Overview

This section summarizes evidence related to the prevalence, assessment and treatment of depression in patients with:

• Cardiovascular disease and cerebrovascular disease
• Diabetes
• Chronic pain
• Geriatric and cognitively impaired patients
• Pregnant and postpartum women

Medical Comorbidity

The importance of the interplay between depression and many medical comorbidities cannot be overstated. Depressed patients often have comorbid conditions.

In the STAR*D trial, study entry subjects had an average of 3.3 general medical conditions (Trivedi, 2006b).

A study utilizing the second cohort of STAR*D patients reported a prevalence of significant general medical conditions of 50% in the study population (Yates, 2007).

A long list of medical conditions has been associated with increased risk for depression; these include chronic pain, diabetes, cancer, HIV, Parkinson's disease, cardiovascular and cerebrovascular disease, and multiple sclerosis (Kozhimnanil, 2009; Egede, 2005; Katon, 2004b).

Undiagnosed or undertreated depression has been associated with worsened outcomes in cancer, cardiovascular disease and other conditions (Hedeyati, 2010; Lichtman, 2008).

Conversely, one would expect that effective identification and treatment of comorbid depression would be associated with improved medical outcomes. Studies have demonstrated an association between effective treatment of depression and improved adherence to medical treatment for conditions such as cardiovascular disease (Ciechanowski, 2000). However, other suspected benefits of antidepressant therapy, such as decreased mortality after MI or CABG, have been more difficult to prove. See the "Implementation Tools and Resources Table" for more information.

The following conditions are particularly important for screening, given the findings.

Cardiovascular and Cerebrovascular Disease

Interplay of risks

Some studies have shown that major depression is associated with an increased risk of developing coronary artery disease (Wulsin, 2003; Rugulies, 2002).

Studies have shown an increased risk of mortality in patients after myocardial infarction by as much as fourfold (Lichtman, 2008; Frasure-Smith, 1995), while other analyses have disputed this (Jiang, 2005; Nicholson, 2006).

Moderate to severe depression before CABG surgery and/or persistent depression after surgery increases the risk of death after CABG more than twofold higher than non-depressed patients (Blumenthal, 2003).

Depression is three times more common in patients after acute myocardial infarction than in the general population and, notably, young women are at particularly high risk for depression after myocardial infarction (Lichtman, 2008).

A meta-analysis in regards to the relationship between stroke and depression found a pooled hazard ratio of 1.45, on par with the association smoking and obesity have with stroke (Pan, 2011).

Potential explanations

Several possible mechanisms are proposed to explain why depression increases the risk of developing cardiovascular disease. These include behavioral issues such as increased smoking, obesity, sedentary lifestyle, and lack of adherence to medication.

A prospective study found that the association between depression and cardiovascular events disappeared after controlling for physical activity and other health behaviors (Whooley, 2008), suggesting depression's negative impact on activity and behavior may account for its contribution to cardiac risk.

Biologic phenomena associated with depression such as increased inflammatory processes (elevated C-reactive protein or cytokine levels), increased platelet dysfunction (heightened platelet aggregation or adhesiveness), and abnormalities in endothelial function may also explain possible mechanisms for an increased risk (Katon, 2004b).

A cross-sectional study of depressed patients also found that, of their depressive symptomatology, specifically increased sympathetic arousal and insomnia were significantly associated with cardiac disease (Fraguas, 2007).

It has been suggested that the potential causative mechanisms for the association between stroke and depression are similar to those discussed above for cardiovascular disease (Pan, 2011).

Treating depression in this population

A recent meta-analysis suggested that SSRI treatment of depression may improve coronary heart disease prognosis (Pizzi, 2011). In addition, consensus opinion is to treat depressed cardiac patients with a safe drug rather than watchful waiting since they would not only benefit from symptomatic relief of their depressive symptoms but also have a potential improvement in their cardiovascular risk profile (Ballenger, 2001).

Although tricyclic antidepressants are effective against depression, they are associated with cardiovascular side effects including orthostatic hypotension, slowed cardiac conduction, proarrhythmic activity and increased heart rate. SSRIs, by contrast, are well tolerated and have a more benign cardiovascular profile; they would be preferred initial agents for treatment of depression in individuals with cardiovascular disease (Jiang, 2005). SSRI treatment also has been shown to significantly decrease depressive symptoms in coronary heart disease patients (Pizzi, 2011).

The American Heart Association science advisory (Lichtman, 2008) suggests sertraline and citalopram as first-line drugs for patients with coronary heart disease. See "SSRIs and Other Antidepressants" and "Citalopram Warning" in the Pharmacotherapy section.

A recent meta-analysis showed that treatment with SSRIs after stroke in non-depressed patients significantly reduced the risk of development of a depressive episode after stroke (odds ratio [OR] of 0.34). In addition to this, it has been shown that an ongoing exercise program after stroke also significantly reduces depressive symptoms post-stroke (Eng, 2014).

For more information, see also the ICSI Heart Failure in Adults guideline and Stable Coronary Artery Disease guideline.

Diabetes

Major depression is associated with an increased number of known cardiac risk factors in patients with diabetes and a higher incidence of coronary heart disease; therefore, screening and treatment of depression in this patient group should be emphasized (Petrak, 2013; Katon, 2004b).

Individuals with diabetes have two to threefold higher odds of depression than those without diabetes (Jeeva, 2013). Additionally, depression earlier in life increases the risk of developing diabetes (Katon, 2004a). Depressive symptom severity is associated with poor self care and medication compliance in addition to higher health care-related costs (van Dijik, 2013). Patient physical and mental quality of life is also decreased (Petrak, 2013).

High levels of symptoms associated with diabetes that do not correlate with physical or laboratory assessments should prompt the physician to assess for depression (Ludman, 2004).

Treatment goals should focus on improvement of glycemic control and improvement or remission of depression via pharmacologic therapy, psychological therapy or combination of both (Petrak, 2013).

For more information, see the ICSI Diagnosis and Management of Type 2 Diabetes Mellitus in Adults guideline.

Chronic Pain

Depression and pain symptoms commonly coexist, exacerbate or attenuate one another, and appear to share biological pathways and neurotransmitters.

Patients with chronic pain are more likely to have coexisting depression. In 2004, data were examined from primary care centers worldwide by the World Health Organization. They found that 22% of all primary care patients suffer from chronic debilitating pain. Further, chronic pain patients were four times more likely to have comorbid depressive disorder than pain-free primary care patients (Lépine, 2004). The findings also showed that the more diffuse the pain complaints, the greater the risk of depression and the bigger impact on the quality of life.

For more information, see the ICSI Assessment and Management of Chronic Pain guideline.

Geriatrics

Depression in the elderly is widespread, often undiagnosed and usually untreated. It is a common misperception that depression is a part of normal aging. Losses, social isolation and chronic medical problems that older patients experience can all contribute to depression.

Rates and presentation

Out of 31 million elderly Americans (65 years and older), nearly 5 million have clinical depression and 1 million have major depression (Birrer, 2004). The rate of depression in adults older than 65 years of age treated in primary care settings ranges from 17 to 37% (Birrer, 2004) and is between 14 to 42% in patients who live in long-term care facilities (Robinson, 2014). Comorbidities are more common in the elderly. The highest rates of depression are found in those with strokes (30 to 60%), coronary artery disease (up to 44%), cancer (up to 40%), Parkinson's disease (40%), Alzheimer's disease (20 to 40%) and dementia (17 to 31%) (Birrer, 2004). The recurrence rate is also extremely high at 40% (Birrer, 2004).

Similar to other groups, the elderly with depression are more likely than younger patients to underreport depressive symptoms. Major depression in the elderly can present with combined clinical symptoms including depressed mood, somatic complaints and sleep disorders (Chen, 2011). This incudes symptoms such as insomnia, headache, dizziness, appetite disturbances, lack of energy, fatigue, chronic pain, constipation and musculoskeletal disorders.

Treatments and outcomes

The outlook for recovery for the elderly is similar to that for the young when appropriately treated. However, treatment usually has to be continued for longer periods than for the young, since it may take longer to reach remission.

Collaborative care. The IMPACT (Improving Mood: Promoting Access to Collaborative Treatment) study showed improvement in treating the depressed elderly over several measures. Patients were randomized to usual care or collaborative care. The latter involved a team composed of a depression care manager, primary care physician and psychiatrist. The team offered education, behavioral activation, antidepressants, brief behavior-based psychotherapy (problem-solving treatment), and relapse prevention geared to each patient's needs and preferences (Unützer, 2002).

Outcomes from IMPACT included demonstration that collaborative care was more effective than usual care for the elderly, regardless of the elderly patients' ethnicity (Areán, 2005). The intervention group also showed improved physical functioning, less suicidal ideation, improved continuation of antidepressant treatment, fewer depressive symptoms, remission of depression, plus increased quality of life, self-efficacy and satisfaction with care. The intervention lasted for one year. One year later the outcomes for the intervention group were still significantly better than for those who received usual care (Hunkeler, 2006).

Pharmacotherapy and psychotherapy. Pharmacotherapy and psychotherapy are appropriate modalities to treat depression in the elderly (Birrer, 2004). When using pharmacotherapy, the physician must carefully consider how the metabolism of the drug may be affected by physiologic changes in the elderly, their comorbid illnesses and the medications used for them (Ganzini, 1997; Reynolds, 1992). Selective serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants (TCAs) do not have significant differences in efficacy; however, in the elderly, the better option for treatment may be SSRIs due to potential side effects of TCAs (Kok, 2012). In those individuals who do not respond to the different antidepressants alone, augmentation therapies may be appropriate (Nierenberg, 1996). This would include psychostimulants, such as cytomel or methylphenidate, or the addition of lithium (Lavretsky, 2015; Cooper, 2011). Psychotherapy is also appropriate, limited only by cognitive impairments (Cuijpers, 2008). Cognitive and/or behavioral therapy in particular can have a significant impact on major depression compared with other forms of non-pharmacological interventions such as interpersonal psychotherapy, psychodynamic therapy and physical exercise (Pinquart, 2007).

Behavioral strategies. Behavioral activation strategies such as increasing daily involvement in pleasant activities are safe, simple and beneficial in treating depression in this population (Cuijpers, 2007).

Maintenance therapy. With a recurrence risk of 50, 70 and 90% after the first three episodes respectively, once a patient reaches remission, data supports that continuing with antidepressants is effective regardless of patient age (Kok, 2011; Birrer, 2004). After the first episode of major depression, therapy should be continued for at least one year after remission. The second and third episodes should be treated for at least two and three years after remission (Birrer, 2004). Another study showed that maintenance therapy with an SSRI (paroxetine) for two years was shown to be effective in preventing recurrent depression after a first-time major depression in the elderly over 70 years of age (Reynolds, 2006). See also Annotation #7d, "If Patient is Not Improving on Initial Treatment, Utilize Stepped Care Approach," "Consider Other Strategies," for electroconvulsive treatment.

Dementia/cognitive impairment

Assessment. Patients with more severe cognitive impairments cannot reliably answer the PHQ-9 questions. The 19-item Cornell Scale for Depression in Dementia (CSDD) has the best sensitivity (93%) and specificity (97%). A cutoff of greater than or equal to six identifies depression in a demented population (Alexopolous, 1988). This is a clinician-administered tool to help diagnose depression in patients with dementia. It has been used in a variety of settings ranging from outpatient to assisted living to nursing homes. Its accuracy decreased when it was modified to be used by less-trained staff. Its usefulness for ongoing tracking purposes has not been studied (Barca, 2010; Watson, 2009).

Interplay of risks. There is reasonably good evidence that having a major depressive episode increases the risk of developing Alzheimer's dementia (odds ratio of 2.03 with 95% confidence, with a range of odds ratio of 4.55 with 95% confidence ratio when depression occurred less than one year before diagnosis of Alzheimer's dementia to odds ratio of 1.71 when depression occurred more than 25 years earlier) (Ownby, 2006; Green, 2003).

Pregnant and Postpartum Women

Prevalence of depression during pregnancy can vary depending on how depression in pregnancy is defined. DSM-5 acknowledges depression during pregnancy (peripartum onset), which is consistent with recent publications that identify perinatal depression beginning at pregnancy onset or the first 12 months after birth. These periods previously referred to as antenatal depression and postpartum depression. A systematic review by Gaynes, 2005 suggests that between 14 and 23% of pregnant women and 10-15% of postpartum women will experience a depressive disorder (Gaynes, 2005). A review by Milgrom, 2014 cites a point prevalence of 13% at three months after delivery and an average of 9% during each trimester of pregnancy (Milgrom, 2014). According to a large-scale epidemiological study by Vesga-López, 2008, depression during the postpartum period may be more common than at other times in a woman's life (Vesga-López, 2008).

With growing understanding of the systemic impact of perinatal stressors, there is a new body of research examining paternal depression. A recent meta-analysis shows a 10-14% incidence of paternal depression during the perinatal period, with a moderate positive correlation with maternal depression (Paulson, 2010).

Untreated prenatal depression has been associated with negative pregnancy outcomes such as low birth weight and preterm labor, as well as negative effects on children such as developmental delay and cognitive impairment (Milgrom, 2014; Davalos, 2012; Li, 2009). A study of pregnancy-associated suicide in women demonstrates pregnant women with mental health problems are at an increased risk of substance abuse and intimate partner problems (Gold, 2012). Studies are demonstrating that untreated paternal depression has an impact on infant and child development similar to untreated maternal depression (Paulson, 2010).

Screening

Two key strategies facilitate early intervention: routine screening and monitoring of known risk factors (O'Hara, 2014). Increased acceptability is found when patients are educated regarding routine screening (Milgrom, 2014). A large scale study by Kaiser Permanente (Dietz, 2007) found that of those women identified and treated for depression, more than half had recurring indicators for depression. Key risk factors include:

• Previous history of a mood disorder
• Depression or anxiety during pregnancy
• Poor social support
• Stressful life events
• Fragmented or poor sleep
• Substance use
• Past or current abuse
• Premorbid or gestational diabetes
• Difficulty breastfeeding in the first two months postpartum

(Watkins, 2011; Coelho, 2011; Lancaster, 2010; Dørheim, 2009; Goyal, 2009; Pearlstein, 2008)

The United States Preventive Services Task Force 2016 recommendations for depression screening recommend that pregnant and postpartum women be screened for depression even in the absence of treatment protocols, care managers and specialty trained providers (Siu, 2016).

Routine use of a self-report screening instrument that has been validated among pregnant women does not supplant clinical diagnosis (Yonkers, 2009). However, it significantly increases the incidence of systematic case finding over spontaneous detection during routine clinical evaluation (Gjerdingen, 2007). Routine maternal screening is highly recommended, followed by a clinical interview of those scoring above threshold (Yonkers, 2009).

Validated tools for screening pregnant and postpartum women. "Edinburgh Postnatal Depression Scale (EPDS)" and Patient Health Questionnaire (PHQ-2 or PHQ-9).

A 1987 validation study by Cox et al. that included 84 mothers showed that EPDS had satisfactory sensitivity and specificity, and was also sensitive to change in the severity of depression over time. The scale can be completed in about 5 minutes and has a simple method of scoring (Cox, 1987).

PHQ-2 or PHQ-9 questionnaire can also be used to screen pregnant and postpartum women for depression. An observational study by Gjerdingen, 2009 that included 506 women to determine the validity of PHQ-2 and PHQ-9 for identifying postpartum depression during well child visits found that the two-question screen was highly sensitive and the PHQ-9 was highly specific for identifying postpartum depression. Both screens can be easily administered in primary care clinics (Gjerdingen, 2009).

A study of 81 pregnant and 104 postpartum patients by Flynn et al. 2010 on comparative performance of EPDS and PHQ-9 in pregnant and postpartum women seeking psychiatric services for depression found few significant differences in the performance of the PHQ-9 and EPDS in detecting clinician-diagnosed MDD in a psychiatry outpatient sample of pregnant and postpartum women (Flynn, 2011).

Treatment

Psychotherapies

Psychotherapeutic treatment recommendations for mild to moderate perinatal depression are interpersonal therapy (IPT) and cognitive behavioral therapy (CBT) (Cuijpers, 2011a; Cuijpers, 2011b; O'Hara, 2000). Successful IPT treatment of antenatal depression has also improved functioning for six months postpartum. Existing literature clearly suggests that IPT and CBT are more efficacious than routine care for postpartum depression (Cuijpers, 2011a; Cuijpers, 2011b; O'Hara, 2009).

Pharmacotherapy

The recommendation for moderate to severe perinatal depression is antidepressant medication in combination with supportive interventions or psychotherapy (Stewart, 2011). Since partial SSRI treatment during pregnancy does not successfully treat the depression, it is not a recommended option (Wisner, 2009). Clinicians should be cautious in disrupting maintenance antidepressants during pregnancy. In a study of antidepressant discontinuation for pregnant women with a history of recurrent major depression, 68% relapsed, compared with 26% who maintained antidepressant treatment (Cohen, 2006).

Light therapy and integrative medicine

There is promising preliminary evidence for bright light therapy, acupuncture, progressive relaxation, music therapy, reduced sleep deprivation, and exercise (Pearlstein, 2008). Evidence for omega-3 fatty acids is still insufficient; however, they pose little to no risk (Freeman, 2008).

Other

Hormonal treatments such as estrogen or progesterone have not shown clear evidence for efficacy for postpartum depression, and in some cases may worsen symptoms (Pearlstein, 2008).

See the "Implementation Tools and Resources Table" for perinatal decision-making tools and clinical algorithms.

Safety assessment of psychotropic medication during pregnancy and lactation

Treatment of a psychiatric illness during pregnancy involves weighing potential risk of fetal exposure to psychotropic medication against potential adverse effects of an untreated disorder on mother and fetus. In conclusion, there is no zero-risk option. Clinicians must help patients assess these negative effects of depression on mothers and families against the risks and benefits of psychotropic medication and other treatment options (Mian, 2005).

The process of making decisions about the use of any medicine, particularly psychotropics, during pregnancy should be made on a case-by-case basis, weighing the varying amounts of information about the medicine and the patient's underlying disease state.

The available evidence about psychotropic medication in pregnancy is substantial but limited by ethical considerations that preclude prospective controlled trials of pregnant women. Most studies retrospective or reliant on databases that do not allow for accuracy in the determination of fetal exposure or other confounders. We will provide a more global risk assessment across pregnancy, including the perinatal period and lactation. Medications taken during pregnancy are considered teratogenic if they increase the risk of congenital malformations above the baseline risk of 3 to 4%. The most reproductive safety information is available for the tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) (Ferreira, 2007; Mian, 2005; Sivojelezova, 2005).

Among the available pregnancy data, TCAs and SSRIs have not shown any evidence of increased risk of major congenital malformations, with the possible exception of paroxetine. In 2006 the FDA issued a warning that first-trimester paroxetine was associated with an increased risk of major malformations (4% vs. 3%), particularly cardiac malformations (2% versus 1%). The FDA changed the pregnancy labeling from category C to D, indicating that controlled or observational studies in pregnant women have demonstrated a risk to the fetus. Studies have suggested that first-trimester exposure to paroxetine at doses greater than 25 mg a day are associated with a greater risk of cardiac malformations (Bérard, 2007; Thormahlen, 2006).

Based on these findings, paroxetine should not be considered a first-line choice for initiating an antidepressant in pregnancy. For women who are already on paroxetine and planning pregnancy, the risks of paroxetine should be weighed against the risks of discontinuing it. For some high-risk women, severe depression or anxiety could also adversely affect the pregnancy.

Factors to consider when choosing an antidepressant (Ray, 2014).

• If there is a history of positive response to medication prior to pregnancy it is recommended to continue with the same medication. This includes effective antidepressants with unplanned pregnancies.
• Data is limited regarding new or improved medications in the peripartum period. Consider starting older medications if there is not a history of prior antidepressant use as more data is available for safety assessment and education.

Impact of antidepressants on neonates

Prenatal exposure to antidepressants has been associated with transient symptoms of possible medication withdrawal or toxicity in neonates (Austin, 2006). These neonatal syndromes have been described with most TCAs, SSRIs and non-SSRIs and can include jitteriness, irritability, breathing difficulties, bowel obstruction and urinary retention. These symptoms are transient and possibly confounded by physiologic effects from maternal depression and anxiety or other medications administered during delivery (Ferreira, 2007; Austin, 2006; Oberlander, 2006; Sivojelezova, 2005).

Potential persistent pulmonary hypertension of the newborn (PPHN)

While studies have evaluated a possible association between SSRI exposure after 20 weeks' gestation and persistent pulmonary hypertension of the newborn (PPHN), in 2011, the U.S. Food and Drug Administration issued a notification that "given the conflicting results from different studies, it is premature to reach any conclusion about a possible link between SSRI use in pregnancy and PPHN." The FDA advisory committee suggested that health care professionals should "weigh the small potential risk of PPHN that may be associated with SSRI use in pregnancy against the substantial risks associated with undertreatment or no treatment of depression during pregnancy." (http://www.fda.gov/Drugs/DrugSafety/ucm283375.htm) (Kieler, 2011; Austin, 2006; Bérard, 2007; Chambers, 2006).

Breastfeeding while taking antidepressants

For women with depression who require antidepressants, breastfeeding and remaining on medication can be highly compatible ways of caring for themselves and their infants. Clinicians can support nursing mothers with depression by helping them weigh the risks and benefits of different treatment options including supportive interventions and medication if indicated (Davanzo, 2011). If choosing to use antidepressant during breastfeeding, the same antidepressant during pregnancy should be maintained because evidence regarding risk with breastfeeding, excludes use of a different antidepressant during pregnancy (Ray, 2014).

Clinicians should advise nursing women on psychotropic medications to monitor infants for behavioral changes such as excessive sedation, jitteriness or inconsolable crying. Infants who develop these symptoms should be evaluated by their clinician for possible drug toxicity.

For infants who are premature or have any medical problems, mothers on psychotropic medication who choose to breastfeed could consider pumping and storing/discarding breast milk until the infant is healthy and can metabolize medication more efficiently.

Consultation with a pediatrician or neonatologist may be warranted.

SSRIs and Autism

A recently published observational study of 145,456 infants suggested the maternal use of SSRIs during the second and third trimester may increase the risk of autism spectrum disorders of their infants. The women using antidepressants were older and were more likely to have had a previous child with autism spectrum disorder. Women who used SSRIs during the first trimester or during the year prior to their pregnancy did not have an increased risk of ASD (Boukhris, 2016). Additional research and evaluation of this risk is needed before definitive correlations can be made.