跳到主要內容區
  1. 首頁
  2. 歷屆論文
  3. 第二屆論文

台北市0-6歲兒童生長常模之建立及肥胖盛行率之分析

論文名稱(中/英文) 研究生 指導教授

台北市0-6歲兒童生長常模之建立及肥胖盛行率之分析
Establishing a growth norm for 0-6-year-old children and analyzing the

prevalence of obesity in Taipei City

 李碧慧 黃奕清博士
[  摘要  ]
身體發育測量是國民健康的基本資料，由兒童身高、體重的資料，可瞭解其生長發育情形，並作為評估營養狀況的重要依據，亦可用以發現體重過重、肥胖或發育不良之個案，俾早期預防、輔導與治療，以奠定健康基礎。
本研究總樣本數為47,474人，其中男生為24,687人（52%），女生為22,787人(48%)。資料分析求出臺北市0~6歲兒童每一年齡層之身高、體重、頭圍（0~2歲）之平均值（mean）、中位數(median)、標準差（standard deviation）、眾數（mode）、最大值(max)、最小值(min)及百分位數（percentile;3rd, 5th, 10th, 15th, 25th, 50th, 75th, 85th, 90th, 95th, and 97th）。以及身體質量指數（body mass index, BMI）、重高指數（weight for length index, WLI）、超出平均體重、身高別體重等變項之平均值、中位數、標準差及百分位數。
    本研究以身高別之體重大於（中位數加二個標準差），估算各年齡層肥胖之百分比；以BMI≧第85百分位數為過重，BMI≧第95百分位數為肥胖；以重高指數計算其盛行率1.2>WLI≧1.1者為過重的界定點，WLI≧1.2者為肥胖；超出平均體重10~20%為過重，超出平均體重20%以上（含20%）為肥胖。並以Pearson相關係數分析身體質量指數、重高指數、超出平均體重與年齡、身高、體重的積差相關。以Kappa一致性分析身體質量指數、重高指數及超出平均體重等指標篩選肥胖的一致性。同時為使各原始百分位數曲線能為平滑曲線，以多次方迴歸方程式（regression models with polynomials）加以修正，在R-square大於0.99才停止累加次方。
    研究結果顯示無論性別或年齡，重高指數與超出平均體重的一致性最高，兩者篩選出的過重及肥胖盛行率皆高於身體質量指數。身體質量指數與超出平均體重的一致性最低。當以身體質量指數（BMI）為篩選指標時，過重盛行率（BMI≧第85百分位數）平均為10.1%、肥胖盛行率（BMI≧第95百分位數）平均為5.1%；以重高指數（WLI）為篩選指標時，過重盛行率（1.2>WLI≧1.1）平均為12.7%、肥胖盛行率（WLI≧1.2）平均為5.5 %；以超出平均體重為篩選指標時，過重盛行率（超出平均體重10~20%）平均為12.3%、肥胖盛行率（超出平均體重20%以上）肥胖盛行率平均為7%；以身高別之體重中位數＋2SD，肥胖盛行率平均為3.5%。
    三種肥胖篩檢指標各有其特色，重高指數與體重有最大的相關，但卻也與身高有很高的相關；超出平均體重計算容易且與體重有最大的相關（r =0.35），然而易受到身高的影響；身體質量指數與身高有最低的相關，但卻也與體重有較低的相關。
[ Abstract ]
Measuring the growth of body is the basic information of people’s health. From the data of children’s height and weight, we can understand the state of their growth and evaluate their nutrient condition. In addition, we can discover the cases of overweight, obesity or dysplasia in advance to prevent, give advice or cure in early stage to establish the foundation of people’s health.The total investing samples are 47,474 children, in which there are 24,687 (52%) boys and 22,787 (48%) girls. Based on the data, we analyzed the mean, median, standard deviation, mode, max, min and percentile (3rd, 5th, 10th, 15th, 25th, 50th, 75th, 85th, 90th, 95th and 97th) of height, weight (of each age level from 0 to 6-year-old) and head circumference (of 0-2-year-old) of children in Taipei. We also analyzed the mean, median, standard deviation and percentile of body mass index (BMI), weight for length index (WLI), average of overweight and weight of different height.
    In this research, the percentage of obesity in each age level by is estimated while weight of different height is over median plus two standard deviations. When BMI ≧the 85th percentile, it is regarded as overweight and when BGI ≧the 95th percentile, is obesity. By WLI, we calculated the prevalence that 1.2 ＞WLI ≧1.1 is the boundary point, which WLI ≧1.2 is considered as obesity. When weight is over mean by 10-20%, it is considered as overweight and over 20% (including 20%) is considered as obesity. We also analyzed the product-moment correlation of BMI, WLI, average of overweight and age, height and weight by Pearson’s correlation coefficient. We analyzed BMI, WLI, average of overweight by Kappa coefficient of agreement to select the agreement of obesity. In order to make the curve of original percentile smooth, it is adjusted by regression models with polynomials until R-square is over 0.99.
    The result shows that, no matter age or gender, WLI has the highest correspondence to average of overweight; the prevalence of overweight and obesity selected by both of them are higher than BMI. BMI has the lowest correspondence to average of overweight. When BMI is taken as the selecting index, the average prevalence rate of overweight (BGI ≧the 85th percentile) is 10.1%, and the average prevalence rate of obesity (BGI ≧the 95th percentile) is 5.1%. When WLI is taken as the selecting index, the average prevalence rate of overweight (1.2 ＞WLI ≧1.1) is 12.7%, and the average prevalence rate of obesity (WLI ≧1.2) is 5.5%. When average of overweight is taken as the selecting index, the average prevalence rate of overweight (over mean by 10-20%) is 12.3%, and the average prevalence rate of obesity (over 20%) is 7%. When taking the median of weight of different height + 2SD as the index, the average prevalence rate of obesity is 3.5%.
    Each obesity selecting index obtains its own characteristics. WLI is most related to weight, but it also highly related to height. It’s easy to calculate the average of overweight and it is most related to weight (r = 0.35), but it is likely to be influenced by height. BMI is lowest related to height, but it’s also lower related to weight.