Section 7: Special Populations: Anxiety in Older Adults

Late-life anxiety is the most prevalent psychiatric condition in the geriatric population, and it is systematically under-recognized, under-treated, and frequently managed with medications that make it worse.

7.1 Prevalence and Under-Recognition

Anxiety is the most common psychiatric condition in older adults, more common than depression, and yet it is the one most likely to be dismissed as "just aging." A 2024 meta-analysis found a global prevalence of clinically significant anxiety symptoms among older adults of approximately 28% (95% CI: 21-35%). Anxiety disorders are more common than depression in this population. (Yan et al., 2024, Journal of Affective Disorders)

Despite this prevalence, treatment rates are abysmal. Data from the 2022 National Survey on Drug Use and Health indicate that among older adults with any mental illness, treatment engagement rates are among the lowest of any demographic group. In primary care settings, anxiety detection rates fall well below prevalence estimates. (SAMHSA, 2023, NSDUH)

The reasons for under-recognition are both patient-side and provider-side. Older adults internalize ageism, normalizing fear, restricted mobility, and insomnia as inevitable consequences of aging. Providers, in turn, attribute anxiety symptoms to medical conditions ("Of course she's anxious, she has COPD") or to "normal aging," a pattern of diagnostic nihilism that denies these patients effective treatment. (Bryant et al., 2008, Journal of Affective Disorders)

7.2 How Anxiety Presents Differently in Aging

The classical anxiety phenotype (excessive worry about future events, with prominent psychological distress) shifts in older adults in three clinically important ways:

Somatic predominance. Older adults are more likely to present with physical complaints (insomnia, fatigue, chronic pain, GI distress, dizziness, chest tightness) than with the cognitive symptoms of worry. Standard screening tools that rely on endorsement of psychological terminology may miss these patients.

Shifted worry content. The worry focus shifts from career, social, and financial concerns to health deterioration, loss of independence, becoming a burden, and fear of falling. Fear of falling is a particularly potent driver of functional decline: it leads to activity restriction, social withdrawal, deconditioning, and paradoxically increased fall risk.

Blurred boundaries with cognitive decline. Anxiety in older adults overlaps phenomenologically with early neurocognitive disorder (concentration difficulties, memory complaints, executive dysfunction). Differentiating anxiety-driven cognitive complaints from prodromal dementia is a common and high-stakes clinical task.

7.3 The Anxiety-Dementia Axis

Perhaps the most significant recent finding in geriatric psychiatry is the clarification of the relationship between chronic anxiety and dementia risk.

The Hunter Community Study, a prospective longitudinal cohort, found that participants with chronic anxiety exhibited a nearly 3-fold increased risk of developing all-cause dementia (HR = 2.80, 95% CI: 1.35-5.72), with an even larger magnitude of risk for vascular dementia specifically. Among participants under age 70, chronic anxiety was linked to a 4.58-fold increased risk. (Khaing et al., 2024, Journal of the American Geriatrics Society)

Several biological mechanisms could explain this connection. Chronic anxiety keeps the stress hormone system (HPA axis) running at high levels, and sustained cortisol exposure is directly toxic to the hippocampus, the brain structure most critical for memory formation. Chronic anxiety also drives low-grade brain inflammation through the same inflammatory pathways (IL-6, TNF-alpha) implicated in cardiovascular damage. Both of these processes may accelerate the accumulation of the amyloid proteins associated with Alzheimer's disease.

However, a critical interpretive caveat applies: the causal direction is not definitively established. Anxiety may be an early symptom of a dementia process that has already begun, rather than a cause of that process. This "protopathic bias" concern also complicates the literature on benzodiazepines and dementia discussed below: if anxiety and insomnia are themselves early prodromal symptoms of dementia, then the benzodiazepine prescriptions those symptoms generate may be markers of the disease, not causes of it.

In Plain Language for Families:

Anxiety in an older adult does not look like anxiety in a younger > person. Instead of saying "I feel worried," an older person might > say "I can't sleep," "My stomach is always upset," or "I'm afraid

of falling." These are not just normal parts of aging. They are > treatable symptoms of a treatable condition. If your older family > member has become more withdrawn, more focused on health complaints, > or more reluctant to leave the house, ask their doctor about > screening for anxiety, not just medical causes.

7.4 Pharmacotherapy: Efficacy and Geriatric-Specific Risks

The pharmacotherapy principles for late-life anxiety are the same as for younger adults with two critical additions: the drug interaction burden of polypharmacy and the heightened vulnerability to specific adverse effects that can mimic the very conditions being treated.

SSRIs remain first-line. A 2025 meta-analysis of 19 RCTs (2,336 older adults) published in The Lancet Psychiatry confirmed that antidepressants significantly reduce anxiety symptoms compared to placebo (standardized mean difference = -1.19, 95% CI: -1.80 to -0.58, indicating a large effect). SSRIs led to significantly greater anxiety reductions than SNRIs in planned subgroup analyses. (Neil-Sztramko et al., 2025, Lancet Psychiatry)

Start low, go slow. Initiate SSRIs at half the standard adult starting dose. Sertraline and escitalopram are generally preferred in older adults due to their relatively clean CYP450 profiles. Avoid paroxetine and fluoxetine, which are potent CYP2D6 inhibitors that can dramatically alter serum levels of commonly co-prescribed medications (beta-blockers, antiarrhythmics, anticoagulants) in the polypharmacy environment that is baseline in geriatric medicine. (Neil-Sztramko et al., 2025, Lancet Psychiatry)

SSRI-induced hyponatremia: a silent geriatric threat. A 2024 meta-analysis of 39 studies (8.1 million patients) found that antidepressant exposure increases hyponatremia risk more than three-fold. SSRIs carry a 5.59% event rate and SNRIs a 7.44% event rate. Among SSRIs, fluoxetine carries the highest risk for hyponatremia (upper 95% CI bound = 5.008), followed by paroxetine (OR = 4.489) and citalopram (OR = 3.598). This risk stratification further supports the preference for sertraline and escitalopram in older adults, as these agents combine favorable CYP450 profiles with lower relative hyponatremia risk compared to fluoxetine and paroxetine. (Letmaier et al., 2024, European Psychiatry)

Practical protocol informed by these risk data: obtain a baseline basic metabolic panel before SSRI initiation. Recheck serum sodium at 2-4 weeks. This is especially critical for patients concurrently taking thiazide or loop diuretics, which independently lower sodium. Symptoms of hyponatremia (confusion, falls, nausea, headache) can mimic worsening anxiety or early dementia, making the diagnosis easy to miss. For patients with documented hyponatremia history, mirtazapine is explicitly designated as the antidepressant with the lowest associated hyponatremia risk. (Letmaier et al., 2024, European Psychiatry)

7.5 The Benzodiazepine Question in Older Adults

Benzodiazepines remain widely prescribed in geriatric populations despite decades of safety concerns. The evidence for this population requires honest presentation of both the risks and the comparative context.

Fall and fracture risk. Meta-analyses consistently show that benzodiazepine use increases fall risk, with elevated risk regardless of whether the agent is short-acting or long-acting, traditional benzodiazepine or Z-drug. (Seppala et al., 2018, JAMDA)

"Iatrogenic aging." Long-term benzodiazepine use induces cognitive symptoms that mimic dementia progression: anterograde amnesia, psychomotor slowing, word-finding difficulty, and blunted processing speed. These effects can be difficult to distinguish from true neurocognitive decline. (Barker et al., 2004, Drugs & Aging)

Mortality risk. Daily benzodiazepine use in older adults is associated with significantly higher all-cause mortality. A 2024 national case-control study of 139,103 nursing home residents with dementia found that new benzodiazepine initiation significantly increased 180-day mortality risk. The elevated mortality risk appears at least partially reversible upon cessation. (Gerlach et al., 2024, JAMA Network Open)

The comparative safety context. The data on benzodiazepine risks in the elderly must be interpreted alongside comparative data on SSRIs. As discussed in Module 3, Section 2.4, Seppala (2018) found that SSRIs carry an OR of 2.02 for falls compared to benzodiazepines' OR of 1.42. (Seppala et al., 2018, JAMDA) Bolton demonstrated that SSRIs have an OR of 1.45 for fractures versus benzodiazepines' OR of 1.10. (Bolton et al., 2008, Journal of Clinical Psychopharmacology) SSRIs have direct effects on bone density via serotonin receptors on osteoblasts that benzodiazepines lack. This comparative data does not make benzodiazepines safe in the elderly; it means that the medications recommended as replacements carry their own substantial risks that must be monitored with equal vigilance.

The Beers Criteria listing. The AGS Beers Criteria lists benzodiazepines as potentially inappropriate medications for older adults. Module 3 provides the full comparative safety analysis and the framework for when deprescribing is appropriate, when maintenance with monitoring carries less risk than forced discontinuation, and how to have the honest risk-benefit conversation with patients already on long-term therapy.

The NNH-to-NNT ratio. The NNH for benzodiazepines treating insomnia in older adults at 2 weeks is approximately 63 for any harm, compared to an NNT of 13 for any improvement. This ratio is unfavorable and supports the consensus recommendation against initiating new chronic benzodiazepine prescriptions in this population. (Glass et al., 2005, BMJ)

Clinical Pearl:

For every older adult started on an SSRI, check sodium at baseline > and again at 2-4 weeks. SSRI-induced hyponatremia mimics the very > symptoms you are treating (confusion, falls, malaise) and is easily > missed in a population already at risk. This one lab draw can > prevent a hospitalization.

Pause and think: An 82-year-old woman with moderate GAD, osteoporosis, and mild cognitive complaints (normal neuropsychological screen) is referred to you. She has been taking lorazepam 0.5 mg nightly for 15 years, prescribed by her now-retired internist. Her daughter wants you to "get her off the Ativan because I read it causes dementia." Based on the evidence in this section, how would you approach this situation?

See a model response

The daughter's concern is understandable but the evidence requires honest presentation. The anxiety-dementia association exists (HR = 2.80), but the causal direction is not established. Both the anxiety and the benzodiazepine prescription may reflect protopathic bias: early prodromal symptoms of a disease process already underway.

The risks of continued lorazepam are real. Fall risk is elevated regardless of agent type. The cognitive effects of long-term use (anterograde amnesia, psychomotor slowing, word-finding difficulty) can mimic dementia progression, making her cognitive complaints harder to evaluate while she remains on the medication.

However, forced discontinuation after 15 years carries its own risks. The comparative safety data show that the SSRIs recommended as replacements carry an OR of 2.02 for falls (versus benzodiazepines' 1.42) and 1.45 for fractures (versus 1.10). In a patient with osteoporosis, that fracture differential is clinically relevant.

The approach: a collaborative risk-benefit conversation, not unilateral deprescribing. If tapering is pursued, the process should be gradual over months. If transitioning to an SSRI, sertraline or escitalopram are preferred for cleaner CYP450 profiles and lower hyponatremia risk. Start at half the standard adult dose. Baseline sodium and recheck at 2-4 weeks.

7.6 CBT in Older Adults

CBT is effective for late-life anxiety, with response rates of approximately 39-42% (meaning roughly four in ten patients no longer meet diagnostic criteria after a full course of treatment). Unlike pharmacotherapy, where symptom recurrence is common after medication cessation, the cognitive restructuring and behavioral skills learned in CBT confer durable, long-term benefits. (CCSMH, 2024, Canadian Guidelines for Seniors Mental Health)

Adaptations for older adults include slower pacing, repetition of key concepts, incorporation of written summaries, involvement of family members in treatment planning, and accommodation of sensory deficits (hearing, vision). Telehealth delivery reduces the access barrier for patients with mobility limitations.

7.7 The Somatic Presentation Trap

When an older adult presents with escalating, treatment-resistant somatic complaints (chronic pain, insomnia, dizziness, GI distress, breathlessness) without corresponding objective findings, anxiety should be high on the differential. The somatic presentation of anxiety in older adults is covered in detail in Section 9.

PsychHQ

Special Populations: Anxiety in Older Adults