Overweight/obesity in childhood is one of the most serious public health challenges of the 21st century [1
] and causes multiple adverse health consequences [2
]. Worldwide, over 42 million children under 5 years of age are estimated to be overweight/obese [1
]. In Sweden, the prevalence of children (aged 2–10 years) with overweight and obesity is ~10–15% [4
]. These facts are concerning as overweight has been found to track from childhood to adolescence [5
]. Therefore, interventions in young children are required, to attempt to prevent the development of overweight/obesity, and consequently, lower the risk of noncommunicable diseases [1
Physical activity (PA) affects positively on energy balance, and thus, it has been considered an important factor in counteracting childhood obesity. The doubly-labeled water (DLW) method is the gold standard in measuring total energy expenditure (TEE) during free-living conditions [8
]. When applied correctly, the DLW method can produce estimates of TEE with an accuracy of 1–3% and a precision of 2–8% [9
]. Moreover, the use of the multipoint protocol (i.e., several postdose samples spread out during the 10–14 day turnover period) instead of the two-point protocol (i.e., postdose samples on the first day of dosing when isotopic equilibrium has been established and in the end of the turn-over period) has been suggested to be used in association studies when maximum precision is desired [10
]. Together with basal metabolic rate (BMR), TEE provides reference estimates of activity energy expenditure (AEE) and physical activity level (PAL), which are essential in identifying associations between energy expenditure in response to PA and body composition.
Considering the impact of PA on energy expenditure it is relevant to examine the associations of PAL and AEE with body fatness in childhood. The few conducted studies in school-aged children have reported that PAL is inversely correlated with fat mass [11
]. However, obesity in very young children has been associated with later disease [1
] and considering that obesity at a young age tracks into later life [5
], it may be of great importance to also consider the relevance of PA to prevent obesity already at the preschool age. However, to date, very few studies [14
] have examined the associations of PAL and AEE with body composition in young children and these studies have reported somewhat contradictory results. For instance, Salbe et al. [14
] and Eriksson et al. [15
] found inverse associations for PAL or AEE with fat mass, although two other studies did not report such associations [16
]. These differences may be due to differences in methods, statistical analysis, and characteristics of study participants. Importantly, the few available studies in preschool-aged children warranted further investigation to increase the knowledge of early PA on body composition indices. This is of importance considering the negative influence of obesity already in the preschool age [1
Furthermore, physical fitness has been found to be an important health indicator [18
]. For example, Ortega et al. [20
] found an inverse relationship between physical fitness levels and obesity in children and adolescents and corresponding associations between fitness and fat mass has also been reported in preschool-aged children [22
Therefore, the purpose of this study was to examine the associations between PAL and AEE assessed using the DLW method with body composition indices and physical fitness measures in healthy Swedish children aged 5.5 years. Our hypothesis was that a higher PAL and AEE are associated with favorable body composition and improved physical fitness.
The participating children’s mothers (n = 40) had an average age of (years) 36 ± 4.2, average height (cm) of 1.67 ± 0.06, average weight (kg) of 68 ± 11.4, average BMI (kg/m2) of 24.3 ± 4.0, and 72.5% of them (n = 29) had completed a university degree. The background information of the children (22 boys and 18 girls) is presented in Table 1
. The children covered a wide range for body composition, TEE, AEE, PAL, and physical fitness.
shows the associations between PAL and AEE with body composition measurements. After adjusting for sex and age, a higher PAL was significantly inversely associated with a lower BMI, BF%, and FMI (standardized β −0.35, −0.41, and −0.45, all p
< 0.036). Similarly, a higher AEE was associated with a lower BMI, BF%, and FMI (standardized β −0.44, −0.44 and −0.47, all p
In the sensitivity analyses, when we also adjusted the models with PAL as a dependent variable for fat-free mass, the significant associations of PAL with BMI, BF%, and FMI became stronger (standardized β −0.42, −0.47, and −0.48, all p < 0.001).
With regards to physical fitness, a 0.1 higher PAL was associated with a 5.28 cm better score for the standing long jump test after adjusting for confounders (standardized β 0.37, p
= 0.017) (Table 3
). Furthermore, a higher AEE was associated with a better score for the standing long jump test (standardized β 0.38, p
= 0.014). There were no statistically significant associations found between PAL or AEE with any other physical fitness tests or FFMI (all p
The main results of this study are that a higher PAL and AEE were associated with a lower BMI, BF%, and FMI as well as with better lower-body muscular strength. This is of great importance, since a healthier body composition in childhood has been found to be associated with a healthier cardiovascular profile later in life and with a lower risk of death [19
]. Furthermore, the positive association between PAL and AEE with lower-body muscular strength is important, as muscular strength has been found to be positively connected to perceived health status and life satisfaction in children [21
]. Additionally, improvement in muscular strength from childhood to later in life has been found to be negatively related to changes in overall adiposity [19
] as well as positively related with skeletal health [18
Previous studies have also reported significant associations between a higher PAL and AEE with a lower BF%, fat mass, and BMI [12
] in children as well as between a higher AEE and a lower fat mass [14
]. However, in a cross-sectional study, Butte et al. [16
] found no relationship between PAL and body composition indices in 4.5-year-olds. Other studies in younger [15
] and older [11
] children have found that PAL has been negatively associated with fat mass but not with fat-free mass. Moreover, Butte et al. [16
] found significant associations between AEE and BMI, but not with BF% or fat mass. It is important to note that fat-free mass was not taken into account in their analyses, and since fat-free mass has been found to have an important role in energy expenditure [33
], this may be a possible reason for the observed differences between the two studies. Rennie et al. [13
] adjusted their analyses for fat-free mass, and they found significant relationships between PAL and AEE with FMI. Similar to our study, the relationships were weaker without adjusting for fat-free mass. However, despite adjusting for fat-free mass, they observed no association between AEE and BMI [9
]. This may be due to differences in sample sizes (100 children vs. 40 children) and study populations (e.g., regarding activity patterns). However, these findings further emphasize the need for proper adjustment for body size when investigating the association between PAL and AEE with body composition indices in children.
We did not observe any significant associations between PAL and FFMI, which is in line with a previous study [16
], but contradictory to one another [13
]. The inconsistent results may be partly explained by the differences in the utilized methodologies as mentioned above, or it could also be due to the fact that PAL reflects overall PA and not a specific intensity level. High-intensity PA has been found to be positively related to FFMI [27
], and thus, it is possible that the children in our study had less high-intensity PA compared to the children in the other study [13
]. Since heart rate has been found to influence energy expenditure [34
], having more time spent in low-intensity PA may result in similar energy expenditure levels as having less time spent in high-intensity PA. Although the influence on body composition may be different. This is important since children differ in terms of how they perform PA, with some children preferring low-intensity PA while the others are more interested in high-intensity PA.
PAL and AEE were both positively associated with better lower-body muscular strength, but not with other components of physical fitness. To our knowledge, there are no previous studies examining the associations between PAL or AEE with physical fitness in preschoolers. Thus, comparison to other studies is not possible. However, our findings are reasonable, since PA in young children may be more focused on maintaining and improving the lower-body muscles (i.e., running, jumping, and cycling). In our previous study [27
], we observed a positive association of high-intensity PA assessed using accelerometry with cardiorespiratory fitness and motor fitness. The lack of significant associations in this study may be due to the fact that the children had less high-intensity PA as discussed earlier, as the accelerometer data showed that they only performed on average 7 min of vigorous-intensity PA per day [27
]. Yet, in order to improve all indices of physical fitness (muscle strength, motor fitness, and cardiorespiratory fitness) enhancing various PA is essential. Nevertheless, more research in larger preschool populations is needed in order to increase knowledge regarding these associations.
The primary strength of our study is the use of the DLW method which is considered the gold standard in measuring total energy expenditure in free-living settings [8
]. The multipoint protocol was used to assess TEE via the DLW method. According to the International Atomic Energy Association (IAEA) handbook [32
], the multipoint protocol is advantageous compared to the two-point method as data from several time points is collected which means data is averaged and the analytical error is reduced. Furthermore, Djafarian et al. [10
] evaluated the two protocols in preschool-aged children and found that there were no significant differences between the average values for TEE between the two protocols. However, they stated that when maximum precision is desired the multipoint protocol should be used, such as for association studies such as this one [10
]. The body composition indices used in this study were based on the 3-component model, which has been validated against the 4-component model and a very high agreement between the models has been found [35
]. Furthermore, to our knowledge, there are no previous studies investigating the association of PAL and AEE (assessed using the DLW method) with physical fitness in preschool-aged children.
A limitation of this study is that we used a predicted BMR based on weight instead of a measured BMR. However, we also fitted PAL based on BMR predicted from fat and fat-free mass using raw data from previously published material in another sample of Swedish preschool-aged children [36
] and the results were consistent (data not shown). The present study is also limited by the relatively small sample size (n = 40), which is common in studies using high cost methods (i.e., DLW). Hence, the findings should be confirmed using larger sample sizes. However, it is important to highlight that the children’s TEE was similar to data in children living in Western countries [37
], and additionally, their body size was similar to Swedish reference data [39
]. Finally, the level of maternal education level was higher than in the general Swedish population (79% versus 52%, respectively) [40
]; however, when adjusting for maternal education in the models, conclusions remained the same.