Journal Article Annotations
2023, 3rd Quarter

Aging

Annotations by Zach Harvanek, MD, PhD
October, 2023

Association between the epigenetic lifespan predictor GrimAge and history of suicide attempt in bipolar disorder.
Shorter telomere length predicts poor antidepressant response and poorer cardiometabolic indices in major depression.
Epigenetic aging and PTSD outcomes in the immediate aftermath of trauma.

PUBLICATION #1 — Aging

Association between the epigenetic lifespan predictor GrimAge and history of suicide attempt in bipolar disorder.

Camila N C Lima, Emese H C Kovács, Salahudeen Mirza, Alexandra Del Favero-Campbell, Alexandre Paim Diaz, Joao Quevedo, Benney M R Argue, Jenny Gringer Richards, Aislinn Williams, John A Wemmie, Vincent A Magnotta, Jess G Fiedorowicz, Jair C Soares, Marie E Gaine, Gabriel R Fries

Abstract: Neuropsychopharmacology. 2023 May;48(6):954-962. doi: 10.1038/s41386-023-01557-9. Epub 2023 Mar 6.

Bipolar disorder (BD) has been previously associated with premature mortality and aging, including acceleration of epigenetic aging. Suicide attempts (SA) are greatly elevated in BD and are associated with decreased lifespan, biological aging, and poorer clinical outcomes. We investigated the relationship between GrimAge, an epigenetic clock trained on time-to-death and associated with mortality and lifespan, and SA in two independent cohorts of BD individuals (discovery cohort – controls (n = 50), BD individuals with (n = 77, BD/SA) and without (n = 67, BD/non-SA) lifetime history of SA; replication cohort – BD/SA (n = 48) and BD/non-SA (n = 47)). An acceleration index for the GrimAge clock (GrimAgeAccel) was computed from blood DNA methylation (DNAm) and compared between groups with multiple general linear models. Differences in epigenetic aging from the discovery cohort were validated in the independent replication cohort. In the discovery cohort, controls, BD/non-SA, and BD/SA significantly differed on GrimAgeAccel (F = 5.424, p = 0.005), with the highest GrimAgeAccel in BD/SA (p = 0.004, BD/SA vs. controls). Within the BD individuals, BD/non-SA and BD/SA differed on GrimAgeAccel in both cohorts (p = 0.008) after covariate adjustment. Finally, DNAm-based surrogates revealed possible involvement of plasminogen activator inhibitor 1, leptin, and smoking pack-years in driving accelerated epigenetic aging. These findings pair with existing evidence that not only BD, but also SA, may be associated with an accelerated biological aging and provide putative biological mechanisms for morbidity and premature mortality in this population.

Annotation

The finding:
The study authors found that individuals with bipolar disorder were biologically older than individuals without bipolar disorder after adjusting for age. This was true for those with bipolar who had never attempted suicide (1.19 year older than controls) and those with bipolar who had previously attempted suicide (2.83 years older than controls). This difference between those who had attempted suicide and those who had not was also significant. In a replication cohort with 95 individuals who have bipolar 1 disorder, those who had previously attempted suicide were again found to be biologically older than those who had not attempted. In general, these results remained significant after accounting for sex, body-mass index, smoking, ancestry, and age. The authors further analyzed specific components of the biological aging clock to suggest this effect could be mediated by plasminogen activator inhibitor 1, leptin, and smoking.

Strength and weaknesses:
Strengths of this study include the use of two independently-recruited cohorts, enhancing generalizability and reliability of findings. The authors also account for multiple covariables, though with multiple models their findings could be difficult to interpret. Weaknesses include that the analysis is cross-sectional and thus causality is unclear: Have the prior suicide attempts driven accelerated aging, or is accelerated aging predictive of suicide attempts? The sample size of each study is also relatively small, and the replication cohort lacked a comparison group of patients without bipolar disorder.

Relevance:
These findings suggest accelerated aging is associated with both bipolar disorder and suicide risk. Notably, lithium has previously been associated with extended lifespan (see the Q1 2023 annotation), though in this study did not find a direct relationship between medication use and accelerated aging. Future studies could investigate biological aging as a potential marker and/or target for bipolar treatment and suicide prevention. The benefits of treatment on aging outcomes remain speculative—for now.

PUBLICATION #2 — Aging

Shorter telomere length predicts poor antidepressant response and poorer cardiometabolic indices in major depression.

Ryan Rampersaud, Gwyneth W Y Wu, Victor I Reus, Jue Lin, Elizabeth H Blackburn, Elissa S Epel, Christina M Hough, Synthia H Mellon, Owen M Wolkowitz

Abstract: Sci Rep. 2023 Jun 23;13(1):10238. doi: 10.1038/s41598-023-35912-z.

Telomere length (TL) is a marker of biological aging, and shorter telomeres have been associated with several medical and psychiatric disorders, including cardiometabolic dysregulation and Major Depressive Disorder (MDD). In addition, studies have shown shorter TL to be associated with poorer response to certain psychotropic medications, and our previous work suggested shorter TL and higher telomerase activity (TA) predicts poorer response to Selective Serotonin Reuptake Inhibitor (SSRI) treatment. Using a new group of unmedicated medically healthy individuals with MDD (n = 48), we sought to replicate our prior findings demonstrating that peripheral blood mononuclear cell (PBMC) TL and TA predict response to SSRI treatment and to identify associations between TL and TA with biological stress mediators and cardiometabolic risk indices. Our results demonstrate that longer pre-treatment TL was associated with better response to SSRI treatment (β = .407 p = .007). Additionally, we observed that TL had a negative relationship with allostatic load (β = – .320 p = .017) and a cardiometabolic risk score (β = – .300 p = .025). Our results suggest that PBMC TL reflects, in part, the cumulative effects of physiological stress and cardiovascular risk in MDD and may be a biomarker for predicting SSRI response.

Annotation

The finding:
In a prospective study of medically healthy, unmedicated depressed patients (n = 48), the authors identified that longer pre-treatment telomere length (associated with younger biological age) predicted a positive response to SSRI treatment after 8 weeks. Positive response was defined as an improvement of at least 50% in HDRS scores (n of responders = 10, n of non-responders = 22). ROC analysis of telomere length demonstrated an AUC of 78.6%. The authors went on to demonstrate that telomere length was associated with increased allostatic load and increased cardiometabolic risk scores. No significant relationships were found with telomerase activity and outcome variables.

Strength and weaknesses:
Strengths of this study include its prospective design with a priori hypotheses based on prior work; some of this work replicates those prior findings as well. The use of medication-free subjects also enhances the interpretability of their study, although it may limit the generalizability. Weaknesses include a relatively small sample size and the lack of follow-up after the 8 week time frame.

Relevance:
Measures of biological aging, such as telomere length, may be useful biomarkers to identify individuals who will respond to SSRI treatment, though further studies are needed to confirm these results and identify potential cut-off points.

PUBLICATION #3 — Aging

Epigenetic aging and PTSD outcomes in the immediate aftermath of trauma.

Anthony S Zannas, Sarah D Linnstaedt, Xinming An, Jennifer S Stevens, Nathaniel G Harnett, Alyssa R Roeckner, Katelyn I Oliver, David R Rubinow, Elisabeth B Binder, Karestan C Koenen, Kerry J Ressler, Samuel A McLean.

Abstract: Psychol Med. 2023 Mar 23:1-10. doi: 10.1017/S0033291723000636. Online ahead of print.

Background:
Psychological trauma exposure and posttraumatic stress disorder (PTSD) have been associated with advanced epigenetic age. However, whether epigenetic aging measured at the time of trauma predicts the subsequent development of PTSD outcomes is unknown. Moreover, the neural substrates underlying posttraumatic outcomes associated with epigenetic aging are unclear.

Methods:
We examined a multi-ancestry cohort of women and men (n = 289) who presented to the emergency department (ED) after trauma. Blood DNA was collected at ED presentation, and EPIC DNA methylation arrays were used to assess four widely used metrics of epigenetic aging (HorvathAge, HannumAge, PhenoAge, and GrimAge). PTSD symptoms were evaluated longitudinally at the time of ED presentation and over the ensuing 6 months. Structural and functional neuroimaging was performed 2 weeks after trauma.

Results:
After covariate adjustment and correction for multiple comparisons, advanced ED GrimAge predicted increased risk for 6-month probable PTSD diagnosis. Secondary analyses suggested that the prediction of PTSD by GrimAge was driven by worse trajectories for intrusive memories and nightmares. Advanced ED GrimAge was also associated with reduced volume of the whole amygdala and specific amygdala subregions, including the cortico-amygdaloid transition and the cortical and accessory basal nuclei.

Conclusions:
Our findings shed new light on the relation between biological aging and trauma-related phenotypes, suggesting that GrimAge measured at the time of trauma predicts PTSD trajectories and is associated with relevant brain alterations. Furthering these findings has the potential to enhance early prevention and treatment of posttraumatic psychiatric sequelae.

Keywords:
Aging; DNA methylation; PTSD; epigenetics; stress.

Annotation

The finding:
In this study, the authors recruited 289 individuals who presented to the emergency department within 72 hours of experiencing trauma, collected blood samples to assess epigenetic age, and then followed subjects longitudinally to determine if they developed PTSD after 6 months. The authors identified accelerated biological age (as measured via GrimAge, an epigenetic clock) as predictive of increased risk for the development of PTSD, with those in the highest third of GrimAge having a 44% increased relative risk compared to those in the lowest third. Accelerated GrimAge predicted worse intrusive memories and nightmares though was not related to other symptom clusters. Accelerated GrimAge was also associated with reduced amygdala volume 2 weeks after the traumatic event in a subset of 63 patients who also had neuroimaging data.

Strength and weaknesses:
Strengths of this study include its prospective design, including the collection of data shortly after the traumatic event and follow-up for incident traumatic symptoms. The inclusion of neuroimaging data is also a strength, although it is unclear if these findings represent changes shortly after trauma or whether they reflect changes that were already present. Weaknesses include the lack of a control group, and the lack of a second timepoint measuring epigenetic aging.

Relevance:
These findings suggest that risk for the development of PTSD after traumatic events may be predicted by epigenetic aging. Future studies could further use GrimAge to identify of at-risk individuals and potentially predict outcomes after early intervention.