Highlights
– Persistent depressive symptoms measured over 36 months after an MCI diagnosis identify a subgroup at markedly higher risk of progression to Alzheimer’s disease (AD).
– ADNI longitudinal analysis (n=397) found three depressive-symptom trajectories (low, moderate, high) with adjusted hazards for conversion of 1.0 (reference), 2.36, and 3.79, respectively.
– A linear, dose–response relationship was observed between 36‑month Geriatric Depression Scale (GDS) scores and AD-conversion hazard, supporting prognostic value of sustained symptom burden.
Background
Alzheimer’s disease (AD) is a leading cause of dementia worldwide and a major public health challenge. Mild cognitive impairment (MCI) represents a heterogeneous prodromal state: some people remain stable or improve, while others progress to AD dementia. Identifying modifiable predictors of conversion has high clinical and research importance for prognosis, stratified follow-up, and potential intervention.
Depressive symptoms are common in older adults and particularly prevalent in people with cognitive complaints. Epidemiologic and biologic evidence has long suggested an association between late‑life depression and subsequent dementia, but questions remain about temporality, whether depression is a causal risk factor versus an early symptom of neurodegeneration, and whether sustained symptom burden adds prognostic information beyond single‑time assessments.
Study design
Design and cohort: Ding et al. analyzed longitudinal data from 397 participants enrolled in the Alzheimer’s Disease Neuroimaging Initiative (ADNI) with a baseline diagnosis of MCI. The cohort was followed for a maximum of 156 months after a 36‑month landmark.
Depressive symptom measurement: Symptoms were assessed with the 15‑item Geriatric Depression Scale (GDS‑15), a widely used self‑report screening instrument (score range 0–15). Trajectories of depressive symptoms over the first 36 months were derived using group‑based trajectory modeling.
Outcome and analysis: The primary outcome was conversion from MCI to clinically diagnosed AD. Cox proportional‑hazards models estimated hazard ratios (HRs) for conversion associated with depressive‑symptom trajectory groups, adjusting for relevant covariates. Restricted cubic spline analysis assessed the dose–response relationship between 36‑month GDS score and conversion hazard.
Key findings
Trajectory groups: Group‑based modeling identified three distinct and clinically interpretable depressive‑symptom courses over 36 months: persistently low, persistently moderate, and persistently high. This classification captures the sustained burden of symptoms rather than transient episodes.
Primary risk estimates: Compared with the persistently low group, the adjusted HR for conversion to AD was 2.36 (95% CI 1.35–4.13; p=0.003) for the moderate‑symptom group and 3.79 (95% CI 1.86–7.69; p<0.001) for the high‑symptom group. These effect sizes are clinically meaningful: a near‑doubling to almost fourfold increase in hazard depending on sustained symptom level.
Simplified dichotomy: Using a two‑group split (low vs. high) produced an adjusted HR of 1.98 (95% CI 1.25–3.13; p=0.003), suggesting that even a coarser classification of chronic elevated mood symptoms retains prognostic value.
Dose–response relationship: Restricted cubic spline analysis demonstrated a linear, dose‑dependent association (p<0.001), with each one‑point increase in 36‑month GDS corresponding to a proportional rise in conversion hazard. This supports an exposure–response pattern rather than a threshold‑only effect.
Sensitivity and robustness: The long follow‑up (up to 13 years after the landmark) strengthens temporal inference, and modeling of trajectories over 36 months reduces the risk that a single transient depressive episode drives associations. The authors adjusted for available confounders, although specifics of covariates adjusted for (e.g., age, sex, education, baseline cognition, vascular comorbidity, antidepressant use, APOE ε4 status) should be inspected in the full report to evaluate residual confounding.
Clinical interpretation
Prognostic value: Persistently elevated depressive symptoms after MCI diagnosis identify a subgroup at substantially increased risk of progression to AD. For clinicians, repeated measurement of mood using brief instruments such as the GDS may add actionable prognostic information beyond a single assessment.
Potential mechanisms: Several, not mutually exclusive, mechanisms could underlie the association. Chronic depression may directly contribute to neurodegeneration through HPA‑axis dysregulation, hippocampal atrophy, increased neuroinflammation, cerebrovascular disease, or lifestyle changes (reduced activity, social isolation) that adversely affect cognitive reserve. Conversely, persistent depressive symptoms may reflect early AD pathology (prodromal neuropsychiatric manifestations), with neurodegenerative changes driving mood disturbance. The observed dose–response and temporal linkage (36‑month trajectory prior to long‑term conversion) strengthen—but do not prove—causality in either direction.
Strengths of the study
– Use of a well‑characterized, longitudinal ADNI cohort with standardized clinical follow‑up and long maximum follow‑up (up to 156 months) permits robust time‑to‑event analyses.
– Group‑based trajectory modeling captures chronicity and pattern of depressive symptoms rather than relying on a single assessment that may be transient or measurement error prone.
– Demonstration of a linear dose–response relationship increases confidence that higher sustained symptom burden conveys incrementally higher risk.
Limitations and cautions
– Measurement: The GDS is a validated self‑report screening tool but is not equivalent to a clinical diagnosis of major depressive disorder. Misclassification of depression subtype, severity, and treatment response is possible.
– Residual confounding: Observational data are vulnerable to confounding by factors such as vascular disease, medication use (antidepressants, sedatives), socioeconomic status, comorbid psychiatric disorders, and genetic risk (e.g., APOE ε4). The degree to which these were adjusted for affects causal interpretation.
– Reverse causation: Despite the landmark design, persistent depressive symptoms could still represent a behavioral manifestation of early AD pathology, particularly affective changes secondary to limbic or frontal dysfunction.
– Cohort generalizability: ADNI participants are a research cohort with selection biases (higher education, underrepresentation of some racial/ethnic groups) that may limit generalizability to community or clinical settings.
– Treatment implications not tested: The study is prognostic and cannot determine whether treating depressive symptoms reduces subsequent dementia risk or delays conversion.
Expert commentary and contextual evidence
Prior meta-analyses and reviews have linked late‑life depression to increased dementia risk. For example, Ownby et al. (Arch Gen Psychiatry 2006) found depression associated with higher Alzheimer disease risk in pooled analyses, and subsequent reviews (see Byers & Yaffe, Nat Rev Neurol 2011) have reinforced this relationship while highlighting heterogeneity and possible reverse causation. The Ding et al. ADNI analysis advances the field by focusing on symptom trajectories after MCI diagnosis and showing a sustained, dose‑dependent prognostic signal.
Clinical guidelines (for example, the NIA‑AA research framework) emphasize multimodal biomarkers for staging AD pathology. Incorporating longitudinal neuropsychiatric measurements adds an important and low‑cost clinical dimension to risk stratification, but integration with imaging, CSF, and genetic markers will be necessary to refine mechanistic insight and guide interventions.
Implications for practice and research
For clinicians: Routinely monitor depressive symptoms in patients with MCI using brief validated tools and consider prolonged or recurrent symptom burden as a marker of elevated AD risk. Such patients may warrant closer cognitive surveillance and consideration for more intensive assessment (biomarkers, neuroimaging) depending on the clinical context.
For researchers: Important next steps include (1) triangulation with biomarkers (amyloid, tau, structural MRI) to determine whether persistent depressive symptoms mark comorbid vascular disease, AD pathology, or independent risk pathways; (2) causal inference approaches (e.g., longitudinal mediation analysis, marginal structural models, Mendelian randomization where possible); and (3) randomized controlled trials testing whether effective treatment of persistent depressive symptoms in MCI can reduce or delay progression to dementia.
Conclusion
Ding et al. provide compelling evidence from a longitudinal ADNI cohort that persistently elevated depressive symptoms over the first 36 months after an MCI diagnosis confer a substantially higher, dose‑dependent hazard of progression to Alzheimer’s disease over long‑term follow‑up. While the results do not establish causality, they support routine monitoring of mood symptoms in MCI and underscore the need for mechanistic studies and trials to determine whether reducing sustained depressive burden can modify the course of cognitive decline.
Funding and clinicaltrials.gov
The primary analysis and data derive from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). For full funding disclosures, see Ding X et al., J Affect Disord. 2025;391:120066. ADNI is a public–private partnership with multiple sponsors; ADNI study registration is available via ClinicalTrials.gov (ADNI has been registered; see the original ADNI registry entry for details).
Selected references
– Ding X, Zheng Z, Wang H, Shao Y, Zhu S, Ma Z, Gu X, Xia T; Alzheimer’s Disease Neuroimaging Initiative. Persistent depressive-symptom trajectories predict conversion from mild cognitive impairment to Alzheimer’s disease: A longitudinal ADNI study. J Affect Disord. 2025 Dec 15;391:120066. doi:10.1016/j.jad.2025.120066.
– Ownby RL, Crocco E, Acevedo A, John V, Loewenstein D. Depression and risk for Alzheimer disease: systematic review, meta‑analysis, and metaregression. Arch Gen Psychiatry. 2006;63(5):530–538.
– Byers AL, Yaffe K. Depression and risk of developing dementia. Nat Rev Neurol. 2011;7(6):323–331.
– Jack CR Jr, Bennett DA, Blennow K, et al. NIA‑AA Research Framework: toward a biological definition of Alzheimer’s disease. Alzheimers Dement. 2018;14(4):535–562.
Practical takeaways
– Repeated assessments of depressive symptoms in people with MCI are low cost and can provide important prognostic information.
– Persistent, higher‑burden depressive symptom trajectories are associated with substantially higher hazards of conversion to AD; consider heightened surveillance and multidisciplinary assessment for such patients.
– Interventional trials are required to determine whether treating sustained depressive symptoms modifies dementia risk or delays conversion.
Image prompt for article thumbnail
Generate an image of a concerned older adult sitting with a clinician in a softly lit exam room; clinician holds a tablet displaying a brain MRI and a small line graph showing increasing depressive symptom scores over time. Include subtle overlays of a simplified brain schematic and warm but clinical color tones to convey both emotional and medical context.
