抑郁老人海马CA23DG亚区体积偏小，关联较弱

这项研究在问什么

老年抑郁与阿尔茨海默病之间的关系，是神经影像研究长期关注的问题。此前不少研究提示，一生中经历过抑郁的人，晚年患阿尔茨海默病的风险似乎更高；而海马体既与记忆密切相关，也是阿尔茨海默病最早受累的脑区之一，因此常被当作连接两者的候选环节。问题在于，海马体并不是一块均质的结构，它由CA1、亚下托、CA2、CA3、齿状回等多个亚区组成，不同亚区在抑郁和阿尔茨海默病中的表现可能并不一致，这也被认为是既往研究结果不一致的原因之一。

南加州大学Stevens神经影像与信息学研究所的Danielle Luu等人，利用“健康与衰老脑研究—健康差异”项目的数据，考察抑郁和抗抑郁药使用与老年人海马亚区体积的关系，同时纳入淀粉样蛋白、tau蛋白和APOE4基因等阿尔茨海默病相关因素。论文发表在《Translational Psychiatry》。

样本与测量方式

分析纳入2009名50至90岁的认知未受损成年人，平均年龄约65岁，通过得克萨斯大学北部分校招募西班牙裔、黑人和白人参与者。其中630人被归为有抑郁：要么老年抑郁量表得分达到10分及以上，要么自报曾被医疗人员诊断为抑郁；其余1379人归为无抑郁。参与者接受高分辨率海马MRI扫描，并报告是否使用抗抑郁药。研究者据此计算CA1、亚下托以及由CA2、CA3和齿状回合成的CA23DG三个区域的体积。

部分参与者还接受了PET扫描，分别测量脑内淀粉样蛋白（684人）和tau蛋白（386人）。抑郁诊断与用药均为自我报告，而非来自病历记录，这是解读结果时需要记住的一点。

主要发现与不确定之处

有抑郁的参与者，其CA23DG体积倾向于更小；抑郁与CA1、亚下托体积没有关联。这个关联本身幅度较小，在控制体重指数、体力活动、tau水平和APOE4状态后仍然存在。但把淀粉样蛋白纳入分析后，关联不再具有统计显著性，效应量却与之前相近——作者把这一变化归因于做淀粉样蛋白扫描的人数较少。当前抑郁症状的严重程度、抑郁起病年龄，与任何一个亚区的体积都没有关联。

在探索性分析中，抑郁与较小亚区的联系在男性中似乎更强，但作者对此持谨慎态度，因为性别差异的正式检验并未达到统计显著。

抗抑郁药的结果更复杂。在有抑郁的参与者中，用药者CA1和CA23DG体积倾向于比未用药者更小；控制体重指数和体力活动后，这些关联减弱到接近显著性边界。当抑郁与用药同时进入分析时，抑郁本身与较小CA23DG的关联消退，而用药仍与较小的CA23DG相关。作者提出，这可能反映“适应证偏倚”：抑郁更严重的人本来就更可能被开具药物。用药者报告的当前抑郁症状确实少于未用药者，提示药物在缓解症状，但研究没有治疗前抑郁严重程度和用药时长的信息。

能说明什么，不能说明什么

这项研究为老年抑郁的神经基础补充了亚区层面的观察，但它的横断面设计无法推出因果结论：无法判断抑郁是否导致海马亚区体积变化、体积差异是否先于抑郁，还是两者共同受其他因素影响。淀粉样蛋白和tau相关分析基于较小的子样本，抗抑郁药分析也没有区分药物种类和剂量。作者明确指出，药物本身是否影响海马结构，需要在随机对照试验中检验。

对读者而言，更稳妥的理解是：在这批认知正常的老年人中，抑郁与海马某一亚区的体积存在一个较弱、且对分析样本敏感的统计关联；它不构成对个人脑健康或用药选择的判断依据。

An analysis of more than 2,000 cognitively healthy older adults found that those with depression tended to have a smaller volume in one part of the hippocampus, a brain region central to memory and one of the first affected by Alzheimer’s disease. The association did not appear to be explained by amyloid and tau, two proteins linked to Alzheimer’s disease. The paper was published in Translational Psychiatry.

Alzheimer’s disease is the most common form of dementia in older adults. It gradually disrupts memory, thinking, behavior, and everyday functioning. The disease is accompanied by characteristic changes in the brain, including abnormal clumps of amyloid protein, called plaques, and twisted strands of tau protein, called tangles, along with the loss of nerve cells.

Research suggests that depression may also play a role in the risk of Alzheimer’s disease. People who experience depression during their lifetime appear to have a higher risk of developing Alzheimer’s disease later on. One possible connection between the two conditions involves the hippocampus, which is central to memory and is among the first areas affected by Alzheimer’s disease.

Depression has likewise been associated with changes in the structure of the hippocampus, although previous studies have produced inconsistent findings. Part of the difficulty may be that the hippocampus is not a single uniform structure but consists of several smaller regions (called hippocampal subfields) that can be affected differently by depression and Alzheimer’s disease. Antidepressant medications may also play a role, as some evidence suggests that they could protect against hippocampal shrinkage, while other findings raise questions about their effects in older adults.

Study author Danielle Luu of the Stevens Neuroimaging and Informatics Institute at the University of Southern California and colleagues set out to examine how depression and antidepressant use relate to the volume of hippocampal subfields in older adults without cognitive problems. They also wanted to know whether Alzheimer’s-related factors, including amyloid and tau buildup and the APOE4 gene variant, which raises Alzheimer’s risk, play a role in these relationships.

The researchers considered two possibilities for antidepressants. The medications might be linked to larger subfield volumes by relieving depressive symptoms. Alternatively, they might be linked to smaller volumes, either because of unknown harmful effects on the brain or because doctors are more likely to prescribe them to people with more severe depression.

The researchers analyzed data from the Health and Aging Brain Study-Health Disparities, which recruited Hispanic, Black, and white older adults through the University of North Texas in Fort Worth. The analysis included 2,009 cognitively unimpaired adults between 50 and 90 years of age, with an average age of about 65.

Of these participants, 630 were classified as having depression, either because they scored 10 or higher on the Geriatric Depression Scale or because they reported having been diagnosed with depression by a health care provider. The remaining 1,379 participants were classified as not depressed.

Participants underwent high-resolution MRI scans of the hippocampus and reported whether they used antidepressant medications. The researchers used the scans to calculate volumes of three hippocampal subfields: the CA1, the subiculum, and a composite region that included the CA2, CA3, and the dentate gyrus (CA23DG). Subsets of participants also underwent PET scans measuring amyloid (684 people) and tau (386 people) in the brain.

Participants with depression tended to have a smaller CA23DG, but depression was not associated with the volume of the CA1 or the subiculum. The association was small and held after the researchers accounted for body mass index and physical activity, as well as tau levels and APOE4 status. When amyloid was added to the analysis, the association was no longer statistically significant, but its size was similar, which the authors attribute to the smaller number of participants with amyloid scans.

The severity of current depressive symptoms and the age at which depression began were not associated with the volume of any subfield. In exploratory analyses, the link between depression and smaller subfields appeared stronger in men than in women, but the authors interpret this cautiously because the formal test of a sex difference was not statistically significant.

Among participants with depression, those who used antidepressant medications tended to have smaller CA1 and CA23DG volumes than those who did not. These associations weakened to borderline levels after the researchers accounted for body mass index and physical activity. When antidepressant use and depression were included in the same analysis, the link between depression itself and a smaller CA23DG faded, while antidepressant use remained associated with a smaller CA23DG.

The authors suggest that the medication findings may reflect what is known as indication bias, meaning that people with more severe depression may be more likely to be prescribed medication in the first place. Antidepressant users reported fewer current depressive symptoms than non-users, suggesting the medications were relieving symptoms, but the researchers did not have information on how severe participants’ depression was before treatment or how long they had taken medication.

“It is important to study AD [Alzheimer’s disease] in the context of mental health, since rates of depression rise with age. Moreover, it is crucial to investigate how depression and antidepressant medication use affect the hippocampus, a region of the brain that is heavily impacted by AD. Understanding the relationship between depression and the development of AD is important for exploring preventative measures against brain atrophy and dementia,” study authors concluded.

The study contributes to the scientific understanding of the neural underpinnings of depression in older adults. However, it should be noted that the design of this study does not allow any causal inferences to be derived from the results. Depression diagnoses and medication use were self-reported rather than drawn from medical records, and the analyses involving amyloid and tau were based on smaller subsets of participants.

The researchers did not compare different types or doses of antidepressants. They note that the possibility that the medications themselves affect hippocampal structure should be tested in randomized controlled trials.

The paper, “Depression and hippocampal subfield volume in older adults,” was authored by Danielle Luu, Alicia M. Twisselmann, Victoria R. Tennant, Brandon J. Hall, Arjun Mahajan, Audrey Kim, Jamie Terner, Marylan L. Davison, Koral Wheeler, Christopher R. K. Ching, James Hall, Matthew T. Borzage, Leigh Johnson, Arthur W. Toga, Sid E. O’Bryant, Kristine Yaffe, and Meredith N. Braskie for the HABS-HD Study Team.