METHODS

Children aged 7–18 years with primary obesity were recruited from our obesity and lipid disorder clinic. All obese patients presenting at the clinic for assessment and advice on healthy living were eligible, except those with obesity secondary to an organic condition, a chronic cardiorespiratory problem, skeletal abnormalities, or congenital defects. Our study was reviewed and approved by the Ethics Committee of the Chinese University of Hong Kong. Informed consent from the subjects and their parents was obtained at the beginning of the assessment. In this study an obese subject was defined as one whose body mass index (BMI) was above the 95th centile according to our local sex and age specific BMI reference range.¹³

Each child was examined by one of the two investigators (AL, DC) to assess their cardiopulmonary status. Their weight (kg) and standing height (cm) were measured with a calibrated weighing scale and stadiometer by standard anthropometric methods.¹⁴

The lung function tests were carried out by the same team of technicians according to the recommended standard.¹⁵ Spirometry was measured by a SensorMedics 2130 spirometer using the Enhanced Spirometry Program. The best of at least three technically acceptable values for forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), maximum mid-expiratory flow rate (MMEFR), and flow-volume curves were selected. Functional residual capacity (FRC), residual volume (RV), and total lung capacity (TLC) were measured by body plethysmography (SensorMedics 6200) and expressed in litres corrected for body temperature, atmospheric pressure, and saturation with water vapour. Diffusion capacity of carbon monoxide was measured by single breath technique and the values obtained were corrected for haemoglobin concentration (SensorMedics 6200 Autobox DL, Single Breath Diffusion Capacity DLco SB Program).¹⁶ The pulmonary function results were expressed as percentages of predicted normal values.¹⁷ For the purpose of this study, the threshold of abnormality was identified as under 80% of the predicted value. Restrictive disease was defined as a reduction in TLC to less than 80%; diffusion impairment as a reduction in DLco to less than 80%; and obstructive airway disease as reduced FEV₁/FVC ratio to less than 80%. The various pulmonary deficits were classified as “mild” (>70%), “moderate” (<70% and >60%), “moderately severe” (<60% and >50%), and “severe” (<50%).¹⁸

DEXA measurements were performed with a total body scanner (QDR4500A; Hologic, Waltham, MA, USA) using array mode. This equipment used a switched pulse stable dual energy x ray operating at 100 and 140 kV. An automatic internal reference system with a calibration wheel achieved the bone density measurement. Data were analysed using software version V8.26a:3. The equipment determined whole body as well as regional measurements of bone mass, lean mass, and fat mass. Total fat mass was expressed in kg and as a percentage of body mass. These were calculated by integrating the measurements for the whole body and different automatic default regions as arms, trunk, and legs. The trunk region consists of the area bordered by a horizontal line below the chin, vertical borders lateral to the ribs, and oblique lines passing through the femoral necks. The leg region included all tissue below these oblique lines. The measurements used for correlation assessment were subtotal (total − head) and trunk body fat as a percentage of body mass. The total radiation dose per scan was <1 mSv.

Statistical methods

Percentage predicted was calculated for each respiratory impairment. Median and interquartile range were presented for various interval parameters, including anthropometric measurements and pulmonary function parameters. The BMI values were converted to exact Z scores from the three smooth age specific curves L (lambda), M (mu), and S (sigma) values derived using the LMS method, with the formula:

where the M and S curves correspond to the median and coefficient of variation of BMI at each age, and the L curve allows for the substantial age dependent skewness in the distribution of BMI.^13,19 Spearman’s rank correlation coefficient was used to assess the association between the DEXA measurements, BMI standard deviation score, and the various pulmonary function parameters. SPSS for Windows statistical software (release 10.1, SPSS Inc., Chicago, IL) was used in the analyses. The level of significance was set at 5% in all comparisons, and all statistical testings were two sided.

RESULTS

Sixty four children (16 girls, 48 boys) with primary obesity were recruited. None had evidence of cardiopulmonary disease on history review or clinical examination. Their median age was 12 years (interquartile range (IQR) 10–14) and the median BMI and standard deviation score were 30.1 kg/m² (IQR 27.2–32.8) and 2.42 (IQR 2.13–2.66) respectively. Table 1 shows the demographic characteristics and the results of spirometric, lung volume testing, diffusion capacity, and DEXA scan of the study population. All subjects underwent spirometric and lung volume testing but only 62 had diffusion capacity assessment. Thirty children (46%) were found to have reduced FRC. Abnormal diffusion capacity was detected in 21 subjects (33% of the study population). The deficit was of moderate severity. Three children had obstructive abnormalities documented on their lung function. They were of mild severity and all had a BMI greater than 34 kg/m².

Table 2 shows Spearman’s correlation between the various parameters. A significant correlation was found between BMI standard deviation score and trunk/subtotal body fat (r = 0.48 and r = 0.51 respectively, p < 0.01). BMI was not correlated to any of the pulmonary abnormalities. We were however, able to show significant negative correlation between both trunk and subtotal body fat with FRC, TLC, and RV but not with Dlco.

DISCUSSION

In our cohort of children with primary obesity, we found reduction in FRC to be the commonest pulmonary abnormality. We were also able to show a statistically significant negative correlation between this abnormality and the degree of obesity as assessed by DEXA scan. Such a relation was however, not found when BMI was used as a marker of obesity. Previous studies in obese children were also unable to establish a relation between pulmonary deficit and anthropometric assessments of obesity.^9,10,20 In fact the use of anthropometric and skinfold measurements has been criticised as being unreliable and inaccurate; they are unable to adequately assess visceral adiposity and are liable to operator bias.^21–25 DEXA scan is a more objective and reliable method of assessing adiposity. Using this technique, a total body scan is obtained with dual beam x rays of relatively low energy. From the absorbed energy at the two energy levels, different aspects of body composition can be calculated. This method requires little cooperation of the subject and involves minimal radiation. The procedure can be performed relatively quickly (scanning time 10–25 minutes) and the results are immediately available. This method allows total body fat to be better assessed, and its use in children has been validated and confirmed to be accurate and precise.^21–25

Classically obesity does not affect spirometric values other than maximum voluntary ventilation (MVV) until vital capacity becomes reduced in extreme cases. It has been shown that severely obese adults, with BMI of 45 or above, had a reduction in FVC and FEV₁ compared with age matched controls. They were also found to have increased RV and RV:TLC ratio, and a higher TLC by plethysmography than by helium dilution, indicating air trapping.²⁶ In our series of 64 patients, only three had evidence of obstructive pulmonary deficit and in all cases their BMI was more than 34 kg/m², representing the more severe end of obesity among our study population. This finding supports the notion that obstructive deficit becomes a problem in the group with severe obesity.

Diminished DLco was also found to be a common abnormality in our study population. This may reflect intrinsic changes within the lung in the presence of obesity. Lipid deposition, cellular hyperplasia, alveolar enlargement, and reductions in alveolar surface area relative to lung volume occur in the lung of immature rat made obese by dietary manipulation.²⁷ The reduction in DLco seen in our group of children may suggest structural changes in the interstitium of the lung resulting from lipid deposition and/or decreased alveolar surface area. Interestingly, we found the DEXA scan result to have a negative correlation with static lung volumes but not with DLco. It is possible that reduction in static lung volumes in obese individuals is simply a mass effect which DEXA scan is capable of determining accurately, whereas deficit in DLco is more complex and may involve more than one pathophysiological mechanisms.^27,28

There are certain limitations to our study. Firstly, there was no control group. Although data from a control group would provide further confirmation of the abnormalities detected, we considered it unreasonable to subject normal children to unnecessary radiation. Secondly, our study was a cross sectional assessment of lung function at one time point. To better understand the progress and natural history of the abnormalities identified in this cohort, it would be ideal to undertake a longitudinal study to assess the changes in lung function in relation to body composition over time.

In summary, we found pulmonary abnormalities to be common among obese children, and that reduction in static lung volume was found to be significantly correlated with the degree of obesity. The clinical significance of such a finding in this group of asymptomatic obese children remains unknown. It is important to assess the longitudinal changes and tracking into adulthood of such abnormalities in order to provide appropriate management.