Advances in modern medicine have greatly increased the life span of people living in the United States. However, individuals living with serious mental illness have been left behind. A 2003 study revealed that patients with serious mental illness treated in the US public health sector die, on average, 25 years earlier than people in the general population (
1). Despite multiple calls to action to address this public health issue, the mortality gap remains, with a recent meta-analysis estimating that people with schizophrenia experience an average of 14.5 years of potential life lost (
2). Clearly, further interventions are urgently needed.
Yet finding a solution to this problem is difficult. Reasons for the mortality gap are complex and multifactorial, involving adverse health behaviors, stigma, medication side effects, poverty, and psychiatric and cognitive symptom burden. The situation is compounded by an overwhelming lack of access to ongoing quality medical care and a subsequent pattern of overuse of emergency services and underuse of primary care (
3). For many people with serious mental illness, the behavioral health clinic is their only contact with the medical system, and their psychiatrist is the only physician they regularly see. In recognizing this reality, the American Psychiatric Association published a position paper in 2015 urging psychiatrists to play a role in assessing, diagnosing, and treating the medical problems that contribute to the mortality gap among people with serious mental illness (
4).
Although rates of disease among people with serious mental illness exceed those in the general population for every disease category (
1), cardiovascular disease (CVD) plays a particularly important role in perpetuating the mortality gap. Rates of cardiovascular risk factors, including hypertension, tobacco use, obesity, dyslipidemia, and diabetes mellitus, are greatly elevated among those with serious mental illness (
5,
6). CVD is two- to threefold more common among people with serious mental illness and occurs at a younger age (
7). Overall, CVD is the most common cause of death among people with serious mental illness (
6). Because of this, addressing CVD and its risk factors is a reasonable initial goal for psychiatrists hoping to improve the medical health of their patients with serious mental illness.
Potential Interventions for Psychiatrists: A Sliding Scale
Of necessity, interventions initiated by psychiatrists will exist on a sliding scale, depending on availability of care referral resources, patient preference, and physician knowledge base. Graduates of combined residency programs (internal medicine-psychiatry and family medicine-psychiatry) are ideally situated to provide integrated care to complex patients with comorbid medical and psychiatric conditions, and the recent increase in the number of such programs in the United States provides hope for improving outcomes for patients with serious mental illness (
8). However, the number of physicians who complete combined residency programs is overall minimal compared with patient needs. If the mortality gap is to be adequately addressed, psychiatrists who have completed categorical residencies will need to intervene as well. Comfort with addressing medical health may require additional training or continuing medical education, although primary care skills are increasingly being emphasized in categorical psychiatry residency training. Recent psychiatry residency graduates affirm that their programs have adequately trained them to recognize and initiate treatment for common medical conditions (
9).
Increasing psychiatrists’ comfort with addressing physical morbidity also requires examination of the therapeutic frame, boundary crossings, and boundary violations in doing so. According to Freud’s initial description in 1911, the therapeutic frame consists of certain norms in the patient-doctor relationship that allow for a safe and contained structure in which to process communication. Among these norms is a lack of physical touch. Although a reductionist and rule-bound adherence to the frame is generally not helpful to the therapeutic dyad, and handshakes and hugs are no longer unheard of in the therapeutic relationship, careful attention to boundary fluidity is necessary. Practice setting, patient diagnosis, physician expertise, and treatment modality will all have an impact on the propriety of psychiatrists performing a physical exam to assess their patients’ physical health status. Risks and benefits must be carefully weighed. Physical examination is generally not recommended (outside of obtaining vital signs) in a practice in which there is no separate examination room and appropriate examination table. Likewise, physical examination of patients with unstable identity and porous boundaries, such as those with borderline pathology, may also be harmful to the therapeutic relationship. The risks of physical examination are likely much lower for psychiatrists treating patients solely with medication management. Fortunately, as I discuss, physical examination is not a necessary component of many of the interventions that psychiatrists can make to improve their patients’ medical morbidity and mortality.
Screening
At a minimum, psychiatrists should screen their patients for cardiovascular risk factors. Clear guidelines exist for screening for metabolic conditions associated with the prescription of second-generation antipsychotics (
Table 1) (
10). For patients not prescribed second-generation antipsychotics, screening recommendations for cardiovascular risk factors vary somewhat depending on organizational guidelines. A reasonable paradigm for cardiovascular risk screening is outlined in
Table 2 and incorporates recommendations from various evidence-based US organizations.
Screening for tobacco use is perhaps the most straightforward of the CVD risk factors. Physicians should be certain to ask about all forms of tobacco (including smoking, chewing, dipping, and vaping) and whether the patient has ever used tobacco, because some individuals may not consider themselves smokers if they use tobacco products intermittently (
11). All adults should be screened annually for obesity by calculation of body mass index (BMI) (
12). BMI is defined as weight in kilograms divided by the square of the height in meters. A patient with a BMI of 25–29.9 m/kg
2 is considered overweight, and one with a BMI of ≥30 m/kg
2 is considered obese. Practitioners must be diligent in avoiding a judgmental stance concerning weight. Rather than focus on appearance, physicians must focus solely on the health risks associated with increasing BMI.
Blood pressure screening is largely noncontroversial and inexpensive and can easily be done in the office. More reliable readings can be done by using a 24-hour ambulatory blood pressure monitor, but this may be logistically difficult to set up (
12). Daytime blood pressures higher than 130/80 mmHg are considered positive for hypertension (
13). Screening for elevated lipids should begin at age 40 for anyone without risk factors for CVD, but physicians could consider screening as early as age 17 if the patient has significant risk factors for CVD. Nonfasting samples are sufficient to determine the need for statin therapy for primary prevention of CVD.
Screening for diabetes, as for hyperlipidemia, requires access to laboratory services and is recommended for all adults ages 18 years or older who have at least one risk factor for diabetes (see
Table 2); for all others, screening is recommended at age 45. Screening is ideally done by measuring fasting blood glucose. Values of 100–125 mg/dL indicate prediabetes, whereas a fasting blood glucose measurement of 126 mg/dL or more indicates a diagnosis of diabetes. Of note, glucometers should not be used to diagnose diabetes because readings may differ by as much as 15% from those obtained via phlebotomy. If fasting values are unavailable (as a result of patients’ difficulty with fasting for laboratory collection), hemoglobin A1c (HbA1c) may be used. Values of 5.7%−6.4% indicate prediabetes, and values of 6.5% or more indicate diabetes. Of note, two HbA1c values are required to confirm the diagnosis of diabetes mellitus. If any abnormal blood glucose measurements are made, screening should be done annually. Among those with normal blood glucose on screening, repeat screening should occur at 3-year intervals.
Behavioral Counseling
Although screening is a start, it is not enough. Interventions are essential to have an impact on the mortality gap for patients with serious mental illness. Behavioral counseling is highly recommended by the American Health Association to decrease overall cardiovascular risk (
13). Given that the American College of Graduate Medical Education lists psychotherapy as one of the essential competency milestones obtained by psychiatry residents during training (
14), it is reasonable to assume that psychiatrists are capable of effectively counseling patients for behavioral change. For maximum effectiveness, such counseling should start with the use of motivational interviewing techniques to assess patients’ readiness to enact change. Inquiries should be nonconfrontational and encourage collaboration. Once patients are ready to change their behavior, collaborative plans may be enacted using various cognitive-behavioral exercises. A great deal of evidence supports the use of cognitive-behavioral therapy and mindfulness-based techniques to reduce cardiovascular risk factors, such as obesity and tobacco use (
15–
17). Multiple online resources, texts, manuals, and workshops are available to guide such structured psychological interventions, and some individual exercises may be easily incorporated into medication management appointments. These exercises include self-monitoring for exercise, food intake, and tobacco use; problem solving; short- and long-term goal setting; and the use of thought records and cognitive restricting (
18). In addition, psychiatrists may address external stressors, underlying psychiatric disorders, and psychological defenses that may make behavioral change difficult.
In addition to individual counseling for behavioral change, a robust literature supports group lifestyle interventions for improving cardiovascular risk (
19), and a growing body of evidence supports the benefits of such interventions for the population with serious mental illness (
20,
21). Psychiatrists may take the lead in supporting, promoting, organizing, or leading group lifestyle interventions offered in their own practice setting. When organizing such interventions, psychiatrists may look to the literature to identify elements associated with greater success; such elements include use of multiple components (exercise, dietary counseling, and behavioral therapy), personalization, more frequent contact, longer duration (longer than three months), manualized programs, active monitoring (of weight, diet, and exercise), and training of treatment providers (
19). Evidence has shown that the use of peers (individuals living with serious mental illness) as treatment team members in lifestyle interventions can also improve outcomes for patients with serious mental illness (
22).
In promoting lifestyle change, psychiatrists must be mindful to tailor interventions to the life circumstances of the patients they are treating. For individuals with serious mental illness, this often means actively addressing adverse social determinants of health (
23). In addition, because individuals with serious mental illness often experience low health literacy, low educational attainment, and cognitive impairment, programs may require simplification of language and content (
24).
Pharmacological Management
As prescribers, the initial step for psychiatrists in approaching the pharmacological management of cardiovascular risk in patients with serious mental illness may be guided by the Hippocratic adage “First, do no harm.” The metabolic effects of psychotropic medications, particularly the second-generation antipsychotics, are well established. Although all of the antipsychotics (including the first-generation agents) have been shown to be obesogenic, weight gain potential varies by individual medication. Among the second-generation agents, clozapine has the greatest risk, followed in descending order of magnitude by olanzapine, quetiapine, risperidone, amisulpride, aripiprazole, and ziprasidone (
28). Weight gained is clinically significant, with an average of 12 kg gained in 40%−90% of patients on olanzapine and up to 31.3 kg gained among patients on clozapine (
29). More recently approved second-generation antipsychotics appear to be less obesogenic, with asenapine associated with a 0.9-kg weight gain in the first three weeks of treatment and iloperidone with a 1.5- to 2.1-kg weight gain (
29). Medications used to treat affective disorders (particularly mood stabilizers) may also cause weight gain. Of patients prescribed valproic acid, 71% gain more than 4 kg, and 20% of individuals taking lithium gain more than 6.3 kg (
29). Antidepressants can also induce weight gain. Of patients taking antidepressants, 10%–20% experience treatment-emergent increases in weight, with mirtazapine and the tricyclic antidepressants consistently associated with more weight gain than the selective serotonin reuptake inhibitors (
29).
Weight gain resulting from psychotropic medication is likely due to interactions with a number of neurotransmitters and neural circuits. Mirtazapine and many of the second-generation antipsychotics (including olanzapine, quetiapine, and clozapine) are serotonin 2c antagonists; in mice, serotonin 2c receptor stimulation promotes anorexia, and mice lacking these receptors become obese. Histamine may help to centrally regulate satiety, and psychotropic medications that block the H1 receptor are associated with greater weight gain. In addition, anticholinergic effects of psychotropics may be obesogenic, whether directly through appetite stimulation or via side effects such as dry mouth (leading to caloric fluid intake) and sedation (inhibiting physical activity) (
29).
In addition to weight gain, second-generation antipsychotics are associated with lipid and glucose abnormalities. These metabolic changes are mediated in part by weight gain but have also been cited independent of adiposity. Olanzapine and clozapine confer the greatest risk of lipid abnormalities; quetiapine and risperidone confer intermediate risk. Elevations in triglycerides are most marked, but second-generation antipsychotics also appear to raise low-density lipoprotein and total cholesterol. Olanzapine and clozapine are also associated with the greatest risk of insulin resistance and glucose dysregulation.
The metabolic syndrome is made up of increased central adiposity, elevated triglycerides, low high-density lipoprotein (HDL), hypertension, and impaired glucose tolerance, and it is a major risk factor for cardiovascular mortality. Clozapine, olanzapine, and chlorpromazine are most closely associated with risk of metabolic syndrome. One meta-analysis showed a prevalence of metabolic syndrome in approximately 50% of patients taking clozapine (
30).
Those who are most at risk for metabolic abnormalities (particularly for weight gain) with second-generation antipsychotics are younger and treatment naive, although all patients should be appropriately counseled when starting psychotropic medications, and a plan for monitoring and healthy habits should be simultaneously prescribed. To decrease cardiovascular risk, polypharmacy should be avoided when possible, and medication with lower potential for adverse metabolic effects should be considered as initial therapy. A dose-dependent relationship between metabolic abnormalities and second-generation antipsychotics has been suggested, particularly for olanzapine (
30). Thus, dose reductions should be considered and medications tapered to the lowest therapeutic dose when feasible. In addition, limited data support starting metformin when treatment with second-generation antipsychotics is initiated, specifically those that are more obesogenic (
31). Metformin acts by inhibiting hepatic glucose production and thus has effects on both weight and glucose regulation. The most common side effect of metformin is gastrointestinal upset, including diarrhea and nausea. To improve tolerability and decrease side effects, metformin should be started at 500 mg/day and titrated to up to 2,000 mg/day in split doses. Lactic acidosis is a rare adverse effect of metformin and is more likely to occur with the accumulation of excess levels. Because metformin is renally cleared, it is therefore contraindicated for individuals with a glomerular filtration rate of less than 30 mL/min and should be held if an individual is to receive intravenous contrast dye (
32).
When medication-associated metabolic abnormalities do occur, psychiatrists may consider switching to a psychotropic agent with less metabolic risk while carefully weighing the potential for psychiatric decompensation. For patients who have gained more than 7% of pretreatment weight or have developed hyperglycemia, hyperlipidemia, or hypertension, physical benefits of switching drugs must be given strong consideration (
30). In addition, several medications have been evaluated as adjunctive strategies for antipsychotic-induced metabolic abnormalities. For treatment of weight gain, the largest body of evidence supports the use of metformin (
6,
33). A smaller body of data shows that the use of adjunctive aripiprazole, topiramate, reboxitine, or sibutramine is superior to placebo; however, clinically relevant weight loss (of ≥7%) was noted only with metformin and aripiprazole (
33). Psychiatrists are generally quite comfortable with prescribing the partial D2 antagonist-agonist aripiprazole; for a discussion of metformin prescribing, please see the preceding paragraph.
Limited data support the use of metformin and rosiglitazone for improving glucose intolerance (
33). Rosiglitazone is a thiazolidinedione that reduces glucose production and increases glucose clearance by improving insulin sensitivity and pancreatic beta-cell function. Rosiglitazone is associated with fluid retention and liver function abnormalities and is contraindicated in patients with New York Heart Association class III and IV heart failure and those with active liver disease or increased liver enzymes (≥2.5 times the upper limit of normal). Rosiglitazone should be started at 4 mg/day and can be titrated to 8 mg/day. Because rosiglitazone may be associated with mild weight gain and is more likely to cause hypoglycemia, metformin may be a better initial agent to prescribe for antipsychotic-emergent glucose dysregulation. In addition, liver enzymes must be monitored among patients prescribed rosiglitazone (
34).
Limited evidence supports the use of metformin, topiramate, or sibutramine to treat patients with antipsychotic-emergent dyslipidemia; in such incidences, these medications have been shown to decrease both total cholesterol and triglyceride levels (
33). The majority of psychiatrists are experienced with prescribing the anticonvulsant and mood stabilizer topiramate, and the logistics of metformin prescribing have previously been listed. Sibutramine is a selective reuptake inhibitor of serotonin, norepinephrine, and, to a lesser extent, dopamine. Sibutramine was approved as a weight loss medication by the Food and Drug Administration in 1997 but was pulled from the US market in 2010 because of an association with cardiovascular events.
In addition to prescribing medications to counteract adverse effects of second-generation antipsychotics, psychiatrists may elect to provide evidence-based pharmacological treatment for medical conditions associated with increased cardiovascular risk. In particular, pharmacological assistance with tobacco cessation greatly improves patients’ ability to successfully quit smoking. Therapies approved by the Food and Drug Administration for smoking cessation include nicotine replacement therapy (NRT), bupropion, and varenicline (
35); all three medications have been shown to improve success in nicotine cessation among patients with serious mental illness without worsening psychiatric symptoms (
36). NRT provides controlled doses of nicotine without the harmful substances found in tobacco products. It improves withdrawal symptoms but does not alleviate them completely because the release of nicotine into the central nervous system is much more gradual than with cigarette inhalation. NRT is available in many formulas, including patches, gum, oral and nasal sprays, lozenges, and inhalers. NRT can safely be used while patients are still using tobacco products; the instruction for patients not ready to quit should be to cut back as much as possible while using NRT. Dosing of NRT should correlate to the amount of tobacco used in a day. For example, when prescribing nicotine patches, anyone smoking 10 or more cigarettes per day should be prescribed a 21-mg patch. Basal delivery formulations (the patch) may also be combined with quick-release formulations (such as gum) to combat cravings. Nicotine patches may cause skin irritation and so should be avoided among patients with chronic dermatologic conditions. To reduce skin irritation, patients should be instructed to place the patch on a different area of the skin each day.
Psychiatrists are likely familiar with prescribing bupropion for patients with depression. Bupropion is a norepinephrine and dopamine reuptake inhibitor that may also act as a nicotine receptor antagonist. Bupropion should be prescribed for at least seven days before the quit date for tobacco cessation in order to build up therapeutic serum levels. Varenicline is a partial agonist of nicotinic acetylcholine receptors (with strongest affinity for alpha
4beta
2 receptors) and a full agonist of the serotonin type 3 receptors. Varenicline is believed to assist with withdrawal symptoms via its agonism of the nicotinic receptor and to prevent the reward of smoking via competitive inhibition of nicotine binding. As with bupropion, varenicline should be started one week before cessation attempt; titrating slowly can improve gastrointestinal side effects (as can taking the medication with a full glass of water). Combination therapy with NRT and bupropion may lead to greater success in quitting tobacco use (
35). Electronic cigarettes (e-cigarettes) have gained in popularity in recent years; however, a recent review found no evidence that e-cigarette use is associated with increased tobacco cessation for cigarette smokers (
37). In addition, recent studies show that dual use of e-cigarettes and tobacco products may be associated with greater risk of cardiovascular disease (
38).
Although a thorough discussion of the evidence-based treatment of other disorders associated with increased cardiovascular risk is outside the scope of this article, psychiatrists may increase their knowledge, comfort, and competency in prescribing by following treatment guidelines and pursuing continuing medical education opportunities.