憂鬱與新陳代謝指標及賀爾蒙之探討

Abstract

憂鬱與肥胖在不同的發展階段都具有相當的影響，乃是不容忽視的公共衛生議題，然而過去的研究對憂鬱與肥胖的關係缺乏一致的定論，對形成憂鬱與肥胖之關聯的原因更是不清楚。為釐清這些議題，這篇論文分別針對非臨床樣本的青少年與老年人，探討憂鬱與肥胖、新陳代謝指標、與賀爾蒙的相關，希望藉由三個獨立的研究，並同時收集心理與生理指標的設計，能夠有助於瞭解這些變項之間的複雜關聯。 研究一、青少年焦慮/憂鬱與新陳代謝指標的遺傳相關：多變項的雙胞胎與手足分析 背景：本研究擬探討青少年焦慮/憂鬱與新陳代謝指標的相關，並估計基因與環境因素的貢獻，並瞭解皮質醇(cortisol)的影響。方法：本研究從台北縣市的國中，招募年齡介於12-18歲的183對同性別雙胞胎，與30對手足配對進入研究。使用的測量工具包括以青少年行為檢核表來測量青少年的焦慮/憂鬱。新陳代謝指標則包括身體質量比、血糖值、胰島素、高密度膽固醇、與血壓。另外也測量皮質醇濃度。統計方法則採用結構方程模式來執行多變項遺傳分析。結果：55%青少年的體重為正常範圍，其中約有84%的青少年的希望有較輕的體重。另外，有38%青少年則屬於低體重分類，其中約有37%希望有較輕的體重。高焦慮/憂鬱分數與低身體質體質量比、低收縮壓、高濃度的高密度膽固醇有相關，也與較低的希望身體質體質量有關，但與皮質醇濃度關係不明顯。進一步的遺傳相關分析指出，焦慮/憂鬱分數與身體質體質量比的遺傳相關為-0.14、與希望身體質體質量比的遺傳相關為-0.18、與收縮壓的遺傳相關為-0.15、與藉由因素分析所得的代謝指標的遺傳相關則為-0.19。結論：青少年的焦慮/憂鬱與一些特定新陳代謝指標具有顯著的遺傳相關，但是幅度不大，而皮質醇濃度並不能解釋所觀察的這些相關。本研究結果可有助瞭解青少年焦慮/憂鬱與新陳代謝指標的關係。 研究二、瘦素與胰島素阻抗的遺傳相關，控制身體質量比的影響：多變項雙胞胎與手足分析 背景：瘦素經常與胰島素濃度、高血壓有關聯，而且這些相關在控制身體質量比的影響之後仍然存在。目前對這些相關瞭解甚少，且也不清楚基因與環境因素的貢獻。本研究擬探討瘦素與新陳代謝指標的相關，以及基因與環境因素之貢獻。方法: 採橫斷式的研究設計，針對從台北縣市的國中生中所收集的同卵雙胞胎(130對)、異卵雙胞胎(68對)、與手足配對(30對)進行研究。測量瘦素濃度與新陳代謝指標，包括身體質量比、血糖值、胰島素、與血壓，胰島素阻抗指標(HOMA-IR)則由模式估計而得。採用結構方程模式來執行多變項遺傳分析。結果：瘦素、身體質量比、血糖值、胰島素、胰島素阻抗指標、與血壓之間的遺傳相關高，介於0.25 與0.66之間。當控制身體質量比的影響之後，瘦素與胰島素、胰島素阻抗指標之間的遺傳相關減弱但仍統計顯著。然而，瘦素、胰島素、胰島素阻抗指標與血壓的遺傳相關則消失。多變項分析結果指出一共同基因同時調控瘦素、胰島素、與胰島素阻抗指標。結論：瘦素與胰島素、瘦素與胰島素阻抗指標之間有明顯的遺傳相關，本研究結果支持，在未來的多變項遺傳分析可以同時納入瘦素與胰島素濃度，以助於相關基因的搜尋。 研究三、憂鬱變化與肥胖的相關：台灣老年人的長期追蹤研究 背景：憂鬱與肥胖常見於老年人，但是憂鬱的變化與肥胖的關聯並不清楚。 本研究擬1)瞭解老年人的憂鬱變化與相關預測因子，2)探討憂鬱變化與肥胖、以及後來新陳代謝指標、與皮質醇濃度的相關。方法:採縱貫式的追蹤研究(1989至1999年)，樣本是3922位台灣的老年人。使用的測量工具是用Center for Epidemiologic Studies Depression Scale (CES-D)問卷來測得憂鬱分數。新陳代謝指標，包括身體質量比、血脂肪值、血糖值、與血壓，以及皮質醇濃度。統計分析採semi-parametric group-based modeling的方法進行憂鬱變化的分析，以Multinomial logistic regression and multivariable linear regression 瞭解憂鬱變化與基線特質及肥胖、新陳代謝指標的相關。結果：共有四組不同憂鬱變化的組別，其中第一類(持續低分)佔有41.8 %，第二類(持續中低分)佔有46.8 %，第三類(後期高分)佔有4.2 %，第四類(持續高分)佔有7.2 %。性別、教育程度、飲酒習慣、規則運動、慢性疾病、自評健康狀態是預測憂鬱變化的重要因子。憂鬱變化與身體質量比為負關聯，，其中身體質量比≧25.0的女性有較少機會是第二、三、四類憂鬱變化(OR分別為0.61、0.48、與0.19)，而身體質量比<18.5的男性, 有較高機會是第三、四類憂鬱變化(OR分別為3.24與4.14)。憂鬱變化與後來的血壓與皮質醇有正關聯。結論：老年人憂鬱變化呈現不同的組別，憂鬱變化與肥胖有負相關，但與後來的血壓、皮質醇正相關。本研究結果助於瞭解老年人憂鬱與肥胖、新陳代謝指標的關聯，老年人的憂鬱變化是值得重視的議題。

Depression and obesity are both increasingly prevalent in different developmental stages. The nature of the relationships between depression and obesity, and other metabolic variables, however, remains inconclusive. The etiology underlying the depression-obesity link is also largely not clear. In this dissertation, three studies were carried out in order to investigate the nature of the association between depression and metabolic function in adolescents, and older adults, respectively; and to examine the underlying determinants of the association. A special focus will be made on the hormonal factors, such as cortisol and leptin, to examine their role in the association. Such a combination of detailed psychological assessments and physiological measurement is expected to help shed light on the interrelation of psychological well-being and metabolic function. Study 1: Genetic Correlation between Anxious/Depression and Metabolic Risk Factors in Adolescents: A Multivariate Twin/Sibling Analysis Background: To examine whether anxious/depression was associated with metabolic risk factors in non-referred adolescents and determine the relative influence of genetic and environmental factors underlying the association. The role of cortisol on the association was also investigated. Method: In a sample of same-sex twins (n = 183 pairs) and sib-pairs (n = 30 pairs) aged 12-18 recruited from middle schools in Taipei, anxious/depression was assessed using the Child Behavior Checklist. Metabolic phenotypes including body mass index (BMI), levels of glucose, insulin, high-density lipoprotein cholesterol (HDL-C), and blood pressure as well as cortisol levels were measured. Desired BMI was also examined. Multivariate genetic analyses were conducted using structural equation modeling. Results: The majority of participants fell within the category of normal weight (55%) or underweight (38%), of which 84% and 37%, respectively, wished to lose weight as revealed by the disparity between actual and desired weight. Higher scores in anxious/depression were associated with lower levels in BMI, desired BMI, and systolic blood pressure as well as with higher levels in HDL-C and, with a borderline significance level, mid-morning cortisol. There were genetic correlations between anxious/depression and BMI (rG = -0.14), desired BMI (rG = -0.18), systolic blood pressure (rG = -0.15), and the metabolic factor (rG = -0.19) derived from factor analysis. Conclusions: There was a small but significant genetic association between anxious/depression and certain metabolic risk factors, which was unlikely explained by cortisol levels, in adolescents mainly of normal- or under-weight. These provide new insights regarding the etiology of both anxious/depression and metabolic profiles. Study 2: Genetic Correlations Between Leptin and Insulin Resistance Independent of Body Mass Index in Adolescents: A Multivariate Twin/Sibling Analysis Background: Leptin levels are frequently correlated with insulin levels, blood pressure, even after adjustment for body mass index (BMI). Nonetheless, the extent to which genes or environmental factors contribute to the covariation among these traits has not been fully understood. We aimed to determine the relative influence of genetic and environmental factors underlying the associations between leptin and metabolic traits. Methods: A cross-sectional twin study was performed in 2002. Participants were recruited from middle schools in Taipei. A sample of monozygotic twins (n = 130 pairs), dizygotic twins (n= 68 pairs), and sib-pairs (n = 30 pairs) aged 12-18 was studied. Serum leptin levels and metabolic phenotypes including BMI, levels of fasting insulin, fasting glucose, and blood pressure were measured. Insulin resistance was determined by homeostasis model assessment (HOMA-IR). Multivariate genetic analyses were conducted using structural equation modeling. Results: Leptin, BMI, insulin, HOMA-IR, and systolic blood pressure tended to be genetically correlated with each other, with genetic correlation ranging from 0.25 to 0.66. After adjusting for BMI, the positive genetic correlations of leptin with insulin levels and HOMA-IR were attenuated but remained significant, while those of leptin, insulin, and HOMA-IR with systolic blood pressure disappeared. Multivariate modeling identified a common genetic factor influenced leptin, insulin, and HOMA-IR. Conclusions: Substantial genetic correlations between leptin and insulin as well as between leptin and insulin resistance were found. These results provide empirical evidence to include both leptin and insulin for future multivariate genetic analyses of metabolic function. Study 3: Depression trajectories and obesity in a longitudinal study of elderly in Taiwan Background: Depression and obesity are common in older adults. However, the course of depression changes over time and their relationship with obesity are not clear. We aimed to 1) characterize trajectories of depressive symptoms and identify predictors of trajectory classes; 2).determine the association between these trajectories and obesity as well as subsequent metabolic function and cortisol levels. Method: A prospective cohort study of older Taiwanese adults (n = 3922) was carried out between 1989 and 1999. Depression was assessed using the Center for Epidemiologic Studies Depression Scale (CES-D). Metabolic variables including body mass index (BMI), lipid profile, levels of glucose, and blood pressure as well as cortisol levels were measured. Trajectory analyses were conducted using semi-parametric group-based modeling. The associations between depression trajectories and baseline characteristics, and BMI categories were examined using multinomial logistic regression. Results: Four distinctive trajectories of depressive symptom were identified: class 1 (“persistent low”; 41.8 %), class 2 (“persistent mild”; 46.8 %), class 3 (“late peak”; 4.2 %), and class 4 (“high-chronic”; 7.2 %). Gender, educational level, regular exercise, chronic disease, and self-assessed health predicted development of trajectory pattern. The pattern of depression trajectory was inversely associated with BMI. Women with BMI≧25.0 were less likely to be in class 2, 3, and 4 (OR = 0.61, 0.48, 0.19 respectively). The odds for men with BMI <18.5 to develop class 3 and 4 were 3.24, and 4.14, respectively. High depressive symptoms were positively linked to subsequent high blood pressure and high levels of cortisol. Conclusions: There existed distinct classes of depressive symptoms changes over time. An inverse association for depression and BMI, while a positive association for depression and high blood pressure was found. These findings may be of interest to health professionals who wish to target depression and obesity to promote the well-being in late life.