Neural systems underlying autobiographical memory dysregulations in depressive and at-risk individuals: A neuroimaging meta-analysis

Chao Cheng, Hei Lam Michelle Tsang, Meng Han, Jie Zhang, Michael Maes, Benjamin Klugah‐Brown, Mercy Chepngetich Bore, Benjamin BeckermedRxiv 2025 · DOI

TL;DR

This neuroimaging meta-analysis challenges the traditional "limbic-centric" view of depression, finding that autobiographical memory (AM) deficits are driven not by the amygdala or hippocampus, but by overactivity in the dorsal anterior cingulate cortex (dACC) and the precuneus. These findings suggest that depression-related memory issues are rooted in a "vicious cycle" where the brain over-prioritizes negative self-referential information while becoming "distanced" from vivid, positive experiences.

What Problem Does This Paper Address?

Autobiographical memory—the ability to recall specific personal experiences—is a cornerstone of our identity. In individuals with depression, this system often malfunctions: memories become "overgeneralized" (vague) or heavily biased toward negative events. While researchers agree this dysfunction helps maintain depressive states, individual fMRI studies have produced a "noise" of inconsistent results. Some point to the amygdala (emotion), others to the hippocampus (memory), and many find nothing at all.

The authors conducted this meta-analysis to filter out the noise and identify the robust neural markers of AM dysfunction across the existing literature, providing a clearer map of how the depressed brain navigates its own past.

Methods

The researchers employed a coordinate-based meta-analysis using a sophisticated technique called SDM-PSI (Seed-based d Mapping with Permutation of Subject Images). Unlike older methods, SDM-PSI allows for a more unbiased estimation of effect sizes by accounting for both positive and negative findings across studies.

Beyond just identifying active regions, the team performed meta-analytic decoding using the Neurosynth database. This allowed them to "reverse-engineer" the likely psychological functions of the identified brain regions (e.g., linking a specific coordinate to "pain" or "social cognition") and map their functional connectivity to large-scale brain networks.

Datasets

The study synthesized data from 13 peer-reviewed fMRI studies published between 1998 and 2023.

  • Total Participants: 602 individuals (341 patients with unipolar depression or at-risk, 261 healthy controls).
  • Tasks: Various experimental paradigms including cued recall, personalized scripts, and prospection (thinking about the future).
  • Inclusion Criteria: Only whole-brain, case-control studies were included to ensure the results weren't biased by pre-selected "regions of interest."

Key Insights and Ideas

  • The "Usual Suspects" are Missing: Surprisingly, the meta-analysis found no significant differences in the amygdala or hippocampus between depressed individuals and controls. This suggests that while these regions are vital for memory formation, they may not be the primary drivers of memory recall dysregulation in depression.
  • The Salience-Default Tug-of-War: The findings highlight two key hubs: the dACC (part of the Salience Network) and the precuneus (a core node of the Default Mode Network).
  • The dACC overactivity is linked to negative affect and "social pain," suggesting depressed individuals "over-process" the emotional conflict of their past.
  • The precuneus overactivity is linked to self-referential processing, potentially reflecting the "fixation" or rumination on negative personal narratives.
  • The "Distancing" Effect: The study found decreased activation in the right anterior insula. Since the insula is responsible for interoception (feeling the state of one's body) and vividness, this suggests that depressed individuals may feel "numb" or "distant" from their memories, lacking the "vividness" that makes healthy recall rewarding.
  • Valence Matters: The exploratory analysis showed that the dACC and precuneus overactivity was specifically driven by negative memory recall, reinforcing the idea of a negative cognitive bias.

Key Results

  • Increased Activation: Depressed individuals showed significantly higher activation in the right paracingulate cortex/dACC ($Z = 4.021$) and the right precuneus ($Z = 3.221$) during AM recall compared to healthy controls.
  • Decreased Activation: A significant cluster of decreased activity was found in the right anterior insula ($Z = -3.307$).
  • Behavioral Decoding: The dACC was strongly associated with terms like "pain," "conflict," and "executive control," while the precuneus was linked to "autobiographical," "theory of mind," and "social cognition."
  • Positive vs. Negative: Negative recall specifically activated the dACC/precuneus network, while positive recall was associated with increased activity in the right inferior temporal gyrus.

Limitations and Future Work

The authors acknowledge several constraints:

  1. Small Study Pool: Only 13 studies met the rigorous inclusion criteria, which limits the statistical power for sub-group analyses.
  2. Valence Data Scarcity: Not all studies separated positive and negative memories, meaning the valence-specific findings are considered "exploratory."
  3. Episodic vs. Semantic: The meta-analysis couldn't fully disentangle "episodic" memories (specific events) from "semantic" ones (general facts about oneself), which may involve different neural pathways.
  4. Demographics: Most studies were conducted in Western populations; future research needs more cross-cultural data to see if these neural markers are universal.

The authors suggest that future interventions, such as real-time fMRI neurofeedback, could specifically target the dACC or precuneus to help patients "re-train" their memory systems.

Summary updated

Abstract

Abstract Background Autobiographical memory (AM) dysfunction has been proposed as a neurocognitive mechanism underlying the development and maintenance of depression. However, case-control neuroimaging studies investigating the neural correlates of AM in depression have yielded inconsistent findings. The present study utilized neuroimaging meta-analyses to identify robust neural markers of AM dysfunction in depression and to characterize the associated behavioural and network-level mechanisms. Methods A pre-registered neuroimaging meta-analysis ( https://osf.io/35xtf ) was conducted, incorporating data from 341 patients with unipolar depression and 261 healthy controls across case-control studies examining AM processing. Meta-analytic network-level and behavioural decoding analyses were performed to aid interpretation of the findings. Results Compared to controls, patients with depression showed increased activation in the right paracingulate cortex (dorsal anterior cingulate, dACC) and precuneus, and decreased activation in the anterior insula during AM recall. Exploratory valence-specific analyses revealed that negative AM recall was associated with increased activity the dACC and precuneus. Meta-analytic decoding linked the dACC to the salience network and to domains related to negative affect and executive control, while the precuneus was associated with the default mode network and with processes related to social cognition and autobiographical memory. Conclusions Findings do not support prevailing models emphasizing altered amygdala and hippocampal function in AM deficits in depression. Instead, they highlight the involvement of core regions within the salience and default mode networks as key neural substrates of AM dysfunction. These regions may contribute to affective, social-cognitive, and mnemonic disturbances that shape the valence-specific nature of AM deficits in depression.
BibTeX
@misc{cheng2025neural,
  author = {Chao Cheng and Hei Lam Michelle Tsang and Meng Han and Jie Zhang and Michael Maes and Benjamin Klugah‐Brown and Mercy Chepngetich Bore and Benjamin Becker},
  title = {{Neural systems underlying autobiographical memory dysregulations in depressive and at-risk individuals: A neuroimaging meta-analysis}},
  year = {2025},
  doi = {10.1101/2025.04.06.25325353},
  howpublished = {medRxiv},
}
Figure