具心血管危險因子之中老年人中樞自主網路內功能性連結與心率變異度及抑鬱情緒間之關係：靜息態磁振造影研究

Abstract

背景與目的: 有心血管危險因子者常有自主神經系統失衡的現象，而此失衡現象常透過較小的心率變異度和抑鬱情緒顯現出來。中樞自主網路與自主神經系統和邊緣系統功能的調節有關。本研究的目的是探討在有心血管危險因子的中老年人中，中樞自主網路內的功能性連結是否會與無心血管危險因子者不同，以及此連結之強度是否與其心率變異度和情緒抑鬱程度相關。 方法: 本研究收錄104 名認知正常且無精神疾病或抑鬱症的中年人（45-64 歲）和老年人（65-80 歲）的資料。資料來源為三個臨床隨機對照試驗（NCT編號 : 02270320、03275038 或05672940）的前測資料。其中，74名（平均年齡：64.0 ± 7.2歲）屬於心血管危險因子組，並且至少有以下三種心血管危險因子中的至少一種——高血壓、糖尿病或血脂異常。其餘30名（平均年齡：64.8 ± 6.6歲）為無心血管危險因子組。所有參與者都接受了蒙特利爾認知評估、老年抑鬱量表（GDS-15）評估和腦部靜息態磁振造影掃描。心血管危險因子組還進行了5分鐘的休息時心率變異度測試。使用的心率變異度參數為正常到正常 R 波間距秒數的標準差（SDNN）、高頻功率（HF）、低頻功率（LF）和低頻對高頻功率之比值（LF/HF）。較高的SDNN和HF數值表示副交感神經控制較強；而較高的 LF和LF/HF表示交感神經控制較強。中樞自主網路中的八個腦區為本研究之興趣腦區，含右側框額皮質、左側框額皮質、右側腦島、左側腦島、內側前額葉皮質、前扣帶皮質、右側杏仁核與左側杏仁核。以 CONN toolbox分析網路內此八個區域間兩兩之連結並用Fisher’s Z 轉換其連結強度數值，連結強度閾值設置為 Family wise error (FWE), p < 0.05。使用獨立樣本 t 檢定比較兩組間中樞自主網路內功能性連結強度之差異。以淨相關分析，控制年齡和性別，分析功能性連結強度與心率變異度及GDS-15分數間的關係。 結果: 在心血管危險因子組，除了內側前額葉皮質和雙側前扣帶皮質、內側前額葉皮質和雙側腦島、以及前扣帶皮質和雙側杏仁核之間的功能性連結為負向之外，中樞自主網路內興趣腦區之間的功能性連結皆為正向。 無心血管危險因子組的功能性連結大多比有心血管危險因子組較弱，但兩組間所有連結強度均未達顯著差異。此外，在心血管危險因子組中，內側前額葉皮質和左側杏仁核間正向功能性連結較強者，其心率變異度較弱（HF較低（r = -0.268，p = 0.027）和LF/HF較高（r = 0.315，p = 0.009））; 且內側前額葉皮質和右側杏仁核間正向功能性連結較強者，其心變異度也較弱（LF/HF較高（r = 0.289，p = 0.017））。此外，在心血管危險因子組中，內側前額葉皮質和右側杏仁核間正向功能性連結較強者，及左側島葉和右側杏仁核的正向功能性連結較弱者，其 GDS-15 分數較高（分別為r = 0.283，p = 0.019；r = -0.318，p = 0.008），亦即情緒較為低落。在無心血管危險因子組中，則未發現這些功能性連結與GDS-15分數間之顯著相關性。 討論: 無論是在有或無心血管危險因子的中老年人中，內側前額葉皮質和杏仁核間的正向功能性連結顯示：隨著年齡的增長，內側前額葉皮質與杏仁核之間缺乏互相抑制性的調節。若衰老和心血管危險因子同時存在，此現象會更明顯。更重要的是，在有心血管危險因子的中老年人中，這種正向功能性連結越強者，其心率變異度愈弱、抑鬱情緒較高，顯示中樞自主網路在調節此族群的心率變異度和情緒方面可能扮演重要角色。 結論: 此研究結果支持在有心血管危險因子的中老年人中，內側前額葉皮質和杏仁核之間的功能性連結，與自主神經系統的平衡和邊緣系統的行為表現間，具有相當重要的關聯性。

Background: People with cardiovascular risks (CVRs) often present imbalanced autonomic nervous system (ANS) function, which is manifested by smaller heart rate variability (HRV) and more depressive mood. The central autonomic network (CAN) has been implicated in the regulation of the ANS and the limbic system functions. The purposes of this study were to investigate that in middle-aged and older adults with CVRs, whether their functional connectivity (FC) within the CAN would differ from those from middle-aged and older adults without CVRs, and whether this connectivity would be associated with their HRV and depressive mood. Method: Baseline data of 104 cognitively normal middle-aged (45-64 years old) and older (65-80 years old) adults who did not have any mental illness or depression and were enrolled in one of three previous randomized controlled clinical trials (Trial No: NCT02270320, NCT03275038, or NCT05672940) were used in this study. Among them, 74 participants (mean age: 64.0 ± 7.2 years) were in the CVR group and had at least one of the following three CVRs－ hypertension, diabetes, or dyslipidemia. The remaining 30 participants (mean age: 64.8 ± 6.6 years) were in the non-CVR group. All participants were assessed with the Montreal Cognitive Assessment (MoCA), the Geriatric Depression Scale-15 (GDS-15), and brain resting-state fMRI scans. The CVR group additionally underwent a 5-minute HRV assessment at resting. The HRV parameters calculated were the standard deviation of normal to normal R wave interval (SDNN), the high frequency power (HF), the low frequency power (LF), and the ratio of the LF power over the HF power (LF/HR). Greater SDNN and HF indicated better parasympathetic control; whereas greater LF and LF/HF ratio indicated more dominant sympathetic control. Eight brain regions in the CAN were set as the regions of interest (ROIs), including the right orbitofrontal cortex (OFC), left OFC, right insula, left insula, medial prefrontal cortex (mPFC), anterior cingulate cortex (ACC), right amygdala, and left amygdala. The FC between each pair of the ROIs was calculated using the CONN toolbox and the Fisher’s Z transformation. The connectivity threshold was set as FWE, p < 0.05. Two sample t-test was used to compare the strength (z-score) of all FC within the CAN between the two groups. Partial correlation analysis was performed to analyze the relationships of the FC with the HRV measures and GDS-15 scores, controlling age and sex. Results: In the CVR group, the FC between the ROIs within the CAN was mostly positive, except for the negative FC between the mPFC and the ACC, between the mPFC and bilateral insula, and between the ACC and bilateral amygdala. Overall, the non-CVR group tended to present weaker FC than the CVR group, but the group differences in all FC values did not the reach significance. Furthermore, in the CVR group, those with stronger positive FC between the mPFC and the left amygdala presented poorer HRV (lower HF power (r = -0.268, p = 0.027) and higher LF/HF ratio (r = 0.315, p = 0.009)); and those with stronger positive FC between the mPFC and the right amygdala also presented poorer HRV (higher LF/HF ratio (r = 0.289, p = 0.017)). In addition, in the CVR group, stronger positive FC between the mPFC and the right amygdala and weaker positive FC between the insula and the right amygdala were associated with higher GDS-15 scores (r = 0.283, p = 0.019; r = -0.318, p = 0.008, respectively), that is, more depressed mood. There were no significant correlations between FC within the CAN and GDS-15 scores in the non-CVR group. Discussion: The positive FC between the mPFC and the amygdala in middle-aged and older adults with and without CVRs suggested a lack of mutual inhibitory regulation between the mPFC and the amygdala as people get older. This phenomenon became worse if aging and CVRs coexisted. More importantly, this positive FC was significantly associated with worse HRV and more depressive mood in middle-aged and older adults with CVRs, suggesting the important roles of the CAN in regulating HRV and mood in this population. Conclusion: Results of this study supported the important associations of the FC between the mPFC and amygdala with functions of the ANS and the limbic systems in middle-aged and older adults with CVRs.