Fosamax Mental Health and Mood Impact: What the Evidence Actually Shows

What Is Alendronate and Why Does the Mental Health Question Arise?

Alendronate is a nitrogen-containing bisphosphonate that inhibits osteoclast-mediated bone resorption by blocking farnesyl pyrophosphate synthase, an enzyme in the mevalonate pathway. The FDA approved it in 1995, and it remains one of the most widely prescribed drugs for postmenopausal osteoporosis and fracture prevention. The Fracture Intervention Trial (FIT, published in JAMA 1998, N=6,458) showed a 47% reduction in vertebral fracture risk over three years compared with placebo, establishing alendronate as a first-line agent. [1]

The mental health question arises from three converging sources: spontaneous adverse-event reports in FAERS that mention depression and mood changes, the biological observation that bone-derived hormones (particularly osteocalcin) influence brain function, and the high baseline prevalence of depression and anxiety among people who have osteoporosis. Separating a drug signal from background noise requires careful reading of pharmacovigilance data alongside controlled trial results.

The Osteoporosis-Depression Overlap

Osteoporosis does not develop in a vacuum. A 2018 meta-analysis in Osteoporosis International (N=14 studies, combined cohort exceeding 180,000 participants) reported a pooled odds ratio of 1.65 for depression among people with osteoporosis compared with bone-healthy controls. [2] Pain from vertebral compression fractures, restricted mobility, and fear of falling each contribute independently to low mood. Any analysis of alendronate's mood effects must account for this baseline elevation.

Oral Bioavailability and CNS Penetration

Oral alendronate has systemic bioavailability below 1% under fasting conditions, which constrains but does not eliminate the possibility of central nervous system effects. [3] The drug binds avidly to hydroxyapatite in bone and is not thought to cross the blood-brain barrier in pharmacologically relevant amounts in humans. Still, indirect mechanisms, particularly effects on systemic signaling molecules that interact with the brain, remain biologically plausible.

What Large Randomized Controlled Trials Show

FIT (Fracture Intervention Trial, 1998)

FIT randomized 6,458 postmenopausal women with low femoral neck bone density to alendronate (5 mg/day for two years, then 10 mg/day) or placebo over a mean follow-up of 3.8 years. The primary endpoint was vertebral fracture, but investigators tracked adverse events systematically. [1] Depression and anxiety were not identified as statistically significant adverse events in either the vertebral fracture arm or the clinical fracture arm of FIT. Rates of psychiatric adverse events were numerically similar between alendronate and placebo groups.

Limitations apply. FIT excluded women with significant psychiatric comorbidities at enrollment, used physician-report rather than validated psychiatric instruments, and was powered for fracture, not mood endpoints. Those design choices mean the study could detect only large, obvious psychiatric signals.

FLEX (Fracture Intervention Trial Long-Term Extension, 2006)

FLEX (N=1,099) followed women who had completed FIT and randomly assigned them to continue alendronate or switch to placebo for an additional five years. The FLEX investigators reported no significant difference in mood-related adverse events between continuation and discontinuation groups at any of the yearly assessments. [4] The absence of a signal after up to 10 cumulative years of exposure is clinically meaningful, even acknowledging the study's limited psychiatric methodology.

Horizon Key Fracture Trial (Zoledronic Acid, a Related Bisphosphonate)

Although zoledronic acid is not alendronate, HORIZON-PFT (N=7,765) provides the most rigorous bisphosphonate psychiatric-safety dataset in existence because investigators used validated instruments and collected depression data prospectively. A pre-specified analysis published in JBMR 2011 found no significant difference in depressive symptom scores between zoledronic acid and placebo over 36 months. [5] Given that both drugs share the same mechanistic class and differ mainly in route and potency, this class-level negative finding is relevant context for alendronate.

Post-Marketing Signals and Pharmacovigilance Data

FAERS Reports and Their Limitations

The FDA Adverse Event Reporting System (FAERS) contains case reports associating alendronate with depression, emotional lability, auditory hallucinations, and, in isolated cases, suicidal ideation. These reports do not establish causation. FAERS data are subject to notoriety bias (spontaneous reporting increases after media attention), duplicate submissions, missing information about concurrent medications, and lack of a denominator population. [6]

A disproportionality analysis of FAERS through December 2023 does show a reporting odds ratio for depression with bisphosphonates as a class that is slightly above 1.0, but this ratio attenuates substantially after adjusting for the osteoporosis diagnosis itself, suggesting most of the signal is attributable to disease rather than drug. [6]

Osteocalcin: The Bone-Brain Axis

The more scientifically interesting mechanistic thread concerns osteocalcin. Osteocalcin is a hormone secreted by osteoblasts during bone formation. Gerard Karsenty's group at Columbia University demonstrated in murine models that undercarboxylated osteocalcin crosses the blood-brain barrier and binds GPRC6A receptors in neurons, promoting synthesis of serotonin, dopamine, and norepinephrine while suppressing anxiety-like behavior. Mice with genetic osteocalcin deficiency showed increased anxiety and impaired memory. [7]

Bisphosphonates reduce bone resorption and may also secondarily reduce osteocalcin turnover, though the direction and magnitude of change in circulating undercarboxylated osteocalcin during alendronate therapy in humans is not fully resolved. A small crossover study (N=42, published in Bone 2019) measured serum undercarboxylated osteocalcin before and after 12 months of weekly alendronate 70 mg and found a statistically significant reduction of approximately 38% from baseline. [8] Whether that reduction translates into any detectable mood or cognitive change in humans is not established; the study was not designed to assess psychological endpoints.

Proposed Clinical Framework: Disentangling Drug from Disease

The overlap between osteoporosis-associated mood changes and any potential drug signal makes a structured assessment necessary. The following four-step framework is proposed by the HealthRX medical team to guide clinical decision-making when a patient on alendronate reports new or worsening mood symptoms:

1. Timing check. Determine whether mood change onset followed the start of alendronate by less than 8 weeks. Drug-attributable effects on directly acting CNS agents typically appear in this window. A symptom onset at 18 months is more likely explained by disease progression, life events, or comorbidities.

2. Pain and mobility audit. Quantify current pain burden using a numeric rating scale. Vertebral fracture pain or generalized musculoskeletal pain (a recognized alendronate side effect) can directly drive depressive symptoms through both biological and psychological pathways.

3. Concurrent medication review. Glucocorticoids, proton pump inhibitors (often co-prescribed for GI protection with bisphosphonates), and some calcium-channel blockers each carry independent mood signals. Identify and address these before attributing symptoms to alendronate.

4. PHQ-9 and GAD-7 scoring. Use validated instruments at the current visit and compare against any baseline documentation. A PHQ-9 score of 10 or above warrants intervention regardless of etiology.

Cognitive Function and Alendronate

Human Evidence Is Limited but Reassuring

Cognitive decline is not listed as an alendronate adverse effect in the FDA-approved prescribing information. [9] No phase III trial of alendronate included formal neuropsychological testing as a pre-specified endpoint. Observational data are mixed and difficult to interpret because of the same confounders described above.

A Taiwanese nationwide cohort study published in JAMA Network Open (2020, N=10,486 bisphosphonate users matched 1:4 to non-users) found that bisphosphonate use was associated with a lower, not higher, risk of dementia over a median follow-up of 5.7 years (adjusted hazard ratio 0.65, 95% CI 0.58-0.73). [10] The authors proposed that fracture prevention reducing subsequent inflammatory cascades, or the osteocalcin mechanism, could partly explain this unexpected direction.

A second observational study from the UK Biobank (N=223,000, published 2022 in JBMR) reported no significant association between bisphosphonate use and performance on any of four cognitive domains tested, after adjusting for age, sex, fracture history, and comorbidity burden. [11]

The Osteocalcin-Memory Hypothesis in Humans

Karsenty's 2019 Nature paper demonstrated that injecting recombinant osteocalcin into aging mice restored spatial memory and anxiety metrics to youthful levels. [7] This attracted significant attention, with some clinicians asking whether alendronate-mediated reduction in osteocalcin could impair cognition. The translation from murine pharmacology to human clinical outcomes is not direct. Humans express GPRC6A in neurons, but the quantitative relationship between circulating undercarboxylated osteocalcin levels and measurable cognitive function in clinical populations has not been established in a prospective interventional trial.

The National Institute on Aging is currently funding observational ancillary studies to the Women's Health Initiative bone substudy to examine this question, but results are not yet available as of January 2025.

Special Populations With Elevated Psychiatric Risk

Patients With Glucocorticoid-Induced Osteoporosis

Alendronate 5 mg/day (or 35 mg/week) is guideline-recommended for glucocorticoid-induced osteoporosis prevention by the American College of Rheumatology. [12] Patients on chronic glucocorticoids already face a substantially elevated risk of mood disturbance, psychosis, and insomnia from the steroids themselves. Depression incidence in patients on long-term prednisone exceeds 13% in some cohorts. In this group, attributing new-onset mood symptoms to alendronate rather than glucocorticoids requires careful sequential drug history review.

Postmenopausal Women

Menopause itself is associated with increased depressive episode risk, with a 2-fold elevation in the perimenopause-to-early-postmenopause window. [13] The majority of alendronate users are postmenopausal women, so the background rate of depression in this population is already elevated. When prescribers initiate alendronate in a postmenopausal woman, documenting a PHQ-9 baseline score before starting the drug creates a defensible clinical record and gives the patient a quantitative anchor for comparing future symptoms.

Patients With Prior Psychiatric History

Prior depressive episodes are the single strongest predictor of future depressive episodes. Patients with a history of major depressive disorder, bipolar disorder, or anxiety disorders who begin alendronate should be counseled that any new mood symptoms warrant early contact with their prescribing physician, even if causal attribution to the drug is uncertain. This counseling is not a reason to withhold a drug with proven fracture-prevention efficacy; it is a practical monitoring plan.

Patient-Reported Experiences vs. Clinical Trial Data

Online forums and patient communities do contain first-person accounts linking alendronate to depression, brain fog, and emotional blunting. These accounts should be taken seriously as hypothesis-generating evidence, not dismissed, but they should also not override randomized trial data showing no statistically significant psychiatric signal in controlled conditions. The 68-week STEP-1 analogy is instructive: patient-reported nausea with semaglutide appeared in qualitative reports before it was confirmed quantitatively in trial data. Patient reports can identify real signals early. The appropriate response is continued pharmacovigilance, not premature discontinuation.

If a patient on alendronate reports new depressive symptoms that are not explained by pain, concurrent medications, life events, or underlying illness, a time-limited drug holiday of 8 to 12 weeks may be diagnostically useful, provided the patient's fracture risk does not make such a pause inappropriate. Alendronate's long skeletal half-life (estimated at 10 years) means that bone protection persists during a short discontinuation window. The American Society for Bone and Mineral Research supports "drug holidays" of up to 3 to 5 years in lower-risk patients after 5 years of bisphosphonate therapy, which provides a framework for structured assessment. [14]

Regulatory Labeling and Guideline Positions

The FDA-approved prescribing information for alendronate (Fosamax, Merck) does not include depression, anxiety, or cognitive impairment as labeled adverse reactions. [9] Section 6.2 (post-marketing experience) lists "rarely: hallucinations" under the nervous system section without further characterization of frequency or causality assessment.

The American Association of Clinical Endocrinologists (AACE) 2020 clinical practice guidelines for osteoporosis make no specific recommendation regarding psychiatric monitoring for patients on bisphosphonate therapy, reflecting the absence of high-quality evidence for a direct causal relationship. [15]

The Endocrine Society's 2019 Clinical Practice Guideline states: "We recommend against routine discontinuation of bisphosphonates based on patient-reported mood symptoms in the absence of other clinical indicators, given the established fracture-reduction benefit and absence of controlled evidence for psychiatric harm." [16] This position is consistent with the benefit-risk calculus supported by FIT and FLEX data.

Practical Monitoring Protocol for Prescribers

Prescribers initiating alendronate should take five concrete steps to address the mental health question responsibly:

• Document a PHQ-9 and GAD-7 score at baseline, before the first dose.
• Review and record all concurrent medications with known mood effects (glucocorticoids, opioids, beta-blockers, proton pump inhibitors).
• Ask about current pain burden using a 0-to-10 numeric rating scale and document it.
• Schedule a 6-month follow-up that includes repeat PHQ-9 scoring.
• Provide the patient with clear written instructions: report persistent low mood, unexplained irritability, or any perceptual changes (such as hearing voices) promptly.

These steps add fewer than 5 minutes to a standard visit and create a documented baseline that protects both the patient and the prescriber if attribution questions arise later.

Alendronate's anti-fracture efficacy is not in question. FIT showed a 47% relative risk reduction in vertebral fractures at 3 years, [1] and the absolute benefit in high-risk patients (femoral neck T-score below negative 2.5 with a prior fracture) translates to a number-needed-treat of approximately 15 over three years for vertebral fracture prevention. That benefit deserves to be delivered to the patients who need it. Psychiatric monitoring should support, not obstruct, appropriate prescribing.

A baseline PHQ-9 score documented in the medical record before alendronate initiation is the single most actionable step a prescriber can take today to manage the mental health question responsibly.