過重與非過重青少年健康相關行為的比較

THE COMPARISON OF HEALTH-RELATED BEHAVIORS BETWEEN OVERWEIGHT AND NON-OVERWEIGHT ADOLESCENTS

陳美燕(Mei-Yen Chen)

指導教授 ： 黃璉華

國立臺灣大學/醫學院/護理學研究所/博士(2005年)

https://doi.org/10.6342/NTU.2005.01241

摘要

過重與非過重青少年健康相關行為的比較
背景：青少年過重問題在已開發及開發中國家有逐漸上升的趨勢，然而過去的研究並沒有從文化環境及健康促進相關行為對青少年過重的影響做探討。 過重或肥胖是指體內貯存過多脂肪的現象，大多數的文獻指出肥胖的原因來自基因遺傳和環境的影響，雖然促成肥胖的原因相當複雜，但最終都與能量的攝入及消耗不平衡有關。事實上至今科學上對能量的攝取或活動的改變以及其他相關因素，如何影響肥胖的機轉仍然不夠清楚。
研究目的：本研究主要目的有：(一)以因素分析和集群分析法測試青少年健康促進量表在過重與非過重群體的概念建構及區辨效度；(二)比較過重與非過重青少年在健康促進及相關行為的差異；(三)探討影響青少年過重的相關影響因素。
研究方法：本研究主要的概念架構是參考護理學者 Orem的自我照顧缺失理論以及Pender的健康促進模式，以引導本研究在探討不同體位青少年之日常生活中，實踐健康促進及其相關行為的自我照顧能力及其環境的影響。研究樣本共660位 (男生351，女生309)，主要來自桃園縣五所國中和兩所高中，其中四所學校屬於郊區、三所屬於都會型地區。研究工具包括青少年健康促進量表、身高、體重計以及基本人口學資料。統計分析方法主要採用類別資料分析、因素分析、集群分析、區變函數分析、以及對數線性模型和邏輯模型檢定。
研究結果：從KMO及Bartlett球形檢定顯示，樣本的選擇適合採行因素分析的分配；因素分析萃取出六個構面的健康促進概念，共可以解釋50.6%的變異量，包括生命欣賞、社會支持、健康責任、營養、壓力處置及運動等行為，其內在一致性信度以Cronbach’s alpha測得0.936。階層集羣分析顯示青少年健康促進量表可以顯著的將過重與非過重青少年，區分成健康型與不健康型兩個集群；非過重青少年顯著的出現在健康型的群組，過重青少年則分布在非健康型的群組，區辨函數分析也支持集群分析所作的兩組分類。此外，過重青少年在執行健康促進行為的頻率，明顯的低於過重組，例如，運動、壓力處置、生命欣賞、健康責任及社會支持等行為。同時過重青少年在看電視及使用電腦的時間也都顯著高於非過重組；而看電視或用電腦的時間越長，其健康促進行為得分就越低。在健康狀態方面，雖然看病次數在兩組間無統計差異，但是過重青少年在自覺健康狀態上顯然比較差。除了健康促進行為對過重有影響外，其他影響青少年過重的基本人口學因素還包括：父母是過重的體位、居住在都市地區及宗教信仰為非基督教相關者、性別、年級、看電視及使用電腦時間；父母的教育程度、家庭結構、族群則無統計上的差異。以對數線性模型進一步做參數估計，發現看電視的時間、居住地區、健康促進行為多寡與青少年過重有直接的影響；而邏輯模型則顯示非過重者少看電視時間、居住於郊區和較多健康促進行為頻率，相較於過重者其相對風險分別是1.6、1.4 和1.3倍的機率。
結論：本研究結果多數支持本研究的假設，亦即青少年健康促進量表具有適當的建構和區辨效度，它可以將所有個案區分成健康型與非健康型的概念，而過重青少年在健康促進行為的實施頻率顯著低於非過重組，同時過重個案也顯著被分類在非健康型的集群裡。研究顯示與青少年過重有關係的重要因素分別是(1)看電視時間；(2)住在都會區；(3)缺乏健康促進行為。這些結果將有助於發現影響過重青少年的相關因素，並可以作為發展改善我國青少年過重問題的政策；針對基層公共衛生人員及校園健康促進人員的角色功能，文中亦有詳細討論可行的策略。

關鍵字

集群分析 ； 邏輯線性模型 ； 類別資料分析 ； 使用電腦 ； 看電視 ； 健康促進行為 ； 過重青少年

並列摘要

Background: Adolescent overweight is increasing all over the developed and developing countries. However, previous studies have generally not looked at the impact of cultural elements and health promotion behaviors on adolescents’ body size. Overweight or obesity is defined as the presence of excess adipose tissue. Most of the literature on etiology of obesity addressed biologic and environmental factors. Although the cause of overweight is complex and involves both genetic and environmental determinants, overweight essentially results from an imbalance of energy intake via diet and energy expenditure. Until now, it is not clear how changes in dietary intake or activity levels and other factors have contributed to recent increases in obesity.
Aims: The purpose of this study was to examine the following questions: (1) to test the construct and discriminated validity of Adolescent Health Promoting behaviors between the overweight and non-overweight adolescents through factor and cluster analysis; (2) to compare the differences of adolescent health promoting behaviors among the overweight and non-overweight adolescents; (3) to find out the factors that associated with the overweight/non-overweight adolescents’ health-related behavior and their body sizes.
Methods: The conceptual framework for this study built on Orem’s self-care deficit theory and Pender’s health-promoting model. Both of these theories lead this study to explore the differences of daily self-care ability in practicing health promoting behaviors and their environmental influences among different body size adolescents. The study was conducted in five junior middle schools and two high schools in Tao-Yuan County, Taiwan. Four of them were located in rural areas, the other three were in urban areas. The study samples consisted of 660 subjects, 351 boys and 309 girls. The instruments included the AHP scale, the height and weight scales and a demographic data sheet. Hypothesis was tested using categorical data analysis, factor analysis, cluster and discriminated analysis, and SAS CATMOD procedure of log-linear and logit model testing.
Results: Findings indicated that AHP was a valid and reliable scale to measure the health promoting behaviors between overweight and non-overweight subjects. Kaiser-Meyer-Olkin (KMO) measures and Bartlett’s sphericity test showed that the samples met the criteria for factor analysis. Factor analysis yielded a six-factor instrument, which explained 50.62% of the variance in the 40 items. The six factors were social support, life appreciation, health responsibility, nutritional behaviors, exercise behaviors and stress management. The Cronbach’s alpha reliability coefficient for the total scale was 0.936. Hierarchical and k-means cluster analyse revealed two distinct patterns of these non-overweight and overweight adolescents: healthy and unhealthy groups. The result of discriminant analysis also supported the two clusters of AHP at construct and discriminant validity scale that the non-overweight adolescents tended to be healthy group and overweight adolescents tended to be the unhealthy group. Overweight adolescents practiced in general significantly less frequent of practicing health-relate behaviors such as exercise behavior, stress management, life appreciation, health responsibility, and social support. While in comparison of the time spent on watching TV and using computer, they gained significant of higher frequency than their counterpart. Interestingly, the longer the time spent on watching TV or using computer, the less they practiced health promoting behaviors. Further the overweight adolescents tended to claim worse health status. Besides for health promoting behavior, other factors related to the adolescent overweight were having overweight parents, male, 8th grader, living in urban area, and with non Christian-related religion background. However, significantly relationship between parental educational level, family structure, ethnics and adolescent’s body size was not found. The log-linear model suggests that there is a direct association between TV watching, living location, health-promoting behavior, and body size. The relative likelihood of being non-overweight in those watching less TV was 1.6 times more than in those overweight watching TV, 1.4 times more likely in those living in a rural area than in those from an urban area, and 1.3 times more likely in those practicing healthy behaviors.
Conclusions: This findings support the hypotheses that adolescent health promoting scale could discriminate participants into two clusters among overweight and non-overweight adolescents. Overweight adolescent was significantly associated with less health promoting behavior. The strongest variables associated with adolescent overweight were (1) a high level of TV watching, (2) living in an urban area, and (3) less health-promoting behavior. These findings can be used to help professionals further develop and refine the knowledge and strategies for adolescent overweight management. Suggestions for professionals in primary health care, ethical and methodological issues were discussed in this paper as well.

並列關鍵字

overweight adolescents ； health promoting behaviors ； watching TV ； categorical analysis ； using computer ； cluster analysis ； log-linear model ； logit model

參考文獻

Peng, C. J., Chang, C. m., Kuo, S. E., Liu, Y. J., Kuo, H. C., & Lin, S. J. (1999). Analysis of anthropometric growth trends and prevalence of abnormal body status in Tainan elementary-school children. Acta Paediatr Tw, 40(6), 406-413.

Backman, K., & Hentinen, M. (1999). Models for the self-care of home-dwelling elderly. Journal of Advanced Nursing, 30(3), 564-572.

Ball, K., Owen, N., Salmon, J., Bauman, A., & Gore, C. J. (2001). Associations of physical activity with body weight and fat in men and women. International Journal of Obesity, 25(6), 914-920.

Barnes, M., Courtney, M. D., Pratt, J., & Walsh, A. M. (2004). School-based youth health nurses: Roles, responsibilities, challenges, and rewards. Public Health Nursing, 21 (4), 316.

Blackie, C., & Gregg, R. (1998). Promoting health in young people. Nursing Standard, 12(36), 39-46.