Why early- and late-onset depression are fundamentally different

By Dr. Liji Thomas, MDReviewed by Lauren HardakerNov 17 2025

By separating early- and late-onset depression, researchers uncovered hidden genetic signatures that reshape our understanding of risk, severity, and brain development.

Study: Genome-wide association analyses identify distinct genetic architectures for early-onset and late-onset depression. Image credit: Bits And Splits/Shutterstock.com

Major depressive disorder (MDD) presents with a wide range of clinical features, reflecting the interaction of genetic and environmental factors. A better understanding of its causes would promote improved clinical management and better outcomes by enabling targeted strategies.

A recent study published in the journal Nature Genetics examined differences in the genetic factors that underpin MDD subtypes of early and late onset.

Defining depression subtypes

Early- and late-onset MDD present and progress differently. The former is linked to psychotic symptoms, suicide attempts, and other physical and mental illnesses. In contrast, late-onset MDD tends to manifest with reduced cognitive performance and increased cardiovascular risk.

However, varying methodologies, small samples, and recall bias, among other reasons, contribute to the challenge of distinguishing between early- or late-onset. The current study leveraged large sample sizes and the availability of longitudinal data, including accurate records of the patient’s age at first diagnosis, a valuable substitute for the age at onset of MDD.

The availability of large samples from the Psychiatric Genomics Consortium and other global biobanks has facilitated genome-wide association studies (GWAS) of MDD. While this approach has revealed genetic variants associated with MDD, the genetic loci linked to specific MDD subtypes remain unexplored.

How depression was analyzed

The current study conducted a GWAS-based meta-analysis of early- and late-onset MDD, thus overcoming clinical heterogeneity. The study drew on Nordic biobanks providing longitudinal health data to investigate 46,708 cases of early-onset MDD and 37,168 instances of late-onset MDD.

The age at first diagnosis closely approximates the true age of onset, with a reported genetic correlation of ~0.95 between the two measures.

Key genetic differences

The scientists found 12 genomic loci (chromosomal locations), comprising 17 significant genes associated with early-onset MDD, vs only two (including four significant genes) for late-onset MDD. These were among the 80 loci in the GWAS of all MDD cases and agree with those reported in previous similar studies. The study also noted that SNP-based heritability was nearly twice as high for early-onset MDD (11.2 %) as for late-onset MDD (6 %), and that early-onset MDD showed lower polygenicity, suggesting fewer causal variants with comparatively larger effect sizes.

Early- and late-onset MDD showed moderate correlation with each other, indicating the existence of differences in their genetic profiles. Early-onset MDD was associated with genetic loci significant in neurodevelopment, suggesting a link to early brain development.  In contrast, in patients with late-onset MDD, only one epigenetic marker was present in male fetal tissues. Neither subtype showed enrichment in adult brain tissues, which the authors suggested may reflect limited statistical power, especially for late-onset MDD.

The two MDD subtypes also differed in the correlation between their genetic architecture and other traits. Early-onset MDD genetic loci showed the highest correlation with suicide attempts, more than twice as high as those of late-onset MDD. Early-onset MDD was more strongly correlated with suicide attempts than the other subtype, but both had similar effects on suicide death.

Mendelian randomization analyses further supported a putative causal effect of early-onset MDD on suicide attempt, whereas the impact on suicide death was similar across subtypes.

The two subtypes also showed differences in their correlation with posttraumatic stress disorder, childhood maltreatment, autism spectrum disorder, and schizophrenia, among others. Early-onset MDD had overlapping genetic associations with heart failure and body mass index markers.

Interestingly, late-onset MDD also showed overlap of its genetic loci with many traits like suicide attempt or suicide death. However, these seemed to be driven mainly by the genes common to both subtypes and were weakened after adjusting for early-onset MDD. In contrast, the reverse was not true. There was limited replication of subtype-specific loci in the UK Biobank, with only one locus showing nominal significance; however, effect sizes across cohorts remained highly correlated.

PRS risk patterns

The researchers also calculated polygenic risk scores (PRS) for both subtypes of MDD. They found significant associations between PRS for early-onset MDD with various clinical indicators and outcomes, including risk of early-onset and lifetime risk of MDD, hospitalization risk, and a change in diagnosis to bipolar disorder or schizophrenia over time.

Certain mental conditions were uniquely linked to the early-onset MDD PRS, including suicide attempt or intentional self-harm and childhood maltreatment-related problems. Additional unique associations included conduct disorder, schizotypal disorder, and other childhood adversity-related ICD-10 codes. Late-onset MDD was more likely to be associated with mental or behavioral illness due to sedative or hypnotic use, or obsessive-compulsive disorder.

The top tenth percentile of early-onset MDD PRS scores had a 26 % risk for suicide attempts in the first ten years following a diagnosis of MDD. The middle 80 % had a somewhat lower risk, at 20 %, while the lowest tenth percentile had a 12 % risk. The relative risk of suicide attempts was reduced by 43 % in the lowest decile compared to the middle group, but the top decile had a 13 % higher risk than the middle.

The risk curves for the middle and top deciles diverged only approximately 5.5 years after diagnosis, and the hazard ratio for the top decile was only marginally significant.

Notably, young people are most likely to attempt suicide, making the early-onset MDD PRS potentially valuable in identifying the high-risk young people in this subgroup.

Implications for psychiatry

The study focused on identifying genetic loci for specific subgroups of MDD, identified by their signs and symptoms, thereby reducing unwanted genetic variation among MDD patients. Similar strategies could be applied to identify additional MDD subgroups defined by features such as vegetative symptoms or psychotic manifestations.

Taken together, our findings can inform precision psychiatry approaches for MDD.

The authors further emphasized that stratifying MDD by age of onset may reveal more biologically coherent subtypes and could aid in developing targeted risk prediction and prevention strategies.

Download your PDF copy now!