Association Between Mental Health Burden and Coronary Artery Disease in U.S. Women Veterans Over 45: A National Cross-Sectional Study

Abstract

Background:

The women Veteran population accessing Veterans Health Administration (VA) care has grown rapidly. Women Veterans exhibit high rates of mental health conditions that increase coronary artery disease (CAD) risk; however, the relationship between specific conditions and increasing mental health burden to CAD in this population is unknown.

Materials and Methods:

Using VA National Patient Care Data for 2009, we identified women Veterans over 45 (N = 157,195). Logistic regression models examined different mental health diagnoses and increasing mental health burden (number of diagnostic clusters) as predictors of CAD.

Results:

CAD prevalence was 4.16%, and 36% of women Veterans were current smokers. Depression exhibited the strongest association with CAD (odds ratio [OR] 1.60, 95% confidence interval [CI] [1.50–1.71]), similar to that of current smoking (OR 1.68 [1.58–1.78]). Controlling for demographic variables, smoking, diabetes, and obesity, each additional mental health diagnosis increased the odds of CAD by 44%.

Conclusions:

Women Veterans over age 45 accessing VA care exhibited a high degree of mental health burden, which is associated with elevated odds of CAD; those with depression alone had 60% higher odds of CAD. For women Veterans using VA, mental health diagnoses may act as CAD risk factors that are potentially modifiable. Novel interventions in primary care and mental health are needed to address heart disease in this growing and aging population.

Introduction

The Veterans Health Administration (VA) experienced a doubling in the number of women Veterans accessing care between 2003 and 2012; women Veterans currently comprise 6.5% of all VA patients.¹ This rate of growth in the number of women Veterans using VA healthcare now exceeds that of men.¹ Of women Veterans using VA, the highest proportion of these patients are women ages 45–64,¹ the age range in which menopause typically begins and the risk for cardiovascular disease (CVD) increases. CVD is the fourth most common visit diagnosis in women VA users ages 45–64, and the second most common diagnosis for the 65 and older group.¹ Women Veterans have higher rates of diagnosed CVD than women in the general population.^2,3 Accordingly, VA has identified cardiovascular health as an important focus for the growing and aging women Veteran population.⁴

An understanding of potentially modifiable factors associated with CVD in this patient population is critical to addressing their elevated risk. Women VA users have a high prevalence of several CVD risk factors, including hypertension, hyperlipidemia, diabetes, and obesity,^5
–7 and are less likely than male Veterans to have low-density lipoprotein under control.^5,8,9

In addition, mental health conditions are prevalent among women Veterans.^2,10 Nearly half of women ages 45–64 who use VA receive a visit diagnosis in this domain,¹ and prior work demonstrates that these women Veterans have a higher risk of CVD when compared to women VA users without mental health conditions.¹¹ Relative to civilian women, women Veterans report higher rates of depressive and anxiety disorders,² as well as comparably limited physical activity.^2,12 Finally, women Veterans are more likely to smoke than are male Veterans¹³ and non-Veteran women.^2,3,14

Although a consistent relationship between CVD and mental health has been observed, the underlying mechanisms are complex and not fully established.^15,16 It is likely that both behavioral/lifestyle factors as well as biologic dysregulations underlie this linkage.¹⁷ Mounting evidence suggests that physiologic changes associated with mental health conditions contribute to CVD equally, if not more, than lifestyle factors do.^17,18

While these processes have not been well studied in populations of women Veterans, prior work has found that Veterans of both genders with mental health conditions experienced increased risk of myocardial infarction¹⁹ and had excess mortality from heart disease, even after adjustment for behavioral factors.²⁰ Among non-Veteran women with diabetes, major depression was an independent risk factor that accelerated development of CVD.²¹ The impact of multiple, or comorbid, mental health diagnoses on CVD has not been well studied, and it is not clear whether women Veterans with more than one such diagnosis are at even greater risk for CVD.

As the population of women Veterans grows and ages, exploration of the mental health-CVD connection among older women Veterans accessing VA care may help tailor treatment planning and inform early detection of individuals at elevated risk, which is critical if mental health multimorbidity increases odds of CVD. To investigate this association among women Veterans using VA care, we focused on one form of CVD, coronary artery disease (CAD), because it is the most common subtype of heart disease²² and provides a circumscribed, highly prevalent outcome for which to examine mental health associations. The relationship of mental health conditions as well as mental health burden (i.e., presence of two or more mental health conditions) on CAD in this cohort has not been described previously.

Our study aims were to (1) determine which mental health conditions have the strongest association with established CAD for midlife and older women Veterans accessing VA care and (2) quantify the effect of increasing mental health burden on CAD in this group.

Materials and Methods

Study population

Our study examined a national sample of female VA users. Inclusion criteria were Veteran status (some VA patients are non-Veteran spouses or employees), at least one VA outpatient visit in fiscal year 2009, and age older than 45 years. Transgender Veterans (n = 290) were excluded. The Institutional Review Boards of VA Boston and VA Connecticut Healthcare Systems approved the study.

Data sources

These VA administrative data were extracted from the National Patient Care Database (NPCD) Outpatient Clinic Files and Inpatient Patient Treatment Files located at the Austin Automation Center, TX; these databases provide demographic and clinical information on all patients in VA.

As detailed previously, ICD-9-CM codes were used to identify the presence of medical and mental health conditions for each subject.²³ We counted conditions that were coded at least once for an inpatient stay or twice for outpatient visits. Outpatient diagnostic codes are typically assigned by healthcare providers, whereas inpatient codes are assigned by professional coding personnel. This methodology has been commonly used in other studies.^24,25

ICD-9-CM codes for similar mental health conditions were aggregated into a smaller set of nine “cluster” variables. These cluster variables coded the presence of any ICD-9-CM code included in the following groupings: (1) posttraumatic stress disorder (PTSD), (2) other anxiety disorders, (3) depressive disorders, (4) bipolar disorders, (5) psychotic disorders, (6) alcohol use disorders, (7) substance use disorders, (8) somatization, and (9) eating disorders. These groupings have been used previously to model associations between medical conditions and mental health exposures.²³ “Mental health burden” was defined as the number of clusters present for each patient.

CAD was identified by the presence of any one relevant ICD-9 code (410.xx, 414.xx, 36.xx). We examined diabetes and obesity as control variables because both are strongly independently related to both mental health conditions²⁶ and CAD.^27,28 In addition, smoking data were derived from clinical reminder data according to a previously validated method.²⁹ Clinical reminders are prompts delivered to clinicians through the VA electronic medical record; multiple clinical reminders are in use and cover preventive care and chronic disease management.⁶ Using these data allowed us to estimate smoking prevalence rates derived from actual clinical encounters with women Veterans.

Because NPCD files code race and ethnicity separately, we constructed a consolidated race variable described previously.²³ Missing race values were replaced using imputation based on random forest recursive partitioning.³⁰

Statistical analyses

We fit logistic regression (LR) models to identify mental health morbidities associated with CAD, adjusting for covariates, and calculated odds ratios (OR) with 95% confidence intervals (CI). Two separate models were created. Model 1 included age (as a five-level categorical variable), race, smoking status, diabetes, and obesity as a baseline model, to which each mental health cluster variable competed for stepwise entry. Cluster variables were chosen for entry (or removal) based on the largest possible improvement in penalized model fit (BIC), with steps halting when fit could no longer be improved. Model 2 focused on mental health burden; it included the same baseline covariates, but the cluster variables found to be predictors by Model 1 were aggregated into a count of present conditions as a single explanatory variable. Analyses were conducted using R statistical software (version 3.2.2, R Core Team, Vienna Austria).

Results

We identified 157,195 women Veterans who met criteria for inclusion in the study. The prevalence of CAD was 4.16% (n = 6,540). As shown in Table 1, mean age was 59.4 years (standard deviation 12.2, range 46–110) and the majority (61%) of the sample was white. Over one-third (35.8%) reported current smoking and 18.7% were past smokers. The prevalence of diabetes and obesity was 14.0% (n = 22,054) and 10.9% (n = 17,116), respectively.

Table 1.

Demographics, Smoking, and Mental Health Characteristics

Variable	N (%)
Age (mean 59.4 years [SD 12.2])
<50	32,678 (11.3)
50–59	68,538 (43.6)
60–69	25,570 (16.3)
70–79	12,677 (8.1)
≥80	17,732 (11.3)
Race n (%)
White	95,878 (61)
Black	39,909 (25.4)
Hispanic	9,844 (6.3)
Other	11,564 (7.4)
Smoking status
Never smoked	71,553 (45.5)
Current smoker	56,257 (35.8)
Past smoker	29,405 (18.7)
Diabetes	22,054 (14.0)
Obesity	17,116 (10.9)
Mental health diagnosis cluster count^a
0	107,638 (68.5)
1	27,949 (17.8)
≥2	21,608 (13.7)

Based on nine diagnostic clusters (depression, anxiety, psychotic, bipolar, somatization, eating, drug, alcohol disorders, and PTSD).

PTSD, posttraumatic stress disorder; SD, standard deviation.

The majority of women Veterans (68.5%) had no mental health diagnosis, 17.8% had a diagnosis in one cluster, and 13.7% carried a diagnosis in two or more clusters. LR analyses demonstrated that depressive disorders, anxiety disorders (other than PTSD), and psychotic disorders were the strongest mental health predictors of CAD after controlling for covariates (Table 2, Model 1). Depression had the strongest association with CAD, with a similar adjusted odds ratio as that of current smoking. Model 2 demonstrated that after controlling for age, race, smoking status, diabetes, and obesity, each mental health diagnosis cluster increased CAD odds by 44%.

Table 2.

Adjusted Stepwise Logistic Regression: Diagnoses Associated with Coronary Artery Disease

Predictor	AOR (95% CI)
Model 1
Age category
<50	1 (Reference)
50–59	1.90 (1.70–2.12)
60–69	4.41 (3.93–4.96)
70–79	8.28 (7.33–9.37)
≥80	11.50 (10.21–12.98)
Race
White	1 (Reference)
Black	0.96 (0.89–1.03)
Hispanic	0.49 (0.41–0.57)
Other	0.34 (0.28–0.40)
Smoking status
Never smoked	1 (Reference)
Current smoker	1.68 (1.58–1.78)
Past smoker	1.11 (1.04–1.19)
Diabetes	3.36 (3.18–3.55)
Obesity	1.55 (1.44–1.67)
Selected mental health diagnostic clusters
Depressive disorders (20.6% of sample)	1.60 (1.50–1.71)
Anxiety disorders^a (7.0% of sample)	1.32 (1.20–1.45)
Psychotic disorders (4.8% of sample)	1.23 (1.12–1.36)
Model 2^b
Present mental health diagnostic cluster count (0–3)	1.44 (1.38–.50)

Not including PTSD.

Includes age, race, smoking status, diabetes, and obesity, and substitutes count of present mental health conditions (0–3) of those clusters selected by Model 1.

AOR, adjusted odds ratio; CI, confidence interval.

Given the known association between depression and smoking,³¹ we conducted follow-up analyses to further establish depression as a predictor of CAD independent of smoking history. As expected, smoking and depression were significantly associated (χ²[2] = 1056.8, p < 0.001), with current smokers most likely to have a depression diagnosis (relative to nonsmokers and past smokers); however, separate LR models of CAD within each smoking category found comparable odds ratios for depression in each (Table 3), suggesting that depression has an independent main effect on the odds of CAD.

Table 3.

Logistic Regression: Odds of Coronary Artery Disease with Depression Stratified by Smoking Category

Within nonsmokers (odds ratios with 95% CIs)
	Odds ratio of coronary artery disease (CAD)	95% CI
Depression present	1.35	1.23–1.48

Within past smokers (odds ratios with 95% CIs)
	Odds ratio of CAD	95% CI
Depression present	1.31	1.16–1.48

Within current smokers (odds ratios with 95% CIs)
	Odds ratio of CAD	95% CI
Depression present	1.44	1.32–1.57

Discussion

We found that women Veterans over age 45 accessing VA care exhibited a high degree of mental health burden, which is associated with elevated odds of CAD. Furthermore, each additional mental health condition in the most predictive three clusters (depression, anxiety, and psychotic disorders) increased the odds of CAD by more than 40%, independent of other risk factors. The rate of CAD found in this study is comparable to previous reports among women Veterans.^1,32 Given that nearly a third (31%) of older women Veterans in this sample have at least one mental health diagnosis, these data identify a considerable population of women Veterans at risk for CAD for whom tailored prevention and intervention efforts may improve long-term prognosis.

Depressive disorders were the cluster of mental health diagnoses most strongly associated with CAD in this study population. Consistent with past Veteran^9,33 and non-Veteran studies^34,35 that show strong associations between depression and CAD, women Veterans with depression alone had 60% higher odds of having CAD than nondepressed women. Our study demonstrates that the magnitude of the relationship between depression and CAD is similar to that of current smoking, a well-established potent CAD risk factor. This connection has recently gained attention as a public health problem; in February 2014, the American Heart Association (AHA) issued a statement identifying depression as a risk factor for adverse medical outcomes in patients with acute coronary syndrome.¹⁶

Accumulating longitudinal data suggest that depression increases CVD risk.^16,36 Recently, prospective studies measuring coronary calcium levels among non-VA women documented an independent association between progression of CAD and depression,^37,38 suggesting that poor mental health accelerates subclinical or nascent CAD, similar to the effect of other well-known cardiac risk factors. In the general population, mental health conditions are associated with poor self-care and adherence to medical treatment.³⁹ In addition to the effect of unhealthy lifestyle in depressed persons,¹⁶ an emerging literature suggests that dysregulation of autonomic, immune-inflammatory, and hypothalamic–pituitary–adrenal axis functions is a potential physiologic mechanism that contributes to the association between depression and development of CAD.^16,17,40

Anxiety disorders (not including PTSD) were the second most predictive cluster of CAD, and demonstrated an additive effect on elevating CAD risk. This finding is concordant with a study which found that decreased heart rate variability, a measure of autonomic inflexibility and correlate of CVD, was more common among participants in whom both depression and anxiety were present.⁴¹

Although not demonstrated specifically in Veterans, the risk of mortality among individuals with CAD has been found to be greater among those with anxiety, especially when it is comorbid with depression.⁴² It remains unclear whether depression and anxiety are associated with CVD through different mechanisms. A recent study found that 24-hour urine epinephrine excretion was positively correlated with anxiety scores but not with measures of depression, potentially suggesting that anxiety and depression may increase CVD risk through different pathways.⁴³

Regardless of the exact mechanisms, symptoms of both depression and anxiety are amenable to treatment in a primary care and/or specialty mental healthcare context along with interventions aimed at improving management of CVD risk factors to maximize health benefits and improve quality of life. For example, in one VA study, Veterans of both genders with chronic mental health disorders and at least one CVD risk factor were randomized to a series of group self-management sessions and monthly care management calls versus usual care. Those in the intervention group had a statistically significant improvement in physical health-related quality of life. Although an improvement in CVD risk factors was not observed at 12-month follow-up, longer follow-up may be needed to achieve change in CVD risk.⁴⁴

Collaborative care models outside VA have demonstrated successful primary⁴⁵ and secondary⁴⁶ prevention of CVD among both older men and women, and suggest that the cardioprotective benefits of treatment are greater earlier in the course of CVD.⁴⁵ Additional promising work among patients with serious mental illness⁴⁷ suggests that targeted prevention in VA primary care or mental health settings may hold promise for decreasing CAD risk among vulnerable patients such as smokers with multimorbid mental health conditions. Future work should inform innovation for addressing such co-occurring physical and mental health conditions.

A second notable, and somewhat unexpected, finding in our study was the 36% prevalence of current smoking in women Veterans over age 45. Recently, VA researchers found a similar rate in a broader age range,⁶ and because prior work suggested higher smoking prevalence in younger women Veterans,³² we anticipated lower rates in this older cohort. Women Veterans have higher rates of smoking than do males¹³ and are more likely than men to smoke to regulate negative affect.⁴⁸ Women Veterans are also less likely than men to receive nicotine replacement therapy in VA.⁴⁹ Smoking, in turn, further magnifies CAD risk in this population.

Our findings suggest that smoking cessation treatments need to be tailored to meet the needs of women Veterans with mental health multimorbidity, however, little evidence currently exists to identify best practices. Prior VA work examining women Veterans' preferences for smoking cessation supported developing consumer-driven interventions, but did not examine the specific needs of women Veterans with mental health multimorbidity.^50,51 While women Veterans are known to use VA care differently than males, creating possible opportunities for intervention, no clear organizational factors associated with smoking cessation by gender have been identified to date.¹³

Other work examining smoking cessation treatment preferences of non-Veteran women demonstrated that individualizing treatment, offering single-gender groups, evening hours, and providing information on women-specific topics relevant to quitting (such as impact of pregnancy and menopause) were considered facilitative.⁵² Simply colocating standard smoking cessation programs in mental health settings has not been effective,⁵³ this may be due to evidence suggesting that individuals with mental health conditions often require higher doses and longer duration of pharmacotherapy and behavioral interventions.^31,54 Clearly, further work is needed to develop effective strategies to mitigate the high rates of smoking among women Veterans with comorbid mental health conditions and CVD risk.

The current study was a cross-sectional analysis of administrative data and did not allow us to infer causality or directionality. These data do suggest that patients with depression and anxiety may benefit from early intervention to reduce their modifiable CVD risk factors. As such, our study data provide important information to clinicians caring for women in both primary care and mental health settings.

Considering the mechanisms underlying the association in these data may prompt clinicians to expand their thinking about the standard approach to CVD—and CAD in particular—and may lend support to collaborative care and novel interdisciplinary program development. Although our findings do not support any one etiologic model, understanding and recognizing the mental health-CAD connection may inform clinical care and inspire future interventions. Since many women Veterans in this age group access care outside VA,¹ it is critical that non-VA clinicians be aware of these findings as well.

Limitations

In addition to the limitations inherent in cross-sectional research, there are several other issues in this study that are important to acknowledge. First, women Veterans using VA tend to have poorer health status than Veterans in the general population²; therefore, our findings may not generalize to women who do not access VA care. Second, women in this cohort may be accessing care outside VA and their diagnostic data may not be captured in these data. Furthermore, administrative code data may result in underestimates of the prevalence of true clinical conditions. This appears to be true for the diabetes and obesity estimates in these data. Finally, the study did not include information on psychiatric medication that could confound the relationship between mental health diagnosis and CAD (e.g., antipsychotic use may cause weight gain and metabolic syndrome).⁵⁵

Conclusions

For women Veterans, depression and anxiety could be viewed as CAD risk factors that are potentially modifiable. These associations should be examined prospectively in either administrative or cohort study data. Additional research to identify potential biologic intervention points between depression, mental health burden, and CAD is needed. In a 2009 study, women Veterans at increased risk for CAD demonstrated low knowledge and awareness of cardiovascular risk factors and treatment.⁷ VA has engaged in enhanced educational programming and outreach through a partnership with the AHA,⁴ and is developing novel gender-specific programs to address heart health. Additional research will be needed to determine older women Veterans' preferences for delivery of these interventions as well as best practices for those with comorbid mental health conditions.

Strengths of this study include the large national sample of women Veterans in midlife and beyond, as well as use of clinically derived smoking data. Combined with extant literature, our findings suggest that primary and secondary prevention interventions for heart disease tailored to the needs of women Veterans should be considered for both mental health and primary care settings, and will become increasingly critical as the expanding population of women Veterans ages. Finally, focused research directed toward determining best practices for smoking cessation among women Veterans with mental health conditions is also necessary to address CVD in this population.

Footnotes

Acknowledgments

Dr. Iverson's contribution was supported by the Department of Veterans Affairs Health Services Research & Development (HSR&D) Services Career Development Award (10-029). The contents of this article do not represent the views of the Department of Veterans Affairs or the United States Government. The authors thank Melissa Skanderson, MSW, for her work on data set acquisition.

Author Disclosure Statement

No competing financial interests exist.

References

Sourcebook: Women Veterans in the Veterans Health Administration. Volume 3. Sociodemographics, Utilization, Costs of Care, and Health Profile. Women's Health Services, Veterans Health Administration; 2014. Available at: www.womenshealth.va.gov/WOMENSHEALTH/docs/Sourcebook_Vol_3_FINAL.pdf Accessed August 28, 2016 .

Lehavot

, Hoerster

, Nelson

, Jakupcak

, Simpson

. Health indicators for military, veteran, and civilian women. Am J Prev Med, 2012; 42:473–480.

Weitlauf

, LaCroix

, Bird

, et al. Prospective analysis of health and mortality risk in veteran and non-veteran participants in the Women's Health Initiative. Womens Health Issues, 2015; 25:649–657.

Petersen

. Heart Disease is the Number One Killer of Women. Inside Veterans Health,. 2014. Available at: www.va.gov/health/NewsFeatures/2014/February/Heart-Disease-is-the-Number-One-Killer-of-Women.asp Accessed November 15, 2015 .

Vimalananda

, Miller

, Christiansen

, Wang

, Tremblay

, Fincke

. Cardiovascular disease risk factors among women veterans at VA medical facilities. J Gen Intern Med, 2013; 28 Suppl 2:S517–S523.

Haskell

, Bathulapalli

, Pham

, et al. Sex differences in patient and provider response to elevated low-density lipoprotein cholesterol. Womens Health Issues, 2014; 24:575–580.

Canter

, Atkins

, McNeal

, Bush

. Risk factor treatment in veteran women at risk for cardiovascular disease. J Surg Res, 2009; 157:175–180.

Goldstein

, Melnyk

, Zullig

, et al. Heart matters: Gender and racial differences cardiovascular disease risk factor control among veterans. Women's Health Issues, 2014; 24:477–483.

Sambamoorthi

, Mitra

, Findley

, Pogach

. Decomposing gender differences in low-density lipoprotein cholesterol among veterans with or at risk for cardiovascular illness. Womens Health Issues, 2012; 22:e201–e208.

10.

Runnals

, Garovoy

, McCutcheon

, et al. Systematic review of women veterans' mental health. Womens Health Issues, 2014; 24:485–502.

11.

Cohen

, Marmar

, Ren

, Bertenthal

, Seal

. Association of cardiovascular risk factors with mental health diagnoses in Iraq and Afghanistan war veterans using VA healthcare. JAMA, 2009; 302:489–492.

12.

Rose

, Farmer

, Yano

, Washington

. Racial/ethnic differences in cardiovascular risk factors among women veterans. J Gen Inter Med, 2013; 28 Suppl 2:S524–S528.

13.

Farmer

, Rose

, Riopelle

, Lanto

, Yano

. Gender differences in smoking and smoking cessation treatment: An examination of the organizational features related to care. Womens Health Issues, 2011; 21(4 Suppl):S182–S189.

14.

McCauley

, Blosnich

, Dichter

. Adverse childhood experiences and adult health outcomes among veteran and nonveteran women. J Womens Health (Larchmt), 2015; 24:723–729.

15.

Cohen

, Edmondson

, Kronish

. State of the art review: Depression, stress, anxiety, and cardiovascular disease. Am J Hypertens, 2015; 28:1295–1302.

16.

Lichtman

, Froelicher

, Blumenthal

, et al. Depression as a risk factor for poor prognosis among patients with acute coronary syndrome: Systematic review and recommendations: A scientific statement from the American Heart Association. Circulation, 2014; 129:1350–1369.

17.

Penninx

, Milaneschi

, Lamers

, Vogelzangs

. Understanding the somatic consequences of depression: Biological mechanisms and the role of depression symptom profile. BMC Med, 2013; 11:129.

18.

Nemeroff

, Goldschmidt-Clermont

. Heartache and heartbreak—The link between depression and cardiovascular disease. Nat Rev Cardiol, 2012; 9:526–539.

19.

Scherrer

, Chrusciel

, Zeringue

, et al. Anxiety disorders increase risk for incident myocardial infarction in depressed and nondepressed Veterans Administration patients. Am Heart J, 2010; 159:772–779.

20.

Kilbourne

, Morden

, Austin

, et al. Excess heart-disease-related mortality in a national study of patients with mental disorders: Identifying modifiable risk factors. Gen Hosp Psychiatry, 2009; 31:555–563.

21.

Clouse

, Lustman

, Freedland

, Griffith

, McGill

, Carney

. Depression and coronary heart disease in women with diabetes. Psychosom Med, 2003; 65:376–383.

22.

Heart Disease Facts.. 2015. Available at: www.cdc.gov/heartdisease/facts.htm Accessed April 5, 2016 .

23.

Gerber

, King

, Pineles

, et al. Hormone therapy use in women veterans accessing veterans health administration care: A National Cross-Sectional Study. J Gen Inter Med, 2015; 30:169–175.

24.

Lurie

, Popkin

, Dysken

, Moscovice

, Finch

. Accuracy of diagnoses of schizophrenia in Medicaid claims. Hosp Community Psychiatry, 1992; 43:69–71.

25.

Walkup

, Wei

, Sambamoorthi

, Crystal

. Sensitivity of an AIDS case-finding algorithm: Who are we missing?. Med Care, 2004; 42:756–763.

26.

Roy

, Lloyd

. Epidemiology of depression and diabetes: A systematic review. J Affect Disord, 2012; 142 Suppl:S8–S21.

27.

van Dieren

, Beulens

, Kengne

, et al. Prediction models for the risk of cardiovascular disease in patients with type 2 diabetes: A systematic review. Heart, 2012; 98:360–369.

28.

Kulie

, Slattengren

, Redmer

, Counts

, Eglash

, Schrager

. Obesity and women's health: An evidence-based review. J Am Board Fam Med, 2011; 24:75–85.

29.

McGinnis

, Brandt

, Skanderson

, et al. Validating smoking data from the Veteran's Affairs Health Factors dataset, an electronic data source. Nicotine Tob Res, 2011; 13:1233–1239.

30.

Stekhoven

, Buhlmann

. MissForest—Non-parametric missing value imputation for mixed-type data. Bioinformatics, 2012; 28:112–118.

31.

Schroeder

, Morris

. Confronting a neglected epidemic: Tobacco cessation for persons with mental illnesses and substance abuse problems. Annu Rev Public Health, 2010; 31:297–314 291p following 314.

32.

Brown

. Smoking prevalence among US veterans. J Gen Intern Med, 2010; 25:147–149.

33.

Shen

, Findley

, Banerjea

, Sambamoorthi

. Depressive disorders among cohorts of women veterans with diabetes, heart disease, and hypertension. J Womens Health (Larchmt), 2010; 19:1475–1486.

34.

Hawkins

, Callahan

, Stump

, Stewart

. Depressive symptom clusters as predictors of incident coronary artery disease: A 15-year prospective study. Psychosom Med, 2014; 76:38–43.

35.

Fiedorowicz

. Depression and cardiovascular disease: An update on how course of illness may influence risk. Curr Psychiatry Rep, 2014; 16:492.

36.

Penninx

. Depression and cardiovascular disease: Epidemiological evidence on their linking mechanisms. Neurosci Biobehav Rev, 2017; 74(Pt B):277–286.

37.

Janssen

, Powell

, Matthews

, et al. Depressive symptoms are related to progression of coronary calcium in midlife women: The Study of Women's Health Across the Nation (SWAN) Heart Study. Am Heart J, 2011; 161:1186–1191 e1181.

38.

Janssen

, Powell

, Matthews

, et al. Relation of persistent depressive symptoms to coronary artery calcification in women aged 46 to 59 years. Am J Cardiol, 2016; 117:1884–1889.

39.

DiMatteo

, Lepper

, Croghan

. Depression is a risk factor for noncompliance with medical treatment: Meta-analysis of the effects of anxiety and depression on patient adherence. Arch Inter Med, 2000; 160:2101–2107.

40.

Stewart

, Rand

, Muldoon

, Kamarck

. A prospective evaluation of the directionality of the depression-inflammation relationship. Brain Behav Immun, 2009; 23:936–944.

41.

Kemp

, Quintana

, Felmingham

, Matthews

, Jelinek

. Depression, comorbid anxiety disorders, and heart rate variability in physically healthy, unmedicated patients: Implications for cardiovascular risk. PLoS One, 2012; 7:e30777.

42.

Watkins

, Koch

, Sherwood

, et al. Association of anxiety and depression with all-cause mortality in individuals with coronary heart disease. J Am Heart Assoc, 2013; 2:e000068.

43.

Paine

, Watkins

, Blumenthal

, Kuhn

, Sherwood

. Association of depressive and anxiety symptoms with 24-hour urinary catecholamines in individuals with untreated high blood pressure. Psychosom Med, 2015; 77:136–144.

44.

Kilbourne

, Barbaresso

, Lai

, et al. Improving physical health in patients with chronic mental disorders: Twelve-month results from a randomized controlled collaborative care trial. J Clin Psychiatry, 2017; 78:129–137.

45.

Stewart

, Perkins

, Callahan

. Effect of collaborative care for depression on risk of cardiovascular events: Data from the IMPACT randomized controlled trial. Psychosom Med, 2014; 76:29–37.

46.

Davidson

, Rieckmann

, Clemow

, et al. Enhanced depression care for patients with acute coronary syndrome and persistent depressive symptoms: Coronary psychosocial evaluation studies randomized controlled trial. Arch Intern Med, 2010; 170:600–608.

47.

Baxter

, Harris

, Khatib

, Brugha

, Bien

, Bhui

. Reducing excess mortality due to chronic disease in people with severe mental illness: Meta-review of health interventions. Br J Psychiatry, 2016; 208:322–329.

48.

Doran

. Sex differences in smoking cue reactivity: Craving, negative affect, and preference for immediate smoking. Am J Addict, 2014; 23:211–217.

49.

Sherman

, Fu

, Joseph

, Lanto

, Yano

. Gender differences in smoking cessation services received among veterans. Womens Health Issues, 2005; 15:126–133.

50.

Katzburg

, Farmer

, Poza

, Sherman

. Listen to the consumer: Designing a tailored smoking-cessation program for women. Subst Use Misuse, 2008; 43:1240–1259.

51.

Katzburg

, Yano

, Washington

, et al. Combining women's preferences and expert advice to design a tailored smoking cessation program. Subst Use Misuse, 2009; 44:2114–2137.

52.

Minian

, Penner

, Voci

, Selby

. Woman focused smoking cessation programming: A qualitative study. BMC Womens Health, 2016; 16:17.

53.

Okoli

, Mason

, Brumley-Shelton

, Robertson

. Providing tobacco treatment in a community mental health setting: A Pilot Study. J Addict Nurs, 2017; 28:34–41.

54.

Ziedonis

, Hitsman

, Beckham

, et al. Tobacco use and cessation in psychiatric disorders: National Institute of Mental Health report. Nicotine Tob Res, 2008; 10:1691–1715.

55.

Haddad

, Sharma

. Adverse effects of atypical antipsychotics: Differential risk and clinical implications. CNS Drugs, 2007; 21:911–936.