity.

MATERIALS/METHODS AND SUBJECTS Sample
Participants included a total of 326 children (girls n = 171, boys n = 151) aged 10-12 years with cross-sectional data on PA EE and body composition from the Physical Activity and Family-based Intervention in Pediatric Obesity Prevention in the School Setting (PESSOA Project), which was a school-based cluster randomized controlled trial involving healthy children in grades 5, 6 and 7 from 14 Portuguese public schools funded by the Portuguese Foundation for Science and Technology (Grant number: PTDC/DES/ 108372/2008). The research protocol was approved by the Portuguese Ministry of Education and all participants and their parents or legal guardians provided written informed consent prior to their participation and in accordance with the World Medical Association's Declaration of Helsinki on human studies [10].
Physical activity energy expenditure PA was assessed by accelerometry (ActiGraph, GT1M model, Fort Walton Beach, FL, USA). Participants were asked to wear the monitor for four consecutive days, including 2 weekdays and 2 weekend days. The devices were activated on the first day (in the morning) and data were recorded in 15-s epochs and later converted to 60-s epochs to estimate metabolic equivalents (METs) from the counts/min value for each minute using the Freedson equation [11]. The METs for each minute were summed for the whole day to derive METs/min/day. Apart from accelerometer non-wear time (i.e., when it was removed during sleep and water-based activities), periods of at least 60 consecutive minutes of zero activity intensity counts were also considered as non-wear time [12]. A valid day was defined as 600 min (10 h) or more of monitored wear time, and all participants with at least 3 valid days (including 1 weekend day) were included in the analyses [12]. The cutoff values to define the time spent in each intensity of PA were based on the Evenson et al. cutoffs [13]. For the current EE analysis, we used 3 and 6 METs to define MPA and VPA, respectively. The device activation, download, and processing were performed using the Actilife software (v.6.9.1., ActigraphTM, Fort Walton Beach, FL, USA).

Body composition
Body weight was assessed to the nearest 0.1 kg wearing minimal clothes and without shoes, and height was measured to the nearest 0.1 cm.
A whole-body Dual Energy X-ray Absorptiometry (DXA) scan was performed in the morning following an overnight fast (Hologic Explorer-W, software QDR v.12.4, Hologic Waltham, USA) to assess total body fat mass. Abdominal fat mass was determined through partial analyses of the DXA scan, based on regions of interest (ROI). The ROI was marked by one rectangle, set between the upper margin of the second lumbar vertebra and the bottom lower margin of the fourth lumbar vertebra, the sides being defined by the continuation of the lateral limits of the rib cage. Fat mass index (FMI) and abdominal FMI were calculated by dividing the fat mass (kg) by the squared height (m 2 ).

Cardiorespiratory fitness
Cardiorespiratory fitness (CRF) was assessed by a cycle test with progressively increasing workload using an electronically braked cycle ergometer (Monark 828E; Vansbro, Sweden). Initial and incremental workloads were 20 watts (W) for children weighing less than 30 kg and 25 W for children weighing 30 kg or more. The workload was increased every 3 min until the maximal effort of the participants was reached. Heart rate was registered continuously (Polar Vantage, Kempele, Finland) throughout the test. Criteria defined for a maximal test were heart rate >185 beats/min and the perception by the observer that the participant was no longer able to continue the test. Peak power output (W max ) and maximal oxygen consumption (VO 2max ) were calculated according to the formula of Hansen et al. [14]. The VO 2max (mL/kg/min) attained from the cycle ergometer test was used in the analysis and is termed "CRF" from here on.

Statistical analysis
Mean ± SD and median (interquartile range) were used to describe normally distributed and skewed variables, respectively. Differences in descriptive characteristics between males and females were determined using independent sample t-tests or the non-parametric alternative. Multiple regression analysis was used to assess the associations between EE at a moderate and vigorous intensity separately (Model 1a and 1b) and combined (Model 2) with FMI and abdominal FMI. The effect of sex and CRF on the relationship between PA EE and FMI and abdominal FMI was assessed by the inclusion of a sex-interaction term and CRF-interaction term, respectively. All linear regression model assumptions were checked and given the presence of heteroscedasticity, robust standard errors were applied in order to assure valid statistical inferences. Collinearity was assessed using the variance inflation factor and all models had a variance inflation factor of <3. Table 1 presents the characteristics of the investigation participants for the whole sample and by sex. Males were taller, had lower total body and abdominal fat mass, and had higher levels of both MPA and VPA (p < 0.05). In terms of the METs spent in MVPA, 25% came from VPA, while the remaining 75% came from MPA. Approximately, 11% of the sample was categorized as overweight, whereas only 1% were obese.

RESULTS
There was no sex interaction on the relationship between PA EE with FMI and abdominal FMI (p > 0.05). Thus, both boys and girls were included together in all analyses, and sex was added as a covariate in all models. In the simple regression analysis, both MPA EE and VPA EE were inversely related to FMI and abdominal FMI ( In Fig. 1, the influence of an equivalent EE dose of MPA and VPA on FMI (Fig. 1A) and abdominal FMI (Fig. 1B) can be visualized. For a given MET/min/day of EE, both FMI and abdominal FMI were less if the EE came from VPA compared to MPA (Fig. 1).

DISCUSSION
We observed that for a given EE dose measured with objective accelerometer data, only VPA EE was inversely related to FMI and abdominal FMI independent of MPA EE and sex. CRF moderated the relationship between PA EE and adiposity, such that MPA EE was positively associated with adiposity in children with high levels of CRF, whereas VPA EE remained inversely associated with adiposity regardless of CRF level, with a caveat of the strength of the association being lower in magnitude in those with higher levels.
The issue of the role of PA intensity on several health-related outcomes has been at the center of discussion for the past decade [6,7]. In the latest PA guidelines, the WHO advocates that children and adolescents should engage in 60 min of MVPA per day, with VPA being incorporated at least 3 days per week [5]. However, the proportion of how these 60 min per day are accumulated between MPA and VPA may have a differential impact on health-related outcomes, including those of adiposity. Our results indicated that when looking at the independent associations of MPA EE and VPA EE with total and regional adiposity, only VPA EE remained significant. In fact, for the same dose of EE equivalent to the PA recommendations (i.e., 200 METs/min/day considering 3.33 METs at MPA), the FMI of children performing PA at a vigorous and   moderate intensity was predicted to be 3.60 and 5.19 kg/m 2 , respectively. A similar trend was also observed for abdominal FMI, which is noteworthy given the known role of central adiposity on cardiometabolic health [1,2]. Thus, it is important to consider that PA intensity may play a bigger role in energy balance that goes beyond just EE with possible implications on body fat. From a physiological standpoint, exercise performed at higher intensities may promote greater fat usage, which could be related to mitochondrial quantitative and qualitative adaptations, such as superior increments in skeletal muscle mitochondrial volume and density and intrinsic mitochondrial fatty acid oxidation [15]. Moreover, in an intensity-dependent manner, PA has also been implicated in the regulation of a plethora of hormones and myokines related to body composition, having possible effects on energy balance and appetite control [16,17]. For example, catecholamines and IL-6 have been shown to drive lipolysis both during and after PA, with implications on fat mobilization and storage [18,19]. Similar to our findings, several reviews have highlighted the potential benefits of objectively measured VPA in children, including those on total and central adiposity [6,7,20]. When looking at the several observational investigations included in these reviews, they all expressed PA intensity in counts/min or in min/day. Such approaches do not account for the EE of MPA and VPA intensities and how they can vary substantially across the energy spectrum (i.e., 3-6 METs for MPA and 6+ METs for VPA). In addition, the min/day at a given intensity can change significantly depending on which cut point is chosen, with values ranging as much as 3000-4012 counts/min for VPA [6]. All of these previous constraints limit the compatibility of MPA and VPA with different health-related outcomes, whereas matching both intensities for the same EE allows us to better understand these relationships [8].
To the best of our knowledge, this is the first investigation to analyze the relationship between MPA and VPA at a matched EE objectively measured with accelerometry on total and regional adiposity assessed by DXA in children. Using the NHANES database, Janssen and Ross [8] also observed that VPA had a greater influence on the prevalence of metabolic syndrome when compared to a similar EE dose of MPA in adults.
Given that CRF has been found to modify the relationship between PA at different intensities and body composition in children [21,22], we also analyzed the possible moderating effect of CRF on the relationship of VPA EE and MPA EE with total and regional adiposity. We observed an interaction effect of CRF on the associations between both intensities and FMI and abdominal FMI, such that MPA EE had an unfavorable association with adiposity in children within the highest tertile of CRF. On the contrary, VPA EE remained inversely associated with adiposity regardless of CRF level, although this relationship was attenuated in children with high CRF. Similar results were previously reported in a sample of children and adolescents [21], where MPA (measured in min/day) was positively associated with waist circumference in children with high CRF, whereas VPA was inversely associated with waist circumference in children with low CRF. However, in contrast to our results, an unfavorable relationship was observed between VPA and waist circumference in children and adolescents with high CRF [21]. Such results highlight once more the importance of performing PA at higher intensities given its favorable relationship with adiposity irrespective of CRF.
The use of EE derived from accelerometer devices to measure free-living activities is not consensual and can be seen as a possible limitation [23], since the EE equations are dependent on the activities used to validate them. Nonetheless, the equation used in this investigation to derive EE from counts/min has been previously validated with reasonable agreement at the group level [11]. Another limitation is our cross-sectional design, which limits our ability to draw conclusions on causality. Despite the current body of literature composed of both observational and experimental evidence [20,[24][25][26], the latest WHO guidelines highlight the uncertainty surrounding the study of PA intensity and its relationship with adiposity [5]. This issue is the result of the complex nature of energy balance, which is not only affected by PA EE, but also by other confounding factors, such as dietary intake, gut microbiome, and genetic variations, which were not accounted for in the current investigation. In addition, although the intensity is a major contributor to the health-related outcomes associated with PA, there are other factors identified in the literature that should not be overlooked, such as the PA type and movement patterns [27]. Finally, a more diverse sample in terms of body composition would aid in a better understanding of the relationship between PA intensity and adiposity.

CONCLUSION
The findings from this investigation suggest that intensity matters and VPA may have additional benefits on total and abdominal adiposity that are independent of EE dose and CRF level of children. On the other hand, MPA EE was not related to FMI and abdominal FMI when accounting for VPA EE. The contribution of the EE within the scope of the energy balance equation is dependent on PA intensity to influence total and abdominal adiposity, and not just total EE. For any given EE, girls and boys that perform more intense PA tend to have improved adiposity phenotypes. These results add to the current body of literature and further stress the importance of incorporating a higher proportion of the total PA METs coming from vigorous intensity into the different domains of everyday living in children, such as in school physical education classes, sports, and during recreational PA. Moreover, future public health policies and PA guidelines should consider the independent roles of MPA and VPA when addressing adiposity outcomes.

DATA AVAILABILITY
Data are available from the corresponding author upon reasonable request. Table 3. Relationships of physical activity energy expenditure at different intensities with fat mass index and abdominal fat mass index in youth with high and low cardiorespiratory fitness adjusted for sex.