妊娠糖尿病與周產期憂鬱之相關性： 相輔性的資料分析研究

Abstract

背景：妊娠糖尿病與周產期憂鬱皆為孕產婦重要的健康議題。雖然非孕產期糖尿病與憂鬱症已證實具有雙向關係，但妊娠糖尿病與孕期及產後憂鬱之相關性目前仍缺乏一致結論。近年兩者盛行率皆呈上升趨勢，且在亞洲地區尤為顯著，其中臺灣數據高於多數國家，然針對亞洲華人，特別是臺灣孕產婦之研究仍相對不足。 目的：探討懷孕早期憂鬱和妊娠糖尿病的關係，及妊娠糖尿病和產後憂鬱的關係。 方法：本研究採世代研究設計，包含前瞻式問卷調查與回溯性病歷數據分析的整合式研究設計。問卷調查採前瞻性世代追蹤研究，共納入573名單胞胎初產婦，排除孕前已知糖尿病及既有精神科診斷者，於孕期第12~24週收案，並於第三孕期、產後一個月及產後三個月進行追蹤調查。次級資料分析採多中心回溯性世代研究，蒐集臺灣四間同體系醫療機構2009年1月1日至2022年12月31日之孕產婦產檢與生產病歷，排除非單胞胎、孕前糖尿病、既有精神科診斷、創傷性生產及病危或死亡個案，最終納入8,771人次生產資料。妊娠糖尿病以口服葡萄糖耐受試驗（oral glucose tolerance test, OGTT）作為診斷依據；憂鬱症狀以愛丁堡產後憂鬱量表（Edinburgh Postnatal Depression Scale, EPDS）評估，量表得分達10分以上者界定為具憂鬱傾向。以羅吉斯迴歸分析妊娠糖尿病與周產期憂鬱之相關性，並以Hayes之PROCESS模式檢驗調節與中介效果。 結果：在檢驗孕期憂鬱對妊娠糖尿病之影響時，結果顯示第二孕期憂鬱與妊娠糖尿病呈顯著負相關（aOR = 0.50，95% CI = 0.26~0.94，p = 0.03）。進一步分析發現，工作狀態對第二孕期憂鬱與妊娠糖尿病之關係具有顯著負向調節效果（β = −1.62，95% CI = −3.03 ~ −0.21，p = 0.02）。在檢驗妊娠糖尿病對產後憂鬱之影響時，次級資料（病歷資料）分析中，妊娠糖尿病與產後48小時內憂鬱呈顯著負相關（aOR = 0.60，95% CI = 0.47~0.76，p < 0.001）。前瞻式問卷資料結果顯示妊娠糖尿病與產後一個月及產後三個月憂鬱皆呈負相關，但未達統計顯著（產後一個月：aOR = 0.88，95% CI = 0.49~1.61，p = 0.68；產後三個月：aOR = 0.76，95% CI = 0.40~1.45，p = 0.41）；進一步以非早產婦女進行中介分析，發現產後一個月社會支持是妊娠糖尿病與產後三個月憂鬱之中介因子（β = −0.25，95% CI = −0.54 ~ −0.002）；妊娠糖尿病者在產後一個月自覺社會支持愈高（β = 2.61，95% CI = 0.03 ~ 5.19，p = 0.047），產後一個月社會支持愈高則產後三個月憂鬱愈低（β = −0.10，95% CI = −0.13 ~ −0.06，p < 0.001）。 結論：本研究發現，第二孕期愈有憂鬱症狀較不會有妊娠糖尿病，尤其在有工作者更是明顯。第二孕期憂鬱與妊娠糖尿病的關係，可能反映行為或生活型態差異（如飲食型態），但其潛在機制仍有待未來研究進一步釐清。另一方面，妊娠糖尿病者在產後48小時內較不憂鬱，在產後一個月和產後三個月亦有較不憂鬱的傾向，這可能與妊娠糖尿病者感受到較高的社會支持有關，尤其是在非早產或較無新生兒健康問題的群體。接受孕期生理疾病的照護可能有益於心理健康。孕產期心理症狀與代謝健康風險之關聯，可能受照護強度、工作狀態與社會支持所影響。除常規孕期生理檢查外，宜結合心理篩檢以作為啟動精準化孕產期照護與社會支持介入之重要切入點。

Background: Gestational diabetes mellitus (GDM) and perinatal depression are both major health concerns for pregnant and postpartum women. Although a bidirectional association between diabetes and depression has been well established in non-perinatal populations, evidence regarding the relationship between GDM and antenatal or postpartum depression remains inconsistent. In recent years, the prevalence of both conditions has been increasing consistently, particularly in Asian countries. Taiwan reports higher rates than many other countries; however, studies focusing on Asian Chinese populations, especially Taiwanese women, remain limited. Objectives: To examine (1) the association between early-pregnancy depressive symptoms and gestational diabetes mellitus, and (2) the association between gestational diabetes mellitus and postpartum depressive symptoms. Methods: This study adopted a cohort design integrating a prospective questionnaire survey with a retrospective medical record analysis. The prospective component enrolled 573 primiparous women with singleton pregnancies, excluding those with pre-existing diabetes or prior psychiatric diagnoses. Participants were recruited between 12 and 24 weeks of gestation and followed up in the third trimester, at 1 month postpartum, and at 3 months postpartum. The secondary data analysis comprised a multicenter retrospective cohort, including prenatal and delivery records from four affiliated medical institutions in Taiwan between January 1, 2009, and December 31, 2022. After excluding non-singleton pregnancies, pre-gestational diabetes, prior psychiatric diagnoses, traumatic deliveries, and critically ill or deceased cases, a total of 8,771 deliveries were included. GDM was diagnosed using the oral glucose tolerance test (OGTT). Depressive symptoms were assessed using the Edinburgh Postnatal Depression Scale (EPDS), with scores ≥10 indicating elevated depressive symptoms. Logistic regression was used to examine associations between GDM and perinatal depression, and moderation and mediation effects were tested using Hayes’ PROCESS models. Results: When examining the effect of antenatal depressive symptoms on GDM, second-trimester depressive symptoms were significantly negatively associated with GDM (adjusted odds ratio [aOR] = 0.50, 95% CI = 0.26–0.94, p = 0.03). Further analysis showed that employment status significantly moderated this association in a negative direction (β = −1.62, 95% CI = −3.03 to −0.21, p = 0.02). Regarding the effect of GDM on postpartum depression, analysis of secondary (medical record) data revealed a significant negative association between GDM and depressive symptoms within 48 hours postpartum (aOR = 0.60, 95% CI = 0.47–0.76, p < 0.001). In the prospective questionnaire data, GDM was negatively associated with depressive symptoms at 1 and 3 months postpartum, although these associations did not reach statistical significance (1 month postpartum: aOR = 0.88, 95% CI = 0.49–1.61, p = 0.68; 3 months postpartum: aOR = 0.76, 95% CI = 0.40–1.45, p = 0.41). Mediation analysis restricted to women without preterm birth indicated that perceived social support at 1 month postpartum mediated the relationship between GDM and depressive symptoms at 3 months postpartum (β = −0.25, 95% CI = −0.54 to −0.002). Women with GDM reported higher perceived social support at 1 month postpartum (β = 2.61, 95% CI = 0.03–5.19, p = 0.047), and higher social support at 1 month postpartum was associated with lower depressive symptoms at 3 months postpartum (β = −0.10, 95% CI = −0.13 to −0.06, p < 0.001). Conclusions: This study found that greater depressive symptoms in the second trimester were associated with a lower likelihood of gestational diabetes mellitus, an effect that was particularly pronounced among employed women. The association between depressive symptoms in the second trimester and GDM may reflect differences in health behaviors or lifestyle factors (e.g., dietary patterns), although the underlying mechanisms warrant further investigation. In addition, women with GDM exhibited lower levels of depressive symptoms within 48 hours postpartum and showed a trend toward lower depression at 1 and 3 months postpartum. This pattern may be partly explained by higher perceived social support among women with GDM, particularly in those without preterm birth or significant neonatal health problems. Engagement with medical care for pregnancy-related physical conditions may confer benefits to psychological well-being. Overall, the association between perinatal psychological symptoms and metabolic health risks may be influenced by care intensity, employment status, and social support. Beyond routine obstetric assessments, integrating mental health screening may serve as an important entry point for initiating precision-based perinatal care and targeted social support interventions.