Background:Objective methods can improve accuracy of physical activity measurement in field studies but uncertainties remain about their use. Methods: Children age 11 years from the Avon Longitudinal Study of Parents and Children (ALSPAC), were asked to wear a uni-axial accelerometer (MTI Actigraph) for 7 days. Results: Of 7159 children who attended for assessment, 5595 (78%) provided valid measures. The reliability coefficient for 3 days of recording was .7 and the power to detect a difference of 0.07 SDs (P ≤ .05) was > 90%. Measures tended to be higher on the first day of recording (17 counts/min; 95% CI, 10-24) and if children wore the monitor for fewer days, but these differences were small. The children who provided valid measures of activity were different from those who did not, but the differences were modest. Conclusion: Objective measures of physical activity can be incorporated into large longitudinal studies of children.

We aimed to measure the prevalence of physical inactivity (PI) during leisure time and to identify variables associated with it in a southern Brazilian adult population. A population-based cross-sectional study was carried out, covering a multiple-stage sample of 1,968 subjects aged 20-69 years. Weekly participation in leisure-time physical activity was addressed. For each activity, energy expenditure was calculated using data on duration, metabolic equivalent, and body weight. Energy expenditures of individual activities were summed to give a weekly total. PI was defined as fewer than 1,000 kilocalories per week. The prevalence of PI was 80.7% (95%CI: 78.9-82.4). After adjusted analyses, the following variables were positively associated with the outcome: female gender, age, living with a partner, and smoking. Schooling and economic status were inversely associated with PI. Chronically undernourished individuals were significantly more likely to be inactive. We found no differences according to skin color or alcohol consumption. In conclusion, the prevalence of PI in this adult population was higher than in populations from developed countries, but the associated variables were similar.

BackgroundObesity is a growing health problem in India and worldwide, due to changes in lifestyle. This study aimed to explore the independent associations between dietary and physical activity exposure variables and total body fat and distribution in an Indian setting.MethodsIndividuals who had participated in the Indian Migration Study (IMS) or the Andhra Pradesh Children And Parents' Study (APCAPS), were invited to participate in the Hyderabad DXA Study. Total and abdominal body fat of study participants was measured using DXA scans. Diet and physical activity (PA) levels were measured using questionnaires.ResultsData on 2208 participants was available for analysis; mean age was 49 yrs in IMS, 21 yrs in APCAPS. Total energy intake was positively associated with total body fat in the APCAPS sample: a 100 kcal higher energy intake was associated with 45 g higher body fat (95% CI 22, 68). In the IMS sample no association was found with total energy intake, but there was a positive association with percent protein intake (1% higher proportion of energy from protein associated with 509 g (95% CI 138,880) higher total body fat). Broadly the same pattern of associations was found with proportion of fat in the abdominal region as the outcome. PA was inversely associated with total body fat in both populations (in APCAPS, one MET-hour higher activity was associated with 46 g (95% CI 12, 81) less body fat; in the IMS it was associated with 145 g less body fat (95% CI 73, 218)). An inverse association was observed between PA and percentage abdominal fat in the IMS but no association was seen in the APCAPS population.ConclusionsIn this Indian population, there was an inverse association between PA and body fat. Associations between body fat and dietary variables differed between the younger APCAPS population and older IMS population. Further longitudinal research is needed to elucidate causality and directions of these associations across the life course.Electronic supplementary materialThe online version of this article (doi:10.1186/s12889-015-1550-7) contains supplementary material, which is available to authorized users.

Measurement of body volume (BV) can be used to estimate body composition using two‐ or multicomponent models. The traditional approach, underwater weighing (UWW), is awkward and unsuitable for many subjects. A newer alternative, whole body air displacement plethysmography (ADP), is less demanding but still unsuitable for young children, who may not remain still during the measurement. We have, therefore, considered whether a novel approach, three‐dimensional photonic scanning, is a viable alternative. Duplicate measurements of body volume were obtained in 22 adults (11 of each sex; mean [SD] BMI, 21.8 [2.5] kg/m2) by UWW, ADP, and a Hamamatsu Bodyline Scanner (HBS) (Hamamatsu, Japan). Subjects wore a tight‐fitting swimming costume for all three measurements, which were performed within one day of each other. Scans lasted 10 seconds, with the subject standing in a predefined position. The body surface skin was reconstructed using a B‐spline‐fitting model. In UWW, lung volume (LV) was measured simultaneously with underwater weight. In ADP and HBS, LV was predicted from weight and height. Results were compared using correlation and Bland and Altman analysis. Correlation analysis indicated that the scanner successfully ranked subjects in terms of BV. However, Bland and Altman analysis demonstrated that, relative to both UWW and ADP, HBS measured BV without bias but with limits of agreement between individuals of >2 liters, equivalent to approximately 20% fat. Scan precision was 0.57 liters, or 4.1% fat. Although HBS cannot yet measure BV with sufficient accuracy to predict fatness, much of the error is probably due to difficulties in standardizing LV during the scan. Simultaneous measurement of LV with volume by HBS is expected to improve limits of agreement substantially. Occlusion is also an important source of error. The method offers many advantages over alternative techniques, because the measurement is brief, noninvasive, and suitable for repeat measurements.

Background:Our study aims to compare the short and full-length International Physical Activity Questionnaires (IPAQ).Methods:Both versions were completed by 186 subjects >14 y living in southern Brazil. Half answered the short and then the long version; the remaining subjects followed the reverse order. Physical inactivity (PI) was defined as <150 min/wk spent in moderate or vigorous activities. The Bland and Altman method and the kappa statistic were used to assess agreement between the continuous and categorical outcomes, respectively.Results:The prevalence of PI was 50% higher with the short IPAQ (42% vs. 28%). The kappa value was 53.7%. Although the correlation coefficient was moderately high (r=0.61), agreement between methods was low.Conclusions:Both analyses used show that the short and full-length IPAQ versions have poor agreement. Utilization of inappropriate statistics would lead to misinterpretation. Researchers should exercise care before comparing studies using different IPAQ versions.