Between the years 2010 and 2050, the percentage of the general population older than age 65 years will increase from 14.9% to 24.5% in developed countries and from 5.8% to 14.4% in developing countries.1,2 Perhaps more notably, the shift will be comprised of a growing number of individuals age 80 to 90 years. Over the same period, the number of people older than age 90 years will increase by 463% in developed countries and 954% in developing countries.1 These trends are particularly relevant to medication treatments in the elderly, as the rate of adverse drug events is estimated to be 2 to 3 times higher in older patients than in those younger than age 30 years.3 Furthermore, it is estimated that as much as one-fifth of all hospital admissions of older adults are attributable to adverse drug effects.4
“Start low and go slow” has been a frequently invoked principle in geriatric pharmacology. Unfortunately, this strategy can lead to unfortunate consequences, such as when symptomatic elderly patients remain on inappropriately low doses despite tolerating the medications and having an inadequate response to treatment. Nevertheless, it is reasonable to initiate treatment at low doses given the considerable chance of developing adverse effects and frank toxicity when medications are started. Aging is accompanied by alterations in metabolic efficiency, protein structure, glycation, oxidation cell damage, and blood vessel distensibility. These changes lead to diminished blood flow to the kidneys, liver, and brain; to changes in drug metabolism; and to numerous other pharmacokinetic and pharmacodynamic factors related to drug effects.5 Beyond advancing age, the impact of comorbid medical conditions linked with aging is often a crucial factor in geriatric psychopharmacology, leading some to focus on the distinction between the fit and frail elderly.2,6 This article reviews the most common pharmacokinetic and pharmacodynamic changes associated with aging and discusses their relevance to geriatric psychopharmacology.
Often defined simply as the way that drugs move through the body, pharmacokinetics includes the processes of absorption, distribution, metabolism, and excretion. By affecting drug levels in blood and tissue, pharmacokinetic processes influence drug effects.7
The overall surface area of the intestinal epithelium, the rate of gut motility, the amount of splanchnic blood flow, and the level of gastric acid secretion all decrease with advancing age. The extent of drug absorption, often depicted as bioavailability or the area under the curve, appears to be largely unaffected in the elderly in the absence of overt disease. However, the rate of absorption, largely mediated by passive diffusion (which partially determines the onset of action of a drug), may be reduced by a decline in gastric emptying, a decrease in intestinal motility, and by subtle changes in the function of the upper esophageal sphincter.5,7,8 Furthermore, many common disease conditions (such as stroke, Parkinson’s disease, and multiple sclerosis) have been shown to decrease gastric motility as well, further modifying the impact of age on absorption.5,9
Once thought to be a predominately passive process, absorption is now conceptualized as consisting of both passive and active components, with the latter also being noteworthy. The cells lining the gut (known as enterocytes) express the drug-metabolizing enzyme CYP 450 3A4 and P-glycoprotein (P-gp), both of which work to regulate active drug absorption.7 P-gp, an efflux pump, is embedded in the intestinal wall and is an important determinant of how much drug is passed to the liver and the general circulation.7 Although the evidence is mixed on the specific influence of aging on the function CYP 3A4 and P-gp, both are involved in important drug and dietary interactions that are particularly relevant to the elderly because they are often taking multiple medications.10 At the intestinal wall, P-gp is particularly influential; inhibitors of its function are associated with increased serum drug concentrations as a result of reduced efflux function, whereas P-gp inducers are associated with decreased serum drug concentrations as a result of increased efflux function.7
Psychotropics that are substrates of P-gp in the gut include tricyclic antidepressants (TCAs) (eg, amitriptyline and nortriptyline), selective serotonin reuptake inhibitors (SSRIs) (eg, sertraline, paroxetine, citalopram), serotonin-norepinephrine reuptake inhibitors (SNRIs) (eg, venlafaxine), antipsychotics (eg, quetiapine, olanzapine), and anticonvulsants (eg, carbamazepine, topiramate).7 Inhibitors of P-gp include TCAs (eg, amitriptyline, imipramine), antipsychotics (eg, haloperidol, risperidone, paliperidone), SSRIs (eg, paroxetine, fluoxetine, sertraline), other medications (eg, disulfiram), as well as garlic, grapefruit juice, and green tea.7 These interactions are of particular concern among the elderly, as they are more likely to be susceptible to adverse consequences (eg, falls related to orthostatic hypotension, bradykinesia). Lastly, P-gp inducers include amitriptyline (which can also act as an inhibitor), trazodone, and St. John’s wort (which also induces CYP 3A4).7
Although it remains unclear whether P-gp function at the intestinal wall decreases with age, data from positron emission tomography scans in older individuals have shown reduced P-gp function at the blood-brain barrier.7,11 This reduced efflux potential could contribute to the increased sensitivity to centrally acting drugs and to toxins that is often observed in the elderly.7
Overall, age-related factors associated with medication absorption are largely related to changes in the rate, but not necessarily the extent, of medication absorption through passive absorption and to the potential for drug and dietary interactions and age-related changes that affect CYP 3A4 and P-gp function, as both are involved in the active aspects of drug transport.
Multiple physiologic changes in the elderly lead to alterations in the distribution of medications through the body after absorption from the gastrointestinal tract. These changes include shifts in body mass composition (the peripheral storage sites in fat and muscle), reductions in body total water, a possible decrease in the concentration of the serum binding protein (ie, albumin) that binds to acidic drugs (eg, valproic acid), and a possible increase in the concentration of the serum binding protein, alpha1-acid glycoprotein, which binds to basic drugs such as amphetamines.7
As the body ages, the percentage of body fat generally increases by 35%, the amount of lean body mass (muscle) decreases, plasma volume decreases by 8%, total body water decreases by 17%, and extracellular body fluid decreases by 40%.2,12 Moreover, the plasma concentration of a drug is inversely related to its volume of distribution, which in turn is dependent on the size of the hydrophilic and lipophilic spaces of the body.5,7 As body fat increases (lipophilic space increases) and total body water decreases (hydrophilic space decreases), the apparent volume of distribution of polar drugs (eg, lithium, ethyl-alcohol) decreases and that of lipophilic drugs (eg, diazepam) increases.10,13–16 Furthermore, the plasma half-life of a drug is related to its volume of distribution; a higher volume of distribution leads to an increase in the half-life and in the time to reach a steady-state concentration (which typically takes four half-lives).5 Although an increase in the ratio of adipose to lean body mass is typical in the elderly, some frail elderly patients experience a decrease in relative fat composition and, therefore, a decrease rather than an increase in the volume of distribution of lipophilic drugs.5,17
Metabolism refers to the bioconversion of a drug to another form, either an active or an inactive metabolite. Many drugs are metabolized in two phases. The first phase is a rate-limiting step, and the second phase is a step that produces polar metabolites available for renal excretion. Importantly, some drugs undergo only phase II metabolism. Many physiologic processes (including hepatic perfusion, activity of drug-metabolizing enzymes [ie, intrinsic drug clearance], and protein binding) affect liver metabolism.8 With advanced age, both liver size and mass decrease about 20% to 30%, and hepatic blood flow decreases by about 20% to 50%.13,18 Additionally, bile flow is reduced and the rates of synthesis of proteins, lipids, and glucose are diminished.5 Although these changes are largely universal, there is significant variability among individuals.7 Unlike other organs, such as the kidney, there are no specific age-related diseases of the liver; moreover, routine clinical tests of liver function do not change significantly with increasing age.5,14,16 In fact, currently used liver function tests are poorly correlated with drug-metabolizing capacity.7
Hepatic metabolism of drugs can essentially be divided into two main categories: those limited by hepatic blood flow (blood-flow limited); and those limited by intrinsic clearance of the drug (capacity limited). Blood-flow–limited drugs exhibit high liver extraction and, therefore, are metabolized to a lesser degree when hepatic blood flow rates decrease with advancing age.5,14,19 Examples relevant to geriatric psychopharmacology include TCAs. Capacity-limited drugs, on the other hand, exhibit low liver extraction, do not depend on liver blood-flow rates, and are instead dependent on the total tissue content of metabolizing enzymes, such as the cytochrome P450 (CYP 450) enzymes, uridine diphosphate glucuronsyl transferases (UGTs), or flavin-containing monooxygenases (FMOs).5,8,19 The metabolic clearance of these drugs by the liver (known as intrinsic clearance) is much less affected by old age. Capacity-limited drugs include diazepam and valproic acid.5
Studies of possible age-related changes in CYP 450 enzyme function have yielded variable results. In vitro studies on CYP 450 enzymes that participate in phase I oxidation reactions have shown no functional changes.14,20,21 However, those conducted in vivo on the same oxidation reactions have suggested a decrease in activity with advancing age.14,22,23 Data from these in vivo studies suggested that the CYP enzymes with the greatest reduction in activity with aging are 1A2 (singular liver metabolism of clozapine, fluvoxamine, and olanzapine) and 3A4/5 (singular liver metabolism of alprazolam, buspirone, carbamazepine, midazolam, and trazodone), 2C8/9, and 2C19.5,7,14 It is noteworthy that 2D6 (singular liver metabolism of desipramine, risperidone, venlafaxine, paroxetine, and thioridazine) has shown no change in activity with aging.5,7 Of note, many other psychotropic medications (eg, haloperidol, diazepam) are metabolized by multiple CYP 450 enzymes and have complicated pathways that are much more difficult to predict.7
Although more information about the relevance of age-related changes in CYP enzyme activity is needed, current data suggest that the major influences on CYP enzyme activity are related to inherited genetic polymorphisms and with drug-diet interactions (metabolic inducers such as tobacco, chronic alcohol use, and St. John’s wort, or metabolic inhibitors such as grapefruit juice) and drug-drug interactions (metabolic inducers such as carbamazepine and phenobarbital, or inhibitors such as bupropion or fluvoxamine).5,7 These factors have been shown to influence the inter-individual variability of CYP activity to a greater degree than age alone.14,16,24 In addition, it seems likely that interactions between age and genetic polymorphisms will be found that are relevant to drug prescribing. For example, one study indicated that the effect of age on the haloperidol serum concentration to dose ratio depends upon the CYP2D6*10 genotype.10,14,25,26
In phase II reactions involving the FMOs, UGTs, and other non-CYP enzymes, oxygen is used only indirectly, and studies have shown little to no impairment in activity with age or with reduced oxygen delivery to hepatic cells.10 Likewise, important phase II pathways (such as glucuronidation, acetylation, and sulfatation) seem to be relatively well preserved in the elderly.14,23
Drug clearance, or the rate at which a drug is removed from the systemic circulation, is accomplished by the combined actions of liver metabolism (discussed above) and renal excretion. Most drugs and metabolites are excreted from the body at a rate that is primarily determined by the glomerular filtration rate (GFR).7 For many adults, the kidney loses up to 30% of its weight between the ages of 30 and 90 years; this leads to the loss of 60% or more of glomeruli, to patchy tubular atrophy, to interstitial fibrosis, and to arteriosclerosis.10 By age 80 years, the GFR for a given individual has often fallen to half of the rate that it was at age 30 years.7 However, these phenomena are not ubiquitous—almost one-third of individuals display no decrease in renal function with age, and a small subset even has an increase in creatinine clearance with aging.2,5,16 Nevertheless, most investigators acknowledge the significant impact of external factors (eg, smoking) and common diseases (eg, hypertension, congestive heart failure, diabetes) on the reduction in renal function and GFR with aging.14 These confounding factors, which are prevalent in the elderly, make large-scale studies difficult to design and to execute.
Whether the original cause of altered kidney function is aging alone or whether it is the result of acquired pathology, the reduction in GFR is a phenomenon observed in the majority of adults.3,7,10 The impact of this change is witnessed in the pharmacokinetics of both water-soluble drugs and their metabolites, with a resultant increased steady-state drug concentration, a greater therapeutic benefit, and a higher risk for adverse reactions.7,8,27 Consequently, it is extremely important to have quick and reliable methods to measure GFR accurately in the elderly. As muscle mass decreases with age, creatinine becomes a less reliable marker of GFR.8 The Cockcroft-Gault formula (which assumes a linear decline in renal function) has been used to estimate GFR using the patient’s serum creatinine level, age, gender, and weight. However, this equation has been shown to underestimate the GFR, leading to under-dosing of medications in the elderly.2 More recently, a better-validated equation with improved accuracy (that accounts for the nonlinear decline of GFR with aging) was developed from the Modification of Diet in Renal Disease Study (MDRD);2 this equation uses the Cockcroft-Gault variables and also accounts for ethnicity.2,7
Changes in renal function are most relevant to psychiatric medications that are water-soluble and not metabolized by the liver (eg, lithium, gabapentin, pregabalin) and those with water-soluble metabolites (eg, risperidone, bupropion, venlafaxine).7 Studies are mixed on the effect of age (as a variable separate from renal pathology) on the clearance of drugs such as lithium. For example, there was significant overlap between age groups in three prominent studies that showed lithium clearance in the young to average between 0.6 and 2.4 L/h, whereas the same values in the elderly ranged from 0.83 to 0.94 L/h. Nevertheless, when administering drugs that are particularly reliant on renal excretion in the elderly, it is wise to reduce the initial dose and/or increase the dosing interval to minimize the risk for adverse reactions and toxicity.7
Whereas pharmacokinetics is defined as the movement of a drug through the body, pharmacodynamics refers to a drug’s effect at its target organ and receptors. Aside from the local drug concentration at a given receptor (the consequence of pharmacokinetic mechanisms), the magnitude of a drug response depends on parameters such as receptor density, receptor affinity, signal transduction pathways, and cell counter-regulatory processes.5,10 Data on pharmacodynamics are limited by the paucity of studies conducted in older individuals and by the limited number of longitudinal studies looking at changes in pharmacodynamic effects of advancing age within the same subjects.8
Despite the limitations of existing data, pharmacodynamic factors are clearly relevant to the practice of geriatric psychopharmacology. The mass and volume of the brain decrease with age (at a rate of around 5% per decade after the age of 40 years), with changes observed much more in the gray matter than the white matter.5,28 Neurotransmitters and receptors in the brain are also adversely affected by advancing age. Several types of dopamine receptors, alpha- and beta-adrenergic receptors, and serotonin 5-HT2A receptor levels decrease steadily in many individuals. Even in the absence of pathology (eg, Alzheimer’s disease), there is an age-related reduction in cholinergic innervation throughout the brain.28 As described above, P-gp functionality is often decreased in the elderly, leading to increased blood-brain barrier permeability and to elevated concentrations of drugs in the central nervous system.7,8
Homeostatic and regulatory function is also impaired in association with aging. Examples include compromised circulatory response, diminished vascular stability, decreased laryngeal reflexes, and reduced thirst response.28 In the setting of these changes in homeostatic and regulatory function, cells generally require longer intervals to recover from metabolic perturbations (eg, caused by the administration of medications), and lead to an increased incidence of toxicity and adverse events.5 Benzodiazepines, one of the most frequently prescribed classes of medications, are particularly affected by changes in pharmacodynamic (as well as pharmacokinetic) mechanisms in the elderly. In general, older adults need one-third to one-half the plasma concentration of benzodiazepines of younger adults to achieve the same drug response.14 Those benzodiazepines (eg, alprazolam, chlordiazepoxide, diazepam) that are metabolized by phase I oxidation reactions in the liver are more affected than those (eg, oxazepam, temazepam, lorazepam) that are not.29 Benzodiazepines are associated with an increased risk for falls and for hip fractures in the elderly that is likely secondary to the effect of the drugs on alertness, cognition, gait, and balance.28 Consequently, benzodiazepines should be used cautiously in the elderly, and the risks and benefits of prescribing should be assessed frequently.
Antipsychotics, another commonly prescribed class of medication, result in an increased incidence of adverse drug reactions (eg, extrapyramidal symptoms, orthostatic hypotension, untoward anticholinergic effects) in the geriatric population.8 Although these adverse reactions are observed more often with use of conventional rather than atypical antipsychotics, these reactions are associated with all antipsychotic medications.10 This phenomenon appears to be largely the result of changes in pharmacodynamics, as older patients experience extrapyramidal side effects, such as bradykinesia, more frequently when D2 receptor occupancy rates are similar to those observed in younger patients. It is noteworthy, however, that certain adverse effects (such as dystonic reactions) are more common in younger individuals.28 An influential study of risperidone showed that older patients also experience adverse effects with drugs at lower D2 occupancy rates.30 Finally, all antipsychotic medications are associated with an increased risk of cerebrovascular events and with all-cause mortality (likely from orthostatic hypotension, thromboembolic effects, or excessive sedation).28
Antidepressants are also associated with unwanted side effects in the elderly. TCAs and the SSRI paroxetine exhibit significant anticholinergic properties to which older patients are particularly vulnerable.28 SSRIs, in general, have also been found to increase the risk for gastrointestinal bleeding in the elderly, which is likely a result of their anti-platelet activity.7,28 This is particularly true when SSRIs are combined with other medications that increase bleeding risk, such as nonsteroidal anti-inflammatory drugs or warfarin. Additionally, studies have shown an association between SSRIs and SNRIs, the syndrome of inappropriate antidiuretic hormone secretion, and subsequent hyponatremia in up to 12% of older adults.28,30 Lastly, recent genetic data on the serotonin transporter among geriatric individuals suggests that certain polymorphisms may predict a patient’s responsiveness to SSRIs as well as their likelihood of experiencing adverse drug events.30
As the elderly will account for an increasingly larger percentage of the world’s population, and as recognition of the value of optimal psychiatric care of this population continues to grow, having a better understanding of the pharmacokinetic and pharmacodynamic factors in psychopharmacology among older adults is imperative.
As described in this review, age-related changes occur across the four major subcategories of pharmacokinetics: absorption, distribution, metabolism, and excretion. As aging occurs, the rate (but not extent) of absorption tends to decrease, generally slowing the onset of action of some drugs. Distribution is altered most significantly by changes in body composition with age—lipophilic drugs build up in fat tissues, have lower serum concentrations related to a greater volume of distribution, and have longer half-lives; whereas hydrophilic drugs, will, in general, have higher serum concentrations and shorter half-lives. Data are mixed on hepatic metabolic changes with aging; phase I oxidation reactions (by CYP 1A2, CYP 3A4/5, 2C8/9, and 2C19) appear to be the most affected by advancing age, although the impact of age-related changes on drug metabolism in the elderly may be relatively less important than the impact of drug-drug interactions related to drug metabolism. This is particularly true for older individuals with comorbid medical illness, given the ubiquity of psychiatric and general medical polypharmacy in this population. Excretion is affected significantly by advancing age, as the majority of individuals develop a gradual decline in their GFR over time, and this is a particularly important factor for drugs (such as lithium) that are not metabolized by the liver. In addition, aging is associated with important changes related to pharmacodynamics, including a reduction in receptor density, a lessened receptor affinity, a reduced cholinergic innervation, a decreased response to signal transduction pathways, and a less robust cell counter-regulatory and homeostatic response. Although “start low and go slow” remains an important guiding principle in geriatric psychopharmacology, increased study of pharmacokinetic and pharmacodynamic changes across the life cycle will undoubtedly lead to a more refined and evidence-based approach to treatment.