There are three primary components that constitute total daily energy expenditure: resting metabolic rate, the thermic effect of food and physical activity thermogenesis. Resting metabolic rate accounts for approximately 50 to 60 % of TDEE. Variation between individuals is small and is essentially a function of body size. The thermic effect of food accounts for only 10 to 15 % of TDEE, whilst the contribution made by physical activity thermogenesis can vary enormously from a meagre 15 % in the sedentary to 50 % in the active (Donahoo et al., 2004; Levine, 2004b). There is more limited data on the components of TDEE in children under 12 years of age children, however, the relative contribution of the three components would appear similar (DeLany et al., 2006; Montgomery et al., 2004).

Levine and colleagues (Levine, 2004a; Levine et al., 2006) have shown that the variation in physical activity thermogenesis is better understood when it is sub-divided into exercise activity thermogenesis and non-exercise activity thermogenesis (NEAT). NEAT is unstructured activity and constitutes all the lifestyle movement patterns and postural adjustments engaged in as a matter of course, rather than the energy expended through more purposeful or structured exercise activity thermogenesis (EAT). NEAT comprises three sub-components. These include body posture, ambulation, and all other movements, essentially fidgeting (Levine et al., 2001b). In comparison to the 1-2 % variation in TDEE contributed by EAT, NEAT varies enormously between individuals and has been shown to play an important role in the maintenance of body weight (Levine et al., 2005). This is best illustrated by Levine's observation that it is not EAT that delineates the lean adult from the obese; rather it is sub-components of NEAT. Their discovery that obese adults sit for 164 minutes per day more than lean adults, whilst the lean stand and ambulate for 152 minutes per day more than the obese, led to the conclusion that the variation in NEAT in adults was primarily related to ambulation and body posture (Levine et al., 2005). The energetic benefit of standing and ambling ranged from 269 to 477 kcal per day. Levine et al., 2005 suggest that if the obese adult adopted the postural and ambulatory patterns of the lean, they would expend an average 350 kcal per day more, and assuming energy intake remained constant, resist further weight gain.

The role of NEAT in paediatric obesity has yet not been evaluated, but may prove to play a key role in modulating weight gain in children. Variation in body posture and ambulation may also distinguish the lean from the obese child. Theoretically, manipulating a child's living environment to enhance NEAT, and create a positive gain in TDEE, could lead to the prevention of excess fat-mass. These concepts have not yet been tested.

The complexity of assessing NEAT and its sub-components has been highlighted by Levine and colleagues in their development and validation of the Physical Activity Monitoring System (PAMS) for this purpose (Levine et al., 2001a; 2001b; 2003). Recently Lanningham-Foster and colleagues (2005) validated the PAMS for use in children, which promises future evidence of the primary predictors of NEAT in children, as well as an understanding of the role NEAT may play in the variation in TDEE in youngsters.

Advances in the objective assessment of physical activity have meant we have amassed a sizable literature on childhood physical activity patterns over the last two decades. Whilst there is minimal evidence of the degree to which specific aspects of activity account for variance in TDEE, there is evidence of the predominant pattern and intensity of activity in children. This information can help build a better understanding of the relative contribution activity of differing intensities makes to the variation in total physical activity.

There is an abundance of data which shows sustained periods of physical activity are rarely apparent during the daily movement repertoire of children (Armstrong et al., 1991; 2006; McManus et al., 1995; Welsman et al., 1997). Through a comprehensive series of studies, Armstrong and colleagues have provided probably the largest heart rate derived activity data set for children (Armstrong et al., 2006). Moderate intensity physical activity has been defined as activity, which elicits a cardiovascular response between 140 to 159 beats·min^-1. This threshold has been found to correspond to walking at 4 km·hr^-1 in both Caucasian and Chinese children (Armstrong, 1998; McManus et al., 2004), and has largely been adopted as an appropriate marker of moderate intensity activity (Livingstone et al., 2003). Likewise, a threshold of >159 beats·min^-1 has generally been accepted as an appropriate marker of vigorous intensity. Very few children sustain 20-minute periods of at least moderate activity (Armstrong et al., 1991; Gilbey et al., 1995; McManus, 2006; McManus et al., 1995). Even a single 10-minute period of sustained moderate intensity activity was absent in at least one quarter of 7 to 12 year olds (Armstrong et al., 1991; Gilbey et al., 1995; McManus, 2006; McManus et al., 1995).

Children's preference for intermittent short-duration movement has long been recognized (Astrand, 1952) and is best illustrated by the work of (Bailey et al., 1995). Through two separate analyses (Bailey et al., 1995; Berman et al., 1998), bouts of activity in children were found to be frequent (83 and 89 bouts per hour in boys and girls respectively), with a mean duration of 20s. The majority of activity bouts were of low-to-moderate intensity, interspersed with very occasional bursts of vigorous intensity movement (median duration 3s). As such there has been a shift away from an emphasis on encouraging sustained periods of physical activity (Sallis et al., 1994), toward accumulating moderate to vigorous physical activity throughout the day (Cavill et al., 2001; Strong et al., 2005).

Figure 1 provides a summary of the time spent being moderately (A) and vigorously (B) physically active per day in 4 to 12 year old boys and girls. These data are drawn from a range of European and Asian studies using heart rate monitoring (Armstrong et al., 1991; Falgairette et al., 1996; Gilbey et al., 1995; Harro, 1997; McManus, 2006; McManus et al., 1995; Sleap et al., 2000; Welsman et al., 1997). Many of the European children attain or are close to attaining the recommended 1 hour per day of moderate intensity activity (Strong et al., 2005). In stark contrast, the Hong Kong Chinese boys and girls accumulated on average 37 minutes per day less moderate intensity activity than the European children. A similar pattern is apparent for the vigorous intensity activity, with Hong Kong Chinese children accumulating far less vigorous activity than either the European or Singaporean children.

Distinguishing between sedentary and low intensity activity is problematic with heart rate monitoring because of extraneous influences upon low level heart rates such as emotions and fitness. A threshold of > 120 beats.min^-1 has been used to quantify lower intensity activity and these data are available for Singaporean and British children (Gilbey et al., 1995; Sleap et al. , 2000). Singaporean boys accumulated 167 minutes per day of light intensity activity, whilst their British counterparts accumulated 118 minutes per day. The Singaporean girls spent 116 minutes of the day in light intensity activity, in comparison to 105 minutes per day in the British girls. If these values are considered relative to the total physical activity level (low, moderate and vigorous activity combined), light intensity activity accounts for between 71 % and 74 % of total physical activity level in the Singaporean boys and girls respectively. In the British boys and girls, light intensity activity accounts for 59% and 64% of total physical activity respectively. Moderate to vigorous activity comprised only 29% of the total active time for Singaporean boys and 26% of total active time for Singaporean girls. Whilst 41% of total active time was spent in moderate to vigorous activity in the British boys, and 36 % in the British girls (Gilbey et al., 1995; Sleap et al., 2000).

Accelerometry studies are more difficult to interpret because of variation in devices used, as well as inconsistent definitions and thresholds for intensity (Freedson et al., 2005). Accelerometers are however, easy to use and affordable and a number of large-scale studies have provided extensive information on the physical activity levels of European children. Dencker and colleagues reported that 9 to 10 yr old Swedish boys and girls accumulated 210 and 190 minutes of moderate intensity activity per day respectively (Dencker et al., 2005). Boys accumulated forty-six minutes of vigorous activity, whilst girls accumulated thirty-five minutes per day. Similarly aged boys and girls from Sweden and Estonia (from the European Youth Heart Study) accumulated 182 and 161 minutes of moderate activity respectively (Ruiz et al., 2006). The boys accumulated 31 minutes of vigorous activity per day, whilst the girls accumulated 22 minutes per day. Further data from over 1000 participants of the European Youth Heart (Ekelund et al., 2004) found boys spent 12 % of their day engaged in activity of at least moderate intensity, whilst girls accumulated somewhat less (8.7 %). This translates to an average of 86 minutes for the boys and 63 minutes for the girls, and is quite similar to heart rate estimates from Armstrong and colleagues (Armstrong et al., 1991; Welsman et al., 1997). Interestingly both data sets are interpreted using a threshold equivalent to walking at 4 km.hr-1. What is clear from this large scale data set is that most of a European child's day is spent being sedentary or engaged in light intensity movement. This is true of both boys and girls who were shown to spend similar amounts of time being sedentary (61.8 and 65.3 % respectively) or engaged in light intensity movement (26 % for both).

The most substantial evidence of the relative contribution activity of differing intensities makes to the variation in overall physical activity levels of children comes from a study of young Scottish children (Montgomery et al., 2004). Total daily energy expenditure was assessed using doubly-labelled water in 2 to 7 year olds, and activity thermongenesis and physical activity level calculated. Specific dimensions of physical activity were assessed using accelerometry, and the relative contribution sedentary, light, and moderate to vigorous intensity activity made to physical activity levels determined. It was found that moderate to vigorous activity explained only about 10 % of the total variance in physical activity level. Instead, the variation in physical activity level appeared to relate primarily to light intensity movement and sedentary behaviour. It was concluded that in children moderate to vigorous activity contributes little to overall physical activity levels. Parallel to Levine's work, these findings would suggest that the most promising approach to increasing TDEE in children would be the conversion of sedentary behaviour to low intensity activity, as opposed to focusing on moderate to vigorous intensity activity.