What Is Polypharmacy? Health Risks and Deprescribing

Polypharmacy, most commonly defined as using five or more medications at the same time, affects a growing share of the population and roughly half of older adults worldwide.¹ The term sounds clinical, but the reality is stubbornly ordinary: a blood-pressure pill, a statin, a diabetes drug, a proton-pump inhibitor for reflux, and maybe a mild pain reliever, and you are already at five. The issue is not that any single prescription is wrong. It is that medications interact, side effects compound, and nobody’s body handles drug number seven the same way it handled drug number one.

What Counts as Polypharmacy

There is no single universal cutoff. A systematic review of how researchers define the term found that the most common threshold was five or more daily medications, used in just under half of the studies examined. But definitions in the medical literature range from as few as two medications to eleven or more.² When someone is on ten or more drugs, the literature often calls that “hyperpolypharmacy,” a tier associated with sharply worse outcomes.³

The five-medication line is admittedly arbitrary. Someone taking five well-chosen, well-monitored drugs with no interactions may be in better shape than someone taking three drugs that clash. Researchers acknowledge this, and some newer frameworks try to distinguish “appropriate polypharmacy,” where every medication has a clear justification and the combination has been reviewed, from “problematic polypharmacy,” where drugs pile up without coordination. In practice, though, the numerical shorthand persists because it is easy to measure at scale, and because the five-drug mark is roughly where interaction risk starts climbing in a way that gets clinicians’ attention.

How Common It Is and How Fast It Is Growing

In the United States, the share of all adults taking five or more medications roughly doubled in two decades, rising from about 8% in 1999–2000 to 17% by 2017–2018. Among adults 65 and older, the jump was even steeper, going from roughly 24% to 44% over the same period.⁴ That means nearly half of older Americans were managing five or more prescriptions by the late 2010s. Even among younger adults aged 20 to 39, polypharmacy went from nearly nonexistent to about 3% of that age group.

The pattern is not identical everywhere. In parts of Asia, polypharmacy rates have been climbing steadily, with Hong Kong and South Korea showing particularly sharp increases. But in Australia and the United Kingdom, the trend has actually reversed slightly, with modest year-over-year declines.⁵ That divergence suggests that rising polypharmacy is not inevitable. Health-system design, prescribing culture, and national deprescribing initiatives all play a role.

Why Medication Lists Keep Getting Longer

The simplest driver is multimorbidity. A person with high blood pressure, type 2 diabetes, osteoarthritis, and gastric reflux will likely end up on a separate medication for each condition, and that is before anyone prescribes a preventive statin or a low-dose aspirin. Clinical practice guidelines are built around single diseases, so following the recommended treatment for each condition independently can quickly stack up. When researchers modeled what would happen if a hypothetical patient’s doctors followed all relevant UK guidelines, the regimen ballooned from 4 separate medications for a new patient to 11 medications and 20 daily dose units for someone whose conditions resisted first-line treatment. None of the guidelines discussed when to stop a drug.⁶ A separate analysis of UK guidelines reached the same conclusion: current recommendations “rapidly cumulate to drive polypharmacy” without telling clinicians how to prioritize when the total burden becomes overwhelming.⁷

Another underappreciated driver is the prescribing cascade, where a drug causes a side effect that is mistaken for a new medical problem, prompting yet another prescription. A blood-pressure medication causes swollen ankles, which gets treated with a diuretic, which drops potassium, which gets treated with a supplement, and so on.⁸ Each step makes clinical sense in isolation. The problem is that nobody steps back to ask whether the ankle swelling was drug-induced in the first place.

Supplements and over-the-counter drugs also contribute. Roughly one in four adults on prescription medications is simultaneously taking dietary supplements, creating interaction risks that often go unmonitored because patients do not think to mention supplements to their doctor, and doctors do not always ask.⁹ Nonprescription medications contribute to about half of major drug-interaction risks.¹⁰

What Happens Inside an Aging Body

Polypharmacy becomes particularly dangerous in older adults not just because they take more drugs, but because their bodies process drugs differently. With age, kidney and liver function decline, which slows how quickly drugs are cleared from the bloodstream. Fat-soluble drugs distribute into a larger volume of body fat, extending the time they linger in the system.¹¹ On top of that, older people tend to be more sensitive to several classes of drugs, including blood thinners, heart medications, and drugs that act on the brain.¹² The net effect is that a dose calibrated for a 50-year-old may produce a stronger and longer-lasting effect in a 75-year-old, and when five or ten drugs are doing this simultaneously, the interactions multiply in ways that are genuinely difficult to predict.

Sex also matters. Women tend to have higher blood concentrations and longer elimination times for the majority of FDA-approved drugs, and those differences strongly predict who experiences adverse reactions. In one analysis of 86 drugs, elevated blood levels in women predicted a higher rate of side effects in women for 96% of drugs where women had the higher levels.¹³ Given that older women are disproportionately represented in polypharmacy populations, this is a compounding vulnerability that standard dosing guidelines often ignore.

The Health Toll

The risks of polypharmacy are not abstract. A large cohort study found that older adults on five or more medications for at least six months had about a 30% higher chance of hospitalization and emergency visits compared to those not on polypharmacy, and a 63% higher chance of death. For those on ten or more drugs, the risk of death was more than two and a half times higher.¹⁴ These are adjusted numbers, meaning researchers accounted for the fact that sicker people take more drugs. Even after controlling for that, the medications themselves were driving worse outcomes.

Falls are one of the most common and dangerous consequences. Persistent polypharmacy increases fall-injury risk by roughly 30%, and when the medication list includes known fall-risk-increasing drugs, such as sedatives, certain antidepressants, and blood-pressure medications that can cause dizziness, the risk climbs to about 48% above baseline.¹⁵ The frustrating part is that even after someone is hospitalized for a fall-related fracture, medication lists often do not change much. One study of older adults who fractured a hip found that over three-quarters were on fall-risk-increasing drugs at admission, and at six-month follow-up the rates were virtually the same.¹⁶

Cognitive effects are another concern. Drugs with anticholinergic properties, a category that includes some antihistamines, bladder medications, and older antidepressants, are known to affect brain function. In a community sample of older adults in Italy, those with a high anticholinergic burden from their medications had more than three times the odds of mild cognitive impairment and more than four times the odds of poor executive function compared to those with a low burden.¹⁷ The insidious thing about anticholinergic effects is that they accumulate across medications. No single drug’s anticholinergic load may seem alarming, but three or four mild offenders together can push someone over the threshold.

Deprescribing and How It Works

Deprescribing is the structured process of tapering or stopping medications that are no longer necessary, no longer effective, or whose harms now outweigh their benefits. It is not the same as simply quitting a drug. Many medications, especially those that affect the brain, the heart, or hormone levels, require slow, monitored dose reductions to avoid withdrawal symptoms or rebound effects.¹⁸

Several screening tools help clinicians identify candidates for deprescribing. The Beers Criteria, maintained by the American Geriatrics Society, flag medications that are potentially inappropriate for older adults. The STOPP/START criteria, developed in Europe, take a somewhat different approach, catching medications that should be stopped (STOPP) and medications that should be started because they are being underused (START). In practice, the two tools have different strengths. In one study of older heart-failure patients, the Beers Criteria flagged potentially inappropriate prescriptions in about 76% of patients, while the STOPP criteria flagged them in about 29%.¹⁹ That does not mean one is better; they simply cast different nets. Using both together catches problems that either alone would miss.

Benzodiazepines, a class of anti-anxiety and sleep medications that are particularly risky in older adults, illustrate why deprescribing requires patience. A recent clinical practice guideline recommends reducing the dose by just 5 to 10% every two to four weeks, and never faster than 25% every two weeks. Even at that pace, some patients experience significant withdrawal symptoms and need the taper slowed further.²⁰ This is one reason why deprescribing rarely happens with a single appointment. It is an ongoing conversation between patient and clinician, with monitoring at each step.

Why Medications Are So Hard to Stop

Even when deprescribing is clearly beneficial, it often does not happen. In one study at a tertiary hospital in India, medication changes were deemed necessary for about 74% of elderly patients, but modifications were actually carried out for only about 36% of those cases.²¹ That gap between “should” and “did” is common across health systems. Computerized alerts have been tried, but the results are discouraging. In one study, electronic reminders about potentially inappropriate medications had to fire about 82 times before a single deprescribing event actually occurred.²²

Emotional barriers run surprisingly deep. Clinicians have described fear of being judged by patients or colleagues if they suggest stopping a medication, frustration when patients resist changes, and even feelings of incompetence when confronting a medication list they did not build. For patients, particularly those with cancer and other serious conditions, there is often a deep-seated belief that doctors are supposed to heal with medications, making the idea of removing drugs feel counterintuitive or even like giving up. Patients also worry about what might happen if a medication they have relied on for years is taken away.

Adherence is part of the picture too. In a study of older adults in Nepal, those with the most complex medication regimens had more than twice the odds of not following their prescriptions correctly. People managing four or more illnesses simultaneously were nearly four times more likely to be nonadherent compared to those with two conditions.²³ When a person cannot keep up with the regimen, missed doses and timing errors create their own set of risks, turning even an appropriate drug list into a hazard.

Pharmacist-Led Reviews and Technology

One of the more promising approaches to managing polypharmacy is having a pharmacist, rather than the prescribing physician, systematically review medication lists. In a randomized trial, pharmacist-led reviews with follow-up significantly reduced the number of uncontrolled health problems among older patients with polypharmacy. The most common issues pharmacists identified were undertreated conditions, poor adherence, and risk of side effects, and their interventions ranged from changing or adding a medication to educating patients about how to take their drugs properly.²⁴

A systematic review of pharmacist-led interventions across the UK and Ireland found that the large majority produced positive results: fewer inappropriate medications, fewer adverse drug reactions, improved prescribing quality, and reduced fall risk. However, none of the studies managed to show reductions in hospitalizations or deaths.²⁵ That gap is telling. Improving prescribing quality is achievable; proving it saves lives at a population level is a much higher bar that may require longer follow-up periods and broader implementation.

Clinical decision-support systems, essentially software that flags problematic medications and interactions, remain a work in progress. A scoping review found wide variation in how these systems are built and evaluated. Most are rule-based, relying on predefined criteria, and few have been rigorously tested for their impact on patient outcomes.²⁶ The promise is real, but the technology is not yet at the point where it reliably changes clinical behavior. As we saw with the 82-alerts-per-deprescribing-event figure, software recommendations are easy to ignore in a busy clinic.

The Financial Weight

Polypharmacy is expensive for patients and health systems alike. In a propensity-matched study of older Americans, those with polypharmacy had average annual healthcare spending of about $14,700, compared to roughly $8,900 for those without. The gap came from everywhere: higher prescription costs, more office visits, more inpatient stays, more outpatient visits. On top of the direct costs, people with polypharmacy were about 70% more likely to report that their prescription medications were unaffordable.²⁷

Targeted intervention programs can offset some of that cost. South Korea’s national polypharmacy management program for hospitalized older adults, which involves structured medication reviews during hospital stays, produced significantly lower hospitalization costs for participating patients and yielded a benefit-cost ratio of 3.8 from the payer’s perspective, meaning every dollar spent on the program returned nearly four dollars in savings.²⁸

Polypharmacy in Psychiatry

Mental health care has its own distinct relationship with polypharmacy. Psychiatric polypharmacy, defined as taking two or more psychiatric medications at once, is common and comes in several forms: two drugs of the same class (say, two antidepressants), drugs from different classes aimed at different symptoms, or augmentation strategies where a second drug is added to boost the effect of the first.²⁹ Some of these combinations have a solid evidence base. Others accumulate through trial and error, and nobody goes back to reassess once a patient stabilizes.

A trend worth watching is the rise in psychotropic polypharmacy among young adults. Between the late 1990s and the late 2010s in the U.S., the prevalence of taking two or more psychotropic medications roughly doubled, with young adults driving most of the increase.³⁰ This matters because the long-term effects of multiple psychotropic medications started in early adulthood are not well studied. The risks are not only pharmacological. Combining antipsychotic medications in psychiatric inpatients was linked to a 54% higher risk of experiencing a first adverse drug event and roughly double the risk of a second one.³¹

End-of-Life Care and Medication Inertia

One of the starkest illustrations of how polypharmacy persists inappropriately is in patients nearing the end of life. Among patients with advanced cancer receiving palliative care, the average number of medications hovers around six, and potentially inappropriate medications are prescribed to roughly half of them. The most common culprits are drugs like blood-pressure medications, cholesterol-lowering statins, and long-term acid suppressants, drugs that take months or years to produce benefits that the patient will never live long enough to receive.³²

The failure to deprescribe at end of life is not primarily a knowledge problem. Clinicians know these medications are unlikely to help. But inertia is powerful: renewing an existing prescription takes seconds, while stopping one requires a conversation, a clinical justification, and the emotional weight of implying that long-term health is no longer the goal. Research has identified critical gaps in medication management for end-of-life cancer patients and called for systematic deprescribing protocols to better align treatment with care goals.³³ This is an area where medication review at the point of care transition, such as when a patient enters hospice, could make a meaningful difference both in reducing pill burden and in redirecting attention toward comfort-focused care.

Specific Drug Combinations and Risk Prediction

Not all polypharmacy carries the same risk. Certain drug combinations are far more dangerous than others, and predicting which patients will end up in the hospital is an active area of research. An analysis of primary-care patients aged 65 to 100 found enormous variability in the risk tied to specific medication combinations. Patients in the highest-risk group for their particular drug mix had more than seven times the odds of an adverse-drug-related or emergency hospital admission compared to those in the lowest-risk group.³⁴ That sevenfold difference underscores that simply counting the number of medications misses most of the picture. Two patients can each take eight drugs, and one can be at dramatically higher risk based purely on which eight drugs those are and how they interact in that person’s body. The development of machine-learning tools that flag specific high-risk combinations, rather than just high drug counts, is one of the more potentially useful directions in this space, though the technology has not yet matured to the point of broad clinical adoption.³⁵