Validity of home-measured height, weight and waist circumference among adolescents

This study assesses the validity of home-measured height, weight and waist circumference among Finnish adolescents from the Fin-HIT cohort. The adolescents were measured by fieldworkers, and were instructed to measure themselves at home with an adult’s assistance. Paired t-test was used for statistical analyses. Home-measured mean height, weight and waist circumference were slightly higher, but BMI lower than measured by the fieldworker. The difference in means was statistically significant for weight (0.51 kg) and waist circumference (1.6 cm), but not for height and BMI. Home-measured height, weight, waist circumference and BMI are sufficiently accurate to be used in epidemiologic studies.

Introduction

The prevalence of overweight and obesity among children and adolescents worldwide has risen by almost 50% during the past three decades.¹ Identification and early intervention of overweight and obese children and adolescents are essential to health policy. In large population surveys, direct measurement of height and weight is often not feasible due to high costs. Instead, data is commonly collected by self-reports. However, self-reported data often underestimates overweight prevalence among adolescents, as reviewed by Sherry et al.². Several studies have shown limited validity of self-reported height and weight among children younger than 14 years.^3–6

It is recommended to use waist circumference measurements besides body mass index (BMI) in order to get a more comprehensive measure of overweight, and especially obesity-related health risks among children and adolescents.^7,8 Only one validation study on self-reported waist circumference among adolescents has been published so far.⁹ In that study, self-reported values were on average higher than measured. However, the adolescents were not given any instructions on how to report waist circumference, and only the younger children attending primary school were asked to seek help from their parents/guardians.

To our knowledge, there are no published studies about the validity of self-reported measurements, when adolescents are asked to measure and weigh themselves. In this study, we report on the validity of home-measured compared to objectively measured height, weight and waist circumference.

Methods

The study population included participants from the prospective Fin-HIT - Finnish Health in Teens cohort, consisting of approximately 11 000 adolescents and 10 000 guardians from South, Middle and North Finland, recruited in 2011–2014 (www.finhit.fi/for-researchers). An informed consent was obtained by the adolescents and one of their guardians. The Coordinating Ethics Committee of the Hospital District of Helsinki and Uusimaa approved the Fin-HIT study.

A sample of 115 adolescents (62 boys, 53 girls; 68% of all invited) aged 9–11 years participated in the validation study in December 2014. The adolescents had their height, weight and waist circumference measured in school in a standardized way by specially trained fieldworkers. The adolescents were asked to remove their shoes, heavy sweaters and pocket items. The fieldworkers reported all the clothes that the adolescents were wearing, and the weight of the clothes was subtracted from the measured weight to form the real weight. Height was measured to the nearest 0.1 centimetre with a portable stadiometer (Seca model 217), weight to the nearest 0.01 kg with portable digital scales (CAS model PB) calibrated before each series of measurements, and waist midway between the hip bones and the ribs to the nearest 0.1 cm with a measuring tape calibrated against a measure once a week. The fieldworkers reported all measurements on a tablet computer using a bar code linked to each participant. About 30% (n = 81) of the measurements were made before lunch, and 70% (n = 33) after lunch, and one child had missing information on the time of measurement. BMI was calculated as weight (kg)/height (m)² and the adolescents were classified as underweight, normal weight, overweight or obese according to age and gender specific cut-offs.¹⁰

After the measurements at school, the adolescents received a measuring tape and written detailed instructions, including pictures, on how to measure and report height, weight and waist circumference at home with assistance from an adult. The instructions corresponded to the measurement protocol in school. If the adolescents did not have scales at home, they were advised to weigh themselves in the school nurse’s office or at a friend’s home. An additional question served as a check question for the weight measurement: ‘When did you weigh yourself?’ with the answer alternatives ‘I weighed myself today’, ‘I weighed myself within the last month’ and ‘I guessed my weight’. The date of the answer was recorded. No additional check questions were asked for height and waist circumference.

Data was analysed using R. The extreme values more than 3 standard deviations (SD) of the mean were removed, thus the final sample size was 113 (62 boys, 51 girls). Paired t-test was used to determine differences in home-measured vs. objectively measured height, weight, waist circumference and BMI, separately for boys and girls. For the measurements without a normal distribution, we first log transformed the data and then we applied the paired t-test on the transformed measurements. Shapiro test was used to check for normal distribution, and both absolute and log transformed values were tested. Variance tests for all measures were also performed. Finally, we tested if home-measured and objectively measured height and weight resulted in different BMI categories defined by Cole et al.¹⁰

Results

Home-measured mean height, weight and waist circumference were slightly higher, but derived BMI lower than measured by the fieldworker. However, the differences were so small that they had an impact on BMI category only in two normal weight adolescents, who were classified at home as underweight and overweight, respectively. Height was normally distributed, while the other measures were not. The absolute difference in mean (for descriptive purposes) and absolute and log transformed p values are reported. The difference in means was significant for weight (0.51 kg; P < 0.001) and waist circumference (1.6 cm; P < 0.001), but non-significant for height (0.3 cm; P = 0.06) and BMI (0.07; P = 0.071) (Table 1). The same trend was observed in both genders without evidence for statistically significant heterogeneity between boys and girls. There was no significant difference in variance for any of the measurements, nor deviance from the normal distribution after log-transformation of the studied variables.

Discussion

To our knowledge, this is the first study to test the validity of home-measured height, weight and waist circumference among adolescents. We found small, statistically significant differences between home-measured and objectively measured weight and waist circumference. These differences could likely be explained by the use of inexact and non-calibrated scales at home. Moreover, at school the adolescents had their body measurements taken during the day while at home the measurements could have been taken any time, also in the evening. As measurements at home were made after measurements at school, it is also likely that the adolescents had increased in weight.

Only two adolescents reported having guessed their weight. Waist circumference is challenging to be measured at home, as the measuring tape might not have been tight enough around the waist, which explains our results.

The strengths of our study were the detailed measurement instructions corresponding to the school measurement protocol. We assumed that many of the adolescents were measured by an adult, as was stated in the instructions.

One limitation of the study is that the adolescents reported their self-measurements after been measured at school. However, the adolescents were only informed about their measurements at school if they asked for them. Information about which adolescents requested to know their weight was not reported. Most adolescents did not weigh themselves on the same day as they were weighed at school. Another limitation is the participation rates (68%), which do not allow us to rule out potential selection bias. However, internal validity of the results should not be affected.

Our results indicate that home-measured height, weight, waist circumference and derived BMI are sufficiently accurate to be used in epidemiologic studies in early phases of adolescence in both sexes in Finland. As home-measurements are substantially less expensive than fieldworker performed measurements, our results may have implications in the planning and execution of future studies on anthropometrical measures in adolescents in similar settings.

Acknowledgements

Mr Jesper Skand for data management. Dr. Rejane Figueiredo for statistical data analysis.

Funding

This work was supported by Folkhälsan Research Foundation; Academy of Finland [grant number 250704]; Life and Health Medical Fund [grant number 1-23-28]; The Swedish Cultural Foundation in Finland [grant number 15/0897]; Signe and Ane Gyllenberg Foundation [grant number 37-1977-43]; and Yrjö Jahnsson Foundation [grant number 11486].

Conflicts of interest: None declared.

References