A large South Korean study of older adults has found that longer cumulative use of proton pump inhibitors (PPIs)—widely prescribed drugs that suppress stomach acid—was associated with a higher likelihood of developing depression. The research, based on more than half a million people followed for up to nine years, also identified a possible association between prolonged PPI exposure and suicide, although the authors stress that the suicide analysis was exploratory and cannot establish that the medicines caused suicidal behavior. The findings add a new dimension to an ongoing debate about the long-term safety of medications such as omeprazole, esomeprazole, lansoprazole and pantoprazole, which are commonly used to treat gastroesophageal reflux disease, peptic ulcers and other acid-related disorders. PPIs are generally considered effective and well tolerated when prescribed appropriately, but their ability to alter the digestive tract’s chemical environment over long periods has raised questions about possible effects beyond the stomach.
The study, published in the journal Discover Mental Health, examined records from the Korea National Health Insurance Service–Senior Cohort. The researchers identified 553,362 older adults and assessed their cumulative PPI use during a four-year period from 2007 through 2010. Participants were then followed from 1 January 2011 to 31 December 2019 to determine who developed depression. Suicide was evaluated as a secondary, exploratory outcome. During follow-up, 50,872 participants developed depression and 57 died by suicide. Because older adults commonly take several medications and often live with multiple chronic illnesses, the investigators used statistical models designed to adjust for measured differences between PPI users and nonusers. Even with those adjustments, the study was observational: the researchers analyzed patterns in existing health records rather than assigning people to take a drug or a placebo.
The central finding concerned cumulative exposure. Among PPI users, every additional 30 days of cumulative use was associated with a subdistribution hazard ratio of 1.020 for incident depression, with a 95 percent confidence interval from 1.018 to 1.023. In practical statistical terms, this represents an estimated 2 percent increase in the relative incidence associated with each additional 30 days of exposure within the analysis—not a 2 percent increase in an individual’s absolute risk, and not proof that one month of treatment directly produces depression. When the researchers compared people with at least 90 days of cumulative PPI use with those who had used the drugs for only one to 29 days, the longer-use group had a subdistribution hazard ratio of 1.428. That estimate corresponds to a 42.8 percent higher relative incidence of depression compared with the shorter-use reference group.
The investigators used Cox proportional hazards and Fine–Gray subdistribution hazard models, methods that account for the fact that death can prevent a person from later receiving a depression diagnosis. This is known as a competing risk. A conventional time-to-event analysis can become misleading when a substantial number of participants die before the outcome of interest occurs, because those individuals are no longer available to experience the event. The Fine–Gray approach estimates differences in the cumulative incidence of depression while incorporating that competing event. The study also considered whether vitamin B12 or magnesium deficiency might help explain the observed association. These mediation analyses suggested a possible indirect pathway involving vitamin B12 deficiency, but the authors describe the result as exploratory rather than definitive.
The biological explanation remains uncertain, but several mechanisms are plausible enough to warrant further study. Stomach acid helps release vitamin B12 from food and supports its subsequent absorption, particularly when the vitamin is bound to dietary proteins. Prolonged acid suppression could therefore contribute to lower B12 availability in some people, although the size and clinical importance of that effect vary. Vitamin B12 is required for red-blood-cell formation, DNA synthesis and the maintenance of the nervous system; deficiency can produce fatigue, cognitive changes, nerve damage and mood symptoms. Magnesium absorption may also be affected in a small subset of long-term PPI users, especially those with other risk factors. In addition, the gut’s microbial environment, which is influenced by gastric acidity, could potentially affect immune signaling or the production of metabolites involved in brain function. None of these mechanisms was proven by the study, and a statistical mediation signal does not demonstrate a complete biological chain from PPI use to depression.
The suicide analysis produced a stronger-looking but much less precise signal. Among PPI users, each additional 30 days of cumulative exposure was associated with a suicide subdistribution hazard ratio of 1.067, with a 95 percent confidence interval of 1.025 to 1.111. Participants with at least 90 days of cumulative use had an estimated suicide incidence 3.767 times that of participants with one to 29 days of use, but the confidence interval ranged from 1.082 to 13.113. That wide interval reflects the small number of suicide deaths—only 57 across the cohort—and means the estimate is statistically unstable. Suicide is also influenced by a complex network of psychiatric, social, medical and environmental factors that are difficult to capture fully in insurance records. The authors therefore caution against interpreting the finding as evidence that PPIs cause suicide, and the result should not be used to prompt abrupt medication discontinuation.
A major challenge in interpreting the depression association is confounding by indication and illness severity. People prescribed PPIs for extended periods may have more persistent gastrointestinal disease, chronic pain, sleep disruption, frailty or other health problems than people who use them briefly. These conditions themselves may increase the risk of depression. Some medications taken alongside PPIs may also influence mood, while early, undiagnosed depression could affect healthcare use and lead to more prescriptions. The study adjusted for available variables, but administrative health data cannot measure every relevant factor, including symptom severity, over-the-counter PPI use, adherence, lifestyle, social isolation and detailed psychiatric history. The exposure assessment also covered the pre-index period rather than directly measuring blood concentrations or confirming that patients took every prescribed dose. These limitations mean the results show an association in a particular population, not a clinical rule that applies equally to every PPI user.
The findings arrive as clinicians are increasingly encouraged to review the necessity and duration of acid-suppressing treatment. PPIs can be essential for preventing gastrointestinal bleeding in high-risk patients, healing ulcers and managing severe reflux or inflammation. For others, they may have been continued automatically after the original indication has resolved. The new results do not overturn the established benefits of appropriate treatment, but they suggest that long-term therapy may deserve periodic reassessment, particularly in older adults taking multiple medicines. A medication review could consider whether the original indication remains, whether the dose is still necessary and whether symptoms can be managed with a shorter course or an alternative strategy. Any change should be guided by a healthcare professional, because stopping PPIs suddenly after prolonged use can cause rebound acid hypersecretion and a return of symptoms.
Future research will need to test whether the relationship persists in different countries and healthcare systems, and whether it changes according to dose, specific PPI, duration, underlying disease or nutritional status. Prospective studies could measure vitamin B12 and magnesium levels before and during treatment, collect detailed information about depression symptoms and distinguish prescribed from over-the-counter use. Researchers may also be able to examine whether correcting documented deficiencies changes the risk of mood disorders. Randomized trials would provide the strongest evidence about causation, although long-term trials involving rare outcomes such as suicide would be difficult and ethically complex. For now, the South Korean analysis serves as a signal for closer investigation: prolonged PPI exposure was linked with depression diagnoses in a very large older population, while the suicide and vitamin B12 findings remain important but preliminary clues rather than proof of harm.
