Introduction: Self-reported measures of habitual physical activity rely completely on the respondent’s ability to provide accurate information on their own physical activity behaviours. Our aim was to investigate if obese adolescents could accurately report their physical activity levels (PAL) using self-reported diaries.
Methods: Total energy expenditure (TEE) was measured using doubly labelled water (DLW) and resting energy expenditure (REE) was measured via indirect calorimetry. Activity energy expenditure (AEE) and PAL values were derived from measured TEE and REE. Self-reported, four-day activity diaries were used to calculate daily MET values and averaged to give an estimated PAL value (ePAL).
Results: Twenty-two obese adolescents, mean age 13.2 years, mean BMI 31 kg/m2, completed the study. No significant differences between mean measured and estimated PAL values were observed (1.37 versus 1.4, P=0.74). Bland Altman analysis illustrated a significant relationship (r=-0.76, P<0.05) between the two methods; thus the bias was not consistent across a range of physical activity levels, with the more inactive overreporting their physical activity.
Conclusion: At an individual level, obese adolescents are unlikely to be able to provide an accurate estimation of their own activity.
Alex’s Notes: Self-reported questionnaires are useful in large-scale studies, no doubt. Unfortunately, they rely on the child (or parent) to recall physical activity behavior information accurately, which we know is fraught with difficulties and grossly inaccurate in obese adults. It stands to reason that children, especially obese children, are not going to perform much better. The current study recruited 22 children/adolescents aged 10-16 years (average 13-years) with a BMI over the 90th percentile (average 31). That means these kids were fatter than 90% of the other children their age. Despite the obesity, they were otherwise healthy.
Total energy expenditure (TEE) was measured using the doubly labelled water (DLW) method, which is considered to be the most accurate method available for the assessment of TEE. Resting energy expenditure (REE) was measured via indirect calorimetry after an overnight fast, and activity energy expenditure (AEE) was measured using simple math (TEE-REE). The PAL value (TEE/REE) was also calculated with math. Physical activity was measured via a four-day self-reported diary that was completed on three week days and one weekend day (the kids were given verbal and written instructions with an example of how to complete the diary), and upon completion, the activities were categorized into nine levels according to their average energy costs, representing multiples of their respective metabolic equivalents (METs), which were then averaged to give a PAL value for each subject.
And not surprisingly,
“The group as a whole were sedentary, with girls reporting on average 4.7 hours per day of “screen time” and boys 3 hours per day. At a group level, there was no statistically significant difference between mean measured and mean estimated PAL values (1.37 versus 1.4, resp., P=0.74). A Bland-Altman analysis illustrated a significant relationship (r=-0.76, P<0.05) between the difference and the mean and between the two methods, demonstrating that the bias (-0.01) was not consistent across the range of PAL values. Subjects with a low measured PAL tended to overreport their activity, while those with a higher measured PAL underreported their activity using the diary.”
What I did find surprising was just how sedentary the children were. On average, time spent in sedentary activities such as sitting and watching TV occupied 45% of the day. Sleeping was another 35%, meaning that 80% of the day was spent basically not moving. In fact, the kids spent more time with “self-care” then both playing outside and organized sport combined.