Stress, mental health and memory: what cognitive training research shows

How does chronic stress affect memory?

Chronic psychological stress disrupts the rhythmic activity of neural networks in the alpha frequency range (8–12 Hz), which is closely tied to working memory, attention and executive control^[1]. A 2022 systematic review and meta-analysis of spectral-analysis EEG studies in Neurobiology of Stress mapped how acute and chronic psychosocial stress alter these oscillatory signatures across the cortex^[3].

A 2025 study in Regulatory Mechanisms in Biosystems tested this relationship where chronic stress is concentrated: young people living near a military front line in Ukraine. After a Perceived Stress Scale (PSS-10) survey of 1,200 students classified 62% with a moderate level of stress, 22% low and 16% high, the authors recorded resting EEG in 160 students aged 18–24 and related alpha-band power to a visual-memory task^[1]. Perceived stress changed where alpha activity mattered: in low-stress men, frontal alpha at the right prefrontal site Fp2 explained 43% of the variation in performance (β = 18.59; R² = 0.43), while moderately stressed participants showed weaker, more unstable associations, read by the authors as reduced neural efficiency under stress^[1].

Do brain-training studies in stressed people show memory gains?

The clearest signal comes from the 2025 frontline-students study. Across four groups of 40 participants each (men and women, low or moderate stress), the share of correctly completed visual-memory tasks rose after one month of training in every group, for example from 76.06% to 89.4% in low-stress men (P = 5.07 × 10⁻¹⁷) and from 78.20% to 93.05% in low-stress women (P = 1.67 × 10⁻¹⁴), about 17–19% across the sample^[1]. Low-stress participants reached the highest post-training scores; moderately stressed participants improved but showed more unstable neural patterns^[1].

A separate 2024 laboratory study in Biological Psychiatry: Cognitive Neuroscience and Neuroimaging approached the question from the other direction, reporting that several weeks of intensive adaptive memory training left healthy adults' working memory more resilient to an acute stressor than that of untrained controls^[5]. Both lines of work measured trained cognitive performance, not stress symptoms.

How does mental health sit between stress and memory?

Stress, mood and memory rarely move independently. A 2021 cross-sectional study in Psychology Research and Behavior Management examined 70 men with HIV who have sex with men (mean age 45.4) in Alicante, Spain, using a five-domain online cognitive assessment alongside measures of health-related quality of life and HIV self-stigma^[2]. Of the five domains, reasoning, memory, attention, coordination and perception, only memory was significantly associated with self-stigma: lower memory tracked with higher stigma (r = −0.276, p < .05), and low-memory participants scored a mean stigma of 19.31 versus 14.69 for high-memory performers (d = 0.66)^[2].

Crucially, that link did not stand on its own. When the mental-health summary of quality of life entered the model, the direct memory–stigma effect was no longer significant, consistent with full mediation, and the model explained 37.4% of the variance in self-stigma (F(3,66) = 13.16, p < .00001)^[2]. The psychological burden, not the memory score alone, carried the relationship.

The Lancet standing Commission on dementia, in its 2024 report, identified depression among 14 modifiable risk factors associated with around 45% of dementia cases worldwide, and highlighted cognitively stimulating activity as protective across the life course^[4]. That places stress-linked mental-health burden inside a larger, population-level picture of cognitive risk.

Is this different from a meditation or relaxation app?

It is a different evidence question. Relaxation, mindfulness and breathing tools are evaluated mainly on stress, mood and sleep outcomes; the studies reviewed here instead measured cognitive performance, visual memory, working memory and standardized assessment scores, in people under stress^[1][2][5]. A relaxation app aims to lower perceived stress; the cognitive-training literature asks whether trained mental abilities improve, and whether stress changes how much they improve^[1].

The two need not conflict, but they answer different questions. None of the cognitive-performance studies above measured whether training reduced participants' stress levels, so each tool should be matched to its tested outcome rather than assumed to stand in for the other.

What the evidence does NOT show

No causal, controlled proof of a stress benefit. The 2025 frontline-students study compared performance before and after training within each group, with no untrained control arm, so practice effects and natural change cannot be separated from a training effect^[1]. The 2021 HIV study was cross-sectional and "did not enable establishing any causality," with a small (N = 70), single-region sample^[2].

No treatment claim. These studies measured cognitive-performance outcomes; they do not establish that cognitive training treats, prevents or relieves chronic stress, anxiety, depression or any stress-related disorder^[1][2]. Reporting that training was studied in stressed or stigmatized populations is not the same as showing it lowers stress.

Narrow, short and untransferred. The populations were specific, Ukrainian students aged 18–24 pre-screened for stable health, and HIV-positive men in one Spanish region, and follow-up was a month or a single time point^[1][2]. Whether any gains persist, generalize, or transfer to everyday stress resilience remains open; the Lancet Commission itself stresses that risk-factor associations are not proof that a given intervention prevents decline^[4].

What this means in practice

The honest summary is narrow but real: chronic stress measurably reshapes the brain activity behind memory^[1][3], memory is the cognitive domain most tied to stress-linked psychological burden in the populations studied^[2], and short cognitive-training studies in stressed people have reported improved task performance, though without the controls needed to prove cause^[1][5]. The evidence supports treating cognitive training as a studied performance tool, not a stress treatment, and keeping mental-health care where stress symptoms are concerned^[2][4].

For the underlying studies, see the full breakdowns of the chronic-stress and visual-memory study and the memory, stigma and quality-of-life study, the cognitive-decline evidence guide, and related reviews on healthy older adults, cognition and insomnia, long COVID brain fog, and multiple sclerosis.