Department of Psychiatry & Behavioral Sciences, Duke University Medical Center, Durham, North Carolina.
Department of Molecular Biophysics & Biochemistry, Yale University, New Haven, Connecticut.
JAMA Netw Open. 2023 Jun 1;6(6):e2317714. doi: 10.1001/jamanetworkopen.2023.17714.
Major depressive disorder (MDD) is a leading cause of global distress and disability. Earlier studies have indicated that antidepressant therapy confers a modest reduction in depressive symptoms on average, but the distribution of this reduction requires more research.
To estimate the distribution of antidepressant response by depression severity.
DESIGN, SETTING, AND PARTICIPANTS: In this secondary analysis of pooled trial data, quantile treatment effect (QTE) analysis was conducted from the US Food and Drug Administration (FDA) database of antidepressant monotherapy for patients with MDD, encompassing 232 positive and negative trials submitted to the FDA between 1979 and 2016. Analysis was restricted to participants with severe MDD (17-item Hamilton Rating Scale for Depression [HAMD-17] score ≥20). Data analysis was conducted from August 16, 2022, to April 16, 2023.
Antidepressant monotherapy compared with placebo.
The distribution of percentage depression response was compared between the pooled treatment arm and pooled placebo arm. Percentage depression response was defined as 1 minus the ratio of final depression severity to baseline depression severity, expressed as a percentage. Depression severity was reported in HAMD-17-equivalent units.
A total of 57 313 participants with severe depression were included in the analysis. There was no significant imbalance in baseline depression severity between the pooled treatment arm and pooled placebo arm, with a mean HAMD-17 difference of 0.037 points (P = .11 by Wilcoxon rank sum test). An interaction term test for rank similarity did not reject the rank similarity governing percentage depression response (P > .99). The entire distribution of depression response was more favorable in the pooled treatment arm than in the pooled placebo arm. The maximum separation between treatment and placebo occurred at the 55th quantile and corresponded to an absolute improvement in depression due to active drug of 13.5% (95% CI, 12.4%-14.4%). The separation between treatment and placebo diminished near the tails of the distribution.
In this QTE analysis of pooled clinical trial data from the FDA, antidepressants were found to confer a small reduction in depression severity that was broadly distributed across participants with severe depression. Alternatively, if the assumptions behind the QTE analysis are not met, then the data are also compatible with antidepressants eliciting more complete response in a smaller subset of participants than is suggested by this QTE analysis.
重度抑郁症（MDD）是全球困扰和残疾的主要原因。早期研究表明，抗抑郁治疗平均可使抑郁症状略有减轻，但这种减轻的分布需要更多的研究。
按抑郁严重程度估计抗抑郁反应的分布。
设计、环境和参与者：在这项对来自美国食品和药物管理局（FDA）的抗抑郁药单药治疗 MDD 患者的汇总试验数据的二次分析中，使用 FDA 数据库进行了定量治疗效果（QTE）分析，涵盖了 1979 年至 2016 年期间向 FDA 提交的 232 项阳性和阴性试验。分析仅限于患有严重 MDD（汉密尔顿抑郁量表 17 项[HAMD-17]评分≥20）的参与者。数据分析于 2022 年 8 月 16 日至 2023 年 4 月 16 日进行。
抗抑郁药单药治疗与安慰剂相比。
比较了汇总治疗组和汇总安慰剂组之间的抑郁反应百分比分布。抑郁反应百分比定义为终末抑郁严重程度与基线抑郁严重程度之比的 1 减去，以百分比表示。抑郁严重程度以 HAMD-17 等效单位报告。
共纳入 57313 名患有严重抑郁症的参与者进行分析。汇总治疗组和汇总安慰剂组之间的基线抑郁严重程度无显著失衡，HAMD-17 差异平均为 0.037 分（Wilcoxon 秩和检验，P=0.11）。秩相似性的交互检验未拒绝主导抑郁反应的秩相似性（P＞0.99）。与汇总安慰剂组相比，汇总治疗组的抑郁反应整个分布更有利。治疗和安慰剂之间的最大分离发生在第 55 分位数，这对应于活性药物引起的抑郁绝对改善 13.5%（95%CI，12.4%-14.4%）。在分布的尾部附近，治疗和安慰剂之间的分离减小。
在这项来自 FDA 的汇总临床试验数据的 QTE 分析中，发现抗抑郁药可使抑郁严重程度略有降低，这在患有严重抑郁症的参与者中广泛分布。或者，如果 QTE 分析背后的假设不成立，那么数据也与抗抑郁药在比 QTE 分析所暗示的更小的亚组参与者中引起更完全的反应兼容。
