Cosmos Study

SSRI Antidepressants Associated with Lower Risk of Blood Cancers

August 14, 2025

Team A:Kersten Bartelt, RNVarun ValEmily Higgs

Team B:Blaine Franklin, PT, DPTGrant Keane

Key Findings

Patients on selective serotonin reuptake inhibitor (SSRI) antidepressants for three to eight years showed an 18% lower risk of developing hematologic cancer compared to those not on an SSRI.
Patients prescribed SSRIs did not have a significant change in the risk of overall cancer or cancer of the lungs, breast, colon, rectum, prostate, pancreas, or skin.

Selective serotonin reuptake inhibitors (SSRIs) are widely prescribed for depression and anxiety disorders. Previous studies suggest SSRIs may enhance T-cell activity, potentially improving immune surveillance against cancers,¹ however, other studies have found mixed associations, including potential increased risk of some cancer types.^2,3

To further explore the relationship between SSRI use and cancer diagnoses across various cancer types, we studied 627,964 adult patients, including patients with no history of SSRI use and those with three to eight years of SSRI use. Patients were matched 1:1 based on legal sex and age, and we accounted for other factors such as demographics, smoking history, BMI, use of other antidepressants, and comorbidities in our analysis.

We found that SSRI use did not significantly change the overall risk of developing cancer, as seen in Figure 1. Similarly, no significant associations were observed for lung, breast, colorectal, prostate, pancreatic, or melanoma cancers. However, patients on SSRIs had an 18% lower risk of hematologic cancers compared to those not on an SSRI. This aligns with prior animal studies suggesting SSRIs may enhance immune surveillance by modulating T-cell activity.¹

Figure 1

Cancer Risk by SSRI Use

Figure 1. Cancer risk for patients with three to eight years of SSRI use compared to those with no SSRI use.

These data come from Cosmos, a dataset created in collaboration with a community of Epic health systems representing more than 300 million patient records from 1,700 hospitals and more than 40,000 clinics from all 50 U.S. states, Lebanon, and Saudi Arabia. This study was completed by two teams that worked independently, each composed of a clinician and research scientists. The two teams came to similar conclusions. Graphics by Brian Olson.

References

Li B, Elsten-Brown J, Li M, et al. Serotonin transporter inhibits antitumor immunity through regulating the intratumoral serotonin axis. Cell. Published online 2025. doi:10.1016/j.cell.2025.04.032
Boursi B, Lurie I, Mamtani R, Haynes K, Yang YX. Anti-depressant therapy and cancer risk: a nested case-control study. Eur Neuropsychopharmacol. 2015;25(8):1147-1157. doi:10.1016/j.euroneuro.2015.04.010
Ashbury JE, Lévesque LE, Beck PA, Aronson KJ. A population-based case-control study of Selective Serotonin Reuptake Inhibitors (SSRIs) and breast cancer: the impact of duration of use, cumulative dose and latency. BMC Med. 2010;8:90. Published 2010 Dec 22. doi:10.1186/1741-7015-8-90

Data Definitions

Study period: 1/1/2017 to 6/1/2025

Study population: inclusion: Patients aged 18 or older as of their first observed face-to-face visit in the study period who have at least one outpatient face-to-face visit every 18 months during the study period

Study population: exclusion

Patients with:

A secondary cancer without evidence of a primary cancer or secondary cancer documented before a primary cancer
Evidence of SSRI use before 7/1/2017 (including patient reported/historical medications)
Patients with SSRI use for less than three years
History of any cancer (ICD-10-CM code C*) including on problem list
No census region

Censoring: First cancer diagnosis
Last outpatient face-to-face visit
Evidence of death

Index date: Case patients: three years after the first SSRI use
Control patients: three years after first outpatient face-to-face visit

Exposures: Continuous usage of an SSRI for at least three years

Outcomes: Breast cancer: ICD-10-CM code C50*
Colorectal cancer: ICD-10-CM code C18*, C19, C20
Lung cancer: ICD-10-CM code C34*
Prostate cancer: ICD-10-CM code C61*
Pancreatic cancer: ICD-10-CM code C25*
Melanoma: ICD-10-CM code C43*
Hematologic cancers: ICD-10-CM code C81*-C96*
Any cancer: ICD-10-CM code C* (excluding secondary cancers)

Secondary cancer: ICD-10-CM: C7B*, C77*, C78*, C79*

Confounders

Age at index: 18-34, 35-49, 50-64, 65+
Legal sex
Race and ethnicity
Social Vulnerability Index quintile
Census region
Has ever smoked
BMI classification

Underweight: < 18.5
Healthy: 18.5 to < 25
Overweight: 25 to < 30
Obese: 30+
Severely Obese: 40+

Evidence of first use of:

- Other antidepressants: ATC code N06A* (excluding N06AB*)
  - Before tracking period
  - During tracking period
- Anti-anxiety medications: ATC Code N05B*
  - Before tracking period
  - During tracking period

History of:

- Depression
- Anxiety
- Bipolar disorder
- Hypertension
- CKD
- Type 2 diabetes

Outpatient face-to-face visit: Encounter with a type of Emergency, Office Visit, Well Child, Follow-up, Telemedicine, Urgent Care, Walk-In, Routine Prenatal, Postpartum Visit, or Fetal Care Consult

SSRI: A prescription with ATC code N06AB*

Continuous usage: Duration in days calculated by: (Quantity / Daily Frequency) * (Refills + 1)
A maximum 90-day gap between a discontinued order and the start of a new order.
A default duration of 365 days was used in the event that one of the variables needed for the duration formula was missing.

Depression: A diagnosis with ICD-10-CM code F32*, F33*, F06.31*, F06.32*, or F34.1*

Anxiety: A diagnosis with ICD-10-CM code F40*, F41*, F06.4*, F93.0*, or F94.0*

Bipolar disorder: ICD-10-CM: F31*

Hypertension: ICD-10-CM code I10

CKD: ICD-10-CM code N18*

Type 2 diabetes: ICD-10-CM code E11*

Race and ethnicity: Patients were classified by self-reported race and ethnicity as Non-Hispanic Black (Black), Hispanic (any race), non-Hispanic White (White), non-Hispanic Asian (Asian), non-Hispanic Other (Other), non-Hispanic multiracial (Multiracial) and Unknown if no race/ethnicity information was present.

Model specifications: Cox proportional hazards model, SSRI patients to non-SSRI patients matched 1:1 on sex and age bucket