SEARCH

SEARCH BY CITATION

Keywords:

  • Antidepressants;
  • BMI;
  • cardiovascular disease;
  • health risks

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

To determine if selective-serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants (TCAs) influence the association between obesity and cardiovascular disease risk, participants from the Third National Health and Nutrition Examination Survey (NHANES III; 1988–1992) and continuous NHANES (1999–2009, n = 18 274) were used. For a given body mass index (BMI), individuals taking SSRIs (n = 219) tended to have significantly better health risk profiles with lower systolic blood pressure (P = 0.002) and higher high-density lipoprotein (P = 0.003) compared with non-users. Conversely, those who used TCAs (n = 116) had significantly worse health risk profiles with higher diastolic blood pressure (P ≤ 0.0001) and triglycerides (P = 0.023) as compared with non-users for a given BMI. Insulin resistance (HOMA-IR) was higher in TCA users and those with larger BMIs, whereby the differences in insulin resistance between TCA users and non-users was greater with higher BMIs (interaction effect: P = 0.013). Furthermore, individuals taking SSRIs were less likely to have cardiovascular disease than non-users (odds ratio, 95% confidence interval = 0.50, 0.33–0.75) for a given BMI, with no differences by TCA use (odds ratio = 0.74, 0.44–1.24). SSRI and TCA use may alter how body weight relates with cardiovascular risk. When prescribing antidepressant medications, it may be necessary to monitor and consider body weight and cardiovascular risk profile of individual patients.

What is already known about this subject
  • Psychopathologies have been linked with increased risk of morbidity and mortality, as well as a poorer quality of life.
  • Various classes of antidepressants are reported to influence body weight differently, including both weight loss and weight gain.
  • Obesity and weight gain are associated with increased cardiovascular risk.
What this study adds
  • Shows that drugs commonly prescribed for depression may influence health risk factors independent of body mass index.
  • The use of selective-serotonin reuptake inhibitors and tricyclic antidepressants may be associated with differential and clinically relevant changes in cardiovascular disease risk independent of body mass index.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

The current prevalence of depression is between 8% and 12% worldwide [1]. Beyond the increased risk of coronary artery disease (CAD), cancer, stroke, epilepsy, Alzheimer's disease, diabetes [2] and all-cause mortality, mood disorders have also been linked to poorer quality of life [2, 3] and increased likelihood of obesity [4]. Since the mid-1990s, the use of antidepressants in the US has increased substantially [5]. The prevalence of selective-serotonin reuptake inhibitors (SSRIs) has risen from 55% to 67% of all antidepressant prescriptions, while the use of tricyclic antidepressants (TCAs) has decreased from 35% to 11% [5]. Even though the use of TCAs has decreased over recent years, they are reported to have a greater efficacy at treating certain types of depression such as melancholic and psychotic [6]. Given that psychopathological patient care is becoming increasingly individualized, it is still of clinical importance to research TCAs as it is likely that they will continue to be prescribed both now and in the future [6].

SSRIs and tricyclics are also commonly used for a wide range of syndromes including anxiety, attention deficit, nervous system disorders and other non-psychiatric disorders [7]. Antidepressant use has been associated with a 1 kg to 3 kg weight gain in 10–40% of patients receiving pharmaceutical treatment [8]. However, the different classes of antidepressants are reported to influence body weight differently [4, 8, 9]. Short-term use of SSRIs is associated with weight loss [10-13] while weight restoration or gain occurs with long-term use [4, 14, 15]. On the other hand, long-term treatment with TCAs has been more frequently linked with weight gain [4, 8, 9, 15, 16]. Furthermore, TCAs have been linked with hypertriglyceridaemia [17, 18] and both TCAs and SSRIs have been associated with blood pressure dysregulation [19-22]. However, because antidepressant use influences both body weight and cardiovascular disease (CVD) risk, it is currently unclear if antidepressant use alters CVD risk independent of obesity. Thus, the objective of this study is to determine if the type of antidepressant used (SSRIs/TCAs) influences the association between obesity and CVD risk.

Methods and procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

Participants

The sample included 96 154 participants from the National Health and Nutrition Examination Survey (NHANES). NHANES is a cross-sectional study that is a US nationally representative survey which utilizes a multistage probability cluster design. The National Center for Health Statistics of the Centers for Disease Control and Prevention collects the data through interviews, physical examinations and laboratory tests. Prior to examination, all participants gave their informed written consent and study protocol was approved by National Center for Health Statistics. Further details pertaining to study design and methods have been previously reported [23, 24].

NHANES III (1988–1994) and NHANES continuous data (1999–2010) were combined to provide sufficient sample size for the analysis of different subclasses of antidepressant use. Participants were excluded from the data set if they were under 18 years of age (n = 41 198), if they were pregnant (n = 1 196) or they had a body mass index (BMI) < 15 kg m−2 or >70 kg m−2 (n = 768). Additionally, participants with missing data for BMI (n = 18 133), antidepressant use (n = 14), systolic blood pressure (SBP: n = 38 512), diastolic blood pressure (DBP: n = 39 576), high-density lipoprotein (HDL: n = 30 809), triglycerides (TGs: n = 56 127), fasting plasma glucose (n = 63 849), glycohaemoglobin (n = 36 994), insulin (n = 64 244), homeostatic model assessment of insulin resistance (HOMA-IR: n = 64 336), chronic disease (cancer: n = 43 723, diabetes: n = 16 942, CVD: n = 43 866, obstructive respiratory disease: n = 43 798), smoking status (n = 43 704) and dietary information were not included (carbohydrates [CHO]: n = 12 724, fat: n = 12 478). This left 18 274 eligible participants.

Survey methods

Age, gender, ethnicity (white, black, other), poverty-income ratio (income, scored 0–5), medication use (user/non-user), smoking status (current smoker/non-smoker) and physical activity (PA) status were assessed by individual questionnaire. Income score is calculated as the family earnings divided by the poverty guidelines provided by the Department of Health and Human Services. If income was missing, mean substitution was performed. PA was coded as: moderate intensity ≥1 time of moderate intensity PA per week; or vigorous intensity ≥1 time of vigorous intensity PA per week; or inactive (0 min of moderate or vigorous PA per week). Moderate PA was defined as a metabolic equivalent score between 3 and 6 while vigorous PA was a metabolic equivalent score >6. When individuals performed both moderate and vigorous PA, they were classified as vigorous PA. If the participant reported doing no PA, they were recorded as inactive. In NHANES III, blood lipids and glycaemic indicators were measured using standard procedures after a 10–16-h fast prior to morning testing or after a 6-h fast for afternoon/evening tests. In NHANES 1999–2004, participants were asked to fast for 8–24 h and ≥9 h for NHANES 2005–2010. Any additional details on blood lipid or glycaemic measurement techniques can be found elsewhere [23, 24]. Chronic disease history was obtained through self-reported doctor diagnoses of cancer, diabetes and CVD (yes/no). CVD was a history of heart failure, angina or stroke in NHANES III and a history of heart failure, angina, stroke, myocardial infarction or coronary heart disease in NHANES 1999–2010. A 24-h dietary recall interview was also performed to obtain dietary information on CHOs and fat (g).

A minimum of three blood pressure (mmHg) readings were taken by trained blood pressure technicians. Anthropometric measures such as weight, height and BMI were also assessed by trained health technicians at the mobile examination centre. Self-reported height and weight were used to substitute missing BMI data in NHANES III.

Prescription medication categorization

Data on prescription medication use were obtained through household interviews. Type of prescription medication usage within the last 30 d was recorded. Whenever possible, the prescription container was examined by the interviewer. In NHANES III, the medications were assigned a 4-digit code based on the National Drug Directory of the Food and Drug Administration. In NHANES continuous, the Lexicon Plus® (Cerner Multum Inc., Denver, CO., USA) system was used to sort medications by therapeutic classification and drug ingredients. The two most common antidepressant drug categories were analysed in this study: (i) SSRIs (n = 219) and (ii) TCAs (n = 116) .

Statistical analysis

Demographic characteristics were stratified by user and non-user for each antidepressant type. Continuous variables are shown as a mean ± standard deviation. Categorical variables are presented as sample size (n) and prevalence (%). Group differences in continuous variables were analysed using t-tests and chi-square tests (χ2) were used for categorical variables. Linear regression analyses were used to examine the influence of antidepressant usage on the association between BMI and health risks using first order drug and BMI interaction and main effect terms. The odds ratio (OR, 95% confidence interval) of prevalent chronic disease was assessed using logistic regression. All regression analyses were adjusted for age, sex, ethnicity, income, smoking status, PA, as well as dietary CHOs and fat intake. Further adjustment for prevalent CVD was performed when assessing blood pressure and blood lipids while diabetes status was adjusted for when analysing glycaemic markers. Due to the amalgamation of continuous and NHANES III data sets, weighted adjustments could not be used. All analyses were completed with SAS (ver. 9.2) (SAS Institute Inc., Cary, NC, USA). Statistical significance was defined as P < 0.05.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

Participant characteristics

Participant characteristics are shown in Table 1. Age did not differ by SSRI use (P > 0.05), while individuals using TCAs were older than their non-user counterparts (P < 0.05). Individuals taking SSRIs and TCAs were more likely to be female than non-users (P < 0.05). Those taking SSRIs had a 1.0 ± 6.3 kg m−2 higher BMI than non-users (P < 0.05) with no difference by TCA use (P > 0.05). SSRI users had a 0.4 ± 1.6 higher income while TCA users had a 0.4 ± 1.4 lower income than their respective non-users (P < 0.05). PA levels did not differ by SSRI use (P > 0.05), while TCA users had significantly lower PA levels than those not using TCAs (P < 0.05).

Table 1. Characteristics of participants, stratified by drug and use category
Drug categorySSRIsTCAs
Use categoryNon-userUserP-valueNon-userUserP-value
  1. Continuous data as mean ± SD and categorical data presented as n (%) within ‘Use category’.

  2. P < 0.05 significantly different from non-users within that drug category.

  3. BMI, body mass index; CHO, carbohydrates; CVD, cardiovascular disease; DBP, diastolic blood pressure; FPG, fasting plasma glucose; HbA1c, glycohaemoglobin; HDL, high-density lipoprotein cholesterol; HOMA-IR, homeostatic model assessment of insulin resistance; Income, poverty-income ratio; n, sample size; SBP, systolic blood pressure; SSRIs, selective-serotonin reuptake inhibitors; TCAs, tricyclic antidepressants; TGs, fasting serum triglycerides.

n (%)18 055 (98.8)219 (1.2) 18 158 (99.4)116 (0.6) 
Age (years)49.3 ± 18.551.2 ± 15.40.0649.2 + 18.558.5 + 17.6<0.0001
Sex (Female)9 083 (50.3)157 (71.7)<0.00019 160 (50.5)80 (69.0)<0.0001
BMI (kg m−2)28.1 ± 6.129.1 ± 6.30.0128.1 ± 6.128.4 ± 7.00.62
Ethnicity      
White8 474 (46.9)159 (72.6)<0.00018 569 (47.2)64 (55.2)0.17
Black3 898 (21.6)15 (6.9) 3 889 (21.4)24 (20.7) 
Other5 683 (31.5)45 (20.6) 5 700 (31.4)28 (24.1) 
Income (0–5)2.5 ± 1.52.9 ± 1.60.0012.5 ± 1.52.1 ± 1.40.003
Smoking status      
Smoker4 222 (23.4)55 (25.1)0.554 246 (23.4)31 (26.7)0.40
Dietary factors      
CHO intake (g)250 ± 117236 ± 1070.07250 ± 117229 ± 1120.05
Fat intake (g)77 ± 4374 ± 360.2077 ± 4372 ± 490.33
Physical activity (≥1 per week)      
Inactive6 619 (36.7)85 (38.8)0.736 642 (36.6)62 (53.5)<0.0001
Moderately active9 126 (50.6)109 (49.8)  184 (50.6)51 (44.0) 
Vigorously active2 310 (12.8)25 (11.4) 2 332 (12.8)3 (2.6) 
Diabetes (%)1 652 (9.2)23 (10.5)0.491 651 (9.1)24 (20.7)<0.0001
CVD1 730 (9.6)34 (15.5)0.0031 743 (9.6)21 (18.1)0.002
Health risk factors      
SBP (mmHg)125 ± 20121 ± 200.004125 ± 20132 ± 210.0002
DBP (mmHg)72 ± 1170 ± 120.00272 ± 1176 ± 100.0006
HDL (mM)1.35 ± 0.401.44 ± 0.440.0021.35 ± 0.401.34 ± 0.370.66
TG (mM)1.57 ± 1.051.56 ± 0.880.821.57 ± 1.051.85 ± 1.070.004
FPG (mM)5.8 ± 1.85.7 ± 1.50.525.8 ± 1.86.0 ± 2.60.30
HOMA-IR3.4 ± 3.93.9 ± 5.20.163.4 ± 3.95.0 ± 8.00.04
HbA1c (%)5.6 ± 1.05.6 ± 0.70.325.6 ± 1.05.8 ± 1.40.08

Individuals taking SSRIs had a higher prevalence of CVD, 4 ± 20 mmHg lower SBP, 2 ± 12 mmHg lower DBP and 0.09 ± 0.44 mM higher HDL than those not using SSRIs (P < 0.05). Those using TCAs had a higher prevalence of diabetes and CVD, 7 ± 21 mmHg higher SBP, 4 ± 10 mmHg higher DBP, 0.28 ± 1.07 higher TG and 1.6 ± 8.0 higher HOMA-IR than those not using TCAs (P < 0.05). There were no significant differences (P ≥ 0.05) in smoking status, CHO or fat intake, fasting plasma glucose and glycohaemoglobin by SSRI or TCA use.

Cardiovascular disease risk factors

For a given BMI, individuals taking SSRIs tended to have significantly better health risk profiles with lower SBP (P = 0.002), marginally lower DBP (P = 0.056) and higher HDL (P = 0.003) compared with non-users (Table 2) with adjustment for age, sex, ethnicity, income, smoking status, PA, total CHOs and fat intake and CVD. Conversely, those who used TCAs had significantly (P < 0.05) worse health risk profiles with higher DBP and TGs as compared with non-users for a given BMI (Table 2). Insulin resistance (HOMA-IR) was higher in TCA users and those with larger BMIs, whereby the differences in insulin resistance between TCA users and non-users was greater with higher BMIs (interaction effect: P = 0.013; Fig. 1).

Table 2. The association of antidepressant usage on health risk factors by drug category
Drug categorySSRIsP-valueTCAsP-value
Predicted LSM valuePredicted LSM value
Non-userUserNon-userUser
  1. Predicted least square adjusted means (LSM ± SEE) values are adjusted for: body mass index, age, sex, ethnicity, income, smoking status, physical activity, total carbohydrates and fat intake. Further adjustment for cardiovascular disease was performed when assessing blood pressure and blood lipids while diabetes status was adjusted for when analysing glycaemic markers.

  2. *Significant interaction effect (see Fig. 1).

  3. DBP, diastolic blood pressure; FPG, fasting plasma glucose; HbA1c, glycohaemoglobin; HDL, high-density lipoprotein; HOMA-IR, homeostatic model assessment of insulin resistance; Income, poverty-income ratio; SBP, systolic blood pressure; SSRIs, selective-serotonin reuptake inhibitors; TCAs, tricyclic antidepressants; TGs, fasting serum triglycerides.

SBP (mmHg)125 ± 1121 ± 10.002125 ± 1127 ± 20.19
DBP (mmHg)71 ± 169 ± 10.05671 ± 175 ± 1<0.0001
HDL cholesterol (mM)1.34 ± 0.0051.41 ± 0.020.0031.34 ± 0.0051.29 ± 0.030.13
TG (mM)1.57 ± 0.011.50 ± 0.070.321.57 ± 0.011.78 ± 0.090.023
FPG (mM)7.1 ± 0.026.9 ± 0.10.457.1 ± 0.026.9 ± 0.10.21
HOMA-IR5.0 ± 0.055.3 ± 0.20.14***
HbA1c (%)6.4 ± 0.016.4 ± 0.060.496.4 ± 0.016.3 ± 0.080.35
figure

Figure 1. The relationship between body mass index (BMI) and homeostatic model assessment of insulin resistance (HOMA-IR) in selective-serotonin reuptake inhibitors (SSRI) (a) and tricyclic antidepressants (TCA) (b) users. Solid line, SSRI or TCA user; dashed line, SSRI or TCA non-user. INT, interaction; ME, main effect.

Download figure to PowerPoint

Likelihood of chronic disease

Neither SSRIs nor TCAs significantly altered the relationship between BMI and cancer (P = 0.30 and P = 0.49 respectively, results not shown). However, individuals taking SSRIs were less likely to have prevalent CVD than non-users (OR, 95% confidence interval = 0.50, 0.33–0.75) for a given BMI, with no differences by TCA use (OR = 0.74, 0.44–1.24) (Fig. 2).

figure

Figure 2. The relationship between body mass index (BMI) and cardiovascular disease (CVD) in selective-serotonin reuptake inhibitors (SSRI) (a) and tricyclic antidepressants (TCA) (b) users. Solid line, SSRI or TCA user; dashed line, SSRI or TCA non-user. INT, interaction; ME, main effect; OR, odds ratio.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

The findings from this study demonstrate that for a given BMI, antidepressant use is associated with differences in health risk factors and CVD risk. Individuals who used SSRIs tended to have a better cardiometabolic profile, while TCA users generally had a worse cardiometabolic profile than was expected given their BMI. Clinically, this may suggest that the differences in body weight observed with antidepressant use may not be associated with the normally expected differences in cardiovascular risk .

Antidepressant use has been linked to both hypotension and hypertension, depending on the type of antidepressant, through alterations to sympathetic activation [19-22]. SSRIs are reported to have inhibitory effects on the sympathetic nervous system, consequently lowering mean arterial pressure [21]. Contrastingly, TCA use was associated with significantly higher resting DBP for a given BMI. Increases in blood pressure with TCA use could be attributed to the blocking of noradrenaline reuptake which is indicative of increases in sympathetic activation [19, 21, 22], or increases in BMI typically observed with their use [4, 8, 9, 15, 16]. However, the hypertensive effects of TCAs were observed independent of BMI, suggesting these differences may be due to alterations in autonomic function. Therefore, both SSRIs and TCAs differentially influence blood pressure independent of body weight.

Past studies have linked depression with obesity [4] and inflammation [22, 25-27], which are both factors in development of dyslipidaemia and decreased HDL [17, 18, 28-30], as well as the pathogenesis of heart failure and CAD [31, 32]. However, SSRI use has been linked with decreased production of inflammatory markers [25, 33], and may potentially explain why SSRI use was associated with higher HDL levels, with and without adjustment for BMI. In contrast, TCA use is often connected with hypertriglyceridaemia that is commonly attributed to weight gain [17, 18]. We extend previous research [17, 18] by demonstrating the association between TCA use and hypertriglyceridaemia is true independent of BMI. The association between TCAs and hypertriglyceridaemia may be due to the higher insulin resistance observed in TCA users. Insulin resistance has been hypothesized to increase TG levels independent of obesity [34, 35]. Thus, our findings suggest that antidepressant use influences blood lipids and insulin sensitivity independent of body weight.

Although depression and CVD are frequently linked [19, 32, 36, 37], research on antidepressant usage in relation to prevalence of CVD is divided [21, 37]. This is likely due to the wide range of cardiac factors that antidepressants can influence, including autonomic regulation, blood pressure and cholesterol levels. We observed that SSRI use was associated with lower SBP for a given BMI. Although 4 mmHg lower SBP may appear to be a small magnitude, this magnitude of decrease has been related to an 19% lower incidence of stroke and 14% lower incidence of ischaemic heart disease [38]. Conversely, a 4 mmHg higher DBP observed with TCA use has been linked with a 16–21% increase in the incidence of CAD [39]. Therefore, SSRI and TCA use may be associated with differential and clinically relevant changes in CVD risk independent of BMI.

This study had limitations and strengths. The data are from cross-sectional surveys and these results do not allow one to infer causation. Moreover, because of the amalgamation of NHANES III and continuous data, the results are not representative of the current US population. It is also worth noting that we investigated two agents that are commonly prescribed as antidepressants but did not strictly examine depressive disorders. SSRIs and TCAs are used for an array of both psychiatric and non-psychiatric disorders and we were unable to ascertain what condition(s) the medications were specifically prescribed for. A strength was the use of chronic disease history data collected through personal interviews which has been shown to generate a low cognitive burden on survey respondents and produce a low recall bias when compared with self-administered questionnaires [40]. Another key strength was the rigorous adjustment for numerous confounding variables that have been shown to be associated with CVD risk in addition to adjustments for prevalent chronic conditions in order to reduce potential bias from previously established indications and contraindications for each antidepressant.

In summary, for a given BMI when compared with non-users, SSRI users tended to have a better cardiometabolic profile, while TCA users generally had a worse cardiometabolic profile. Future longitudinal studies are needed to explore and test for causal relations. When health professionals are prescribing antidepressant medications, further consideration should be given to the cardiovascular risk profile of individual patients.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References

JLK, CIA and MAR conceived the project. KJS carried out the data analysis. Both, JLK and KJS completed data interpretations and writing of the manuscript. KJS performed the search of literature and generation of tables and figures. JLK is the guarantor of this manuscript and takes full responsibility for the integrity and accuracy of data analysis. All the authors were involved in the editing of the manuscript and had final approval of the submitted and published versions.

Disclosure: KJS received an OGS scholarship from the Ontario Ministry of training, colleges and universities as well as a funding agreement provided by York University. JLK, CIA and MAR received a research grant (#131594) from the Canadian Institute of Health Research (CIHR).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods and procedures
  5. Results
  6. Discussion
  7. Conflict of Interest Statement
  8. Acknowledgements
  9. References
  • 1
    Kemp AH, Quintana DS, Gray MA et al. Impact of depression and antidepressant treatment on heart rate variability: a review and meta-analysis. Biol Psychiatry 2010; 67: 10671074.
  • 2
    Ramasubbu R, Beaulieu S, Taylor VH, Schaffer A, McIntyre RS. The CANMAT task force recommendations for the management of patients with mood disorders and comorbid medical conditions: diagnostic, assessment, and treatment principles. Ann Clin Psychiatry 2012; 24: 8290.
  • 3
    Yohannes AM, Willgoss TG, Baldwin RC, Connolly MJ. Depression and anxiety in chronic heart failure and chronic obstructive pulmonary disease: prevalence, relevance, clinical implications and management principles. Int J Geriatr Psychiatry 2010; 25: 12091221.
  • 4
    Serretti A, Mandelli L. Antidepressants and body weight: a comprehensive review and meta-analysis. J Clin Psychiatry 2010; 71: 12591272.
  • 5
    Olfson M, Marcus SC. National patterns in antidepressant medication treatment. Arch Gen Psychiatry 2009; 66: 848856.
  • 6
    Wrobel S. Science, serotonin, and sadness: the biology of antidepressants: a series for the public. FASEB J 2007; 21: 34043417.
  • 7
    Mark TL. For what diagnoses are psychotropic medications being prescribed?: a nationally representative survey of physicians. CNS Drugs 2010; 24: 319326.
  • 8
    Schwartz TL, Nihalani N, Jindal S, Virk S, Jones N. Psychiatric medication-induced obesity: a review. Obes Rev 2004; 5: 115121.
  • 9
    Vanina Y, Podolskaya A, Sedky K et al. Body weight changes associated with psychopharmacology. Psychiatr Serv 2002; 53: 842847.
  • 10
    Croft H, Settle E, Houser T et al. A placebo-controlled comparison of the antidepressant efficacy and effects on sexual functioning of sustained-release bupropion and sertraline. Clin Ther 1999; 21: 643648.
  • 11
    Goldstein DJ, Rampey AH, Roback PJ et al. Efficacy and safety of long-term fluoxetine treatment of obesity – maximizing success. Obes Res 1995; 3(Suppl. 4): 481S490S.
  • 12
    Lawton CL, Wales JK, Hill AJ, Blundell JE. Serotoninergic manipulation, meal-induced satiety and eating pattern: effect of fluoxetine in obese female subjects. Obes Res 1995; 3: 345356.
  • 13
    Michelson D, Amsterdam JD, Quitkin FM et al. Changes in weight during a 1-year trial of fluoxetine. Am J Psychiatry 1999; 156: 11701176.
  • 14
    Fava M, Judge R, Hoog SL, Nilsson ME, Koke SC. Fluoxetine versus sertraline and paroxetine in major depressive disorder: changes in weight with long-term treatment. J Clin Psychiatry 2000; 61: 863867.
  • 15
    Kivimaki M, Hamer M, Batty GD et al. Antidepressant medication use, weight gain, and risk of type 2 diabetes: a population-based study. 2010; 33: 26112616.
  • 16
    Berken H, Stern C. Weight gain of tricyclic antidepressants. J Affect Disord 1984; 7: 133138.
  • 17
    van Reedt Dortland AKB, Giltay EJ, van Veen T, Zitman FG, Penninx BWJH. Metabolic syndrome abnormalities are associated with severity of anxiety and depression and with tricyclic antidepressant use. Acta Psychiatr Scand 2010; 122: 3039.
  • 18
    Kopf D, Westphal S, Luley CW et al. Lipid metabolism and insulin resistance in depressed patients. J Clin Psychopharmacol 2004; 24: 527531.
  • 19
    Coupland N, Wilson S, Nutt D. Antidepressant drugs and the cardiovascular system: a comparison of tricyclics and selective serotonin reuptake inhibitors and their relevance for the treatment of psychiatric patients with cardiovascular problems. J Psychopharmacol 1997; 11: 8392.
  • 20
    Licht CMM, de Geus EJC, Seldenrijk A et al. Depression is associated with decreased blood pressure, but antidepressant use increases the risk for hypertension. Hypertension 2009; 53: 631638.
  • 21
    Licht CMM, Penninx BWJH, de Geus EJC. Effects of antidepressants, but not psychopathology, on cardiac sympathetic control: a longitudinal study. Neuropsychopharmacology 2012; 37: 24872495.
  • 22
    Penninx BWJH, Milaneschi Y, Lamers F, Vogelzangs N. Understanding the somatic consequences of depression: biological mechanisms and the role of depression symptom profile. BMC Med 2013; 11: 129.
  • 23
    Centers for Disease Control and Prevention. Analytic and Reporting Guidelines for the Third National Health and Nutrition Examination Survey. National Center for Health Statistics, Hyattsville, MD, USA. 1996.
  • 24
    Centers for Disease Control and Prevention. Analytic and Reporting Guidelines for the National Health and Nutrition Examination Survey, NHANES. National Center for Health Statistics, Hyattsville, MD, USA. 2013.
  • 25
    Kenis G, Maes M. Effects of antidepressants on the production of cytokines. Int J Neuropsychopharmacol 2002; 5: 401412.
  • 26
    Wozniak G, Toska A, Saridi M, Mouzas O. Serotonin reuptake inhibitor antidepressants (SSRIs) against atherosclerosis. Med Sci Monit 2011; 17: RA205RA214.
  • 27
    Rosenblat JD, Cha DS, Mansur RB, McIntyre RS. Inflamed moods: a review of the interactions between inflammation and mood disorders. Prog Neuropsychopharmacol Biol Psychiatry 2014; 53C: 2334.
  • 28
    Barter P. Effects of inflammation on high-density lipoproteins. Arterioscler Thromb Vasc Biol 2002; 22: 10621063.
  • 29
    Esteve E, Ricart W, Fernández-Real JM. Dyslipidemia and inflammation: an evolutionary conserved mechanism. Clin Nutr 2005; 24: 1631.
  • 30
    McIntyre RS, Soczynska JK, Konarski JZ, Kennedy SH. The effect of antidepressants on lipid homeostasis: a cardiac safety concern? Expert Opin Drug Saf 2006; 5: 523537.
  • 31
    Janssen DGA, Caniato RN, Verster JC, Baune BT. A psychoneuroimmunological review on cytokines involved in antidepressant treatment response. Hum Psychopharmacol 2010; 25: 201215.
  • 32
    Stapelberg NJC, Neumann DL, Shum DHK, McConnell H, Hamilton-Craig I. A topographical map of the causal network of mechanisms underlying the relationship between major depressive disorder and coronary heart disease. Aust N Z J Psychiatry 2011; 45: 351369.
  • 33
    Xia Z, DePierre JW, Nässberger L. Tricyclic antidepressants inhibit IL-6, IL-1 beta and TNF-alpha release in human blood monocytes and IL-2 and interferon-gamma in T cells. Immunopharmacology 1996; 34: 2737.
  • 34
    Olefsky JM, Farquhar JW, Reaven GM. Reappraisal of the role of insulin in hypertriglyceridemia. Am J Med 1974; 57: 551560.
  • 35
    Subramanian S, Chait A. Hypertriglyceridemia secondary to obesity and diabetes. Biochim Biophys Acta 2012; 1821: 819825.
  • 36
    Huffman JC, Celano CM, Beach SR, Motiwala SR, Januzzi JL. Depression and cardiac disease: epidemiology, mechanisms, and diagnosis. Cardiovasc Psychiatry Neurol 2013; 2013: 114.
  • 37
    Wulsin LR. Do depressive symptoms increase the risk for the onset of coronary disease? A systematic quantitative review. Psychosom Med 2003; 65: 201210.
  • 38
    Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet 2002; 360: 19031913.
  • 39
    Mancia G. Blood pressure reduction and cardiovascular outcomes: past, present, and future. Am J Cardiol 2007; 100: 3J9J.
  • 40
    Bowling A. Mode of questionnaire administration can have serious effects on data quality. J Public Health (Oxf) 2005; 27: 281291.