Associations Between Depression, Arterial Stiffness, and Metabolic Syndrome Among Adults in the UK Biobank Population Study: A Mediation Analysis | Cardiology | JAMA Psychiatry | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Mediating Pathway of the Association of Depression With Arterial Stiffness
Mediating Pathway of the Association of Depression With Arterial Stiffness

Direct acyclic graph of a structural model of mediation of the association between depression and arterial stiffness by metabolic syndrome. NDE indicates natural direct effects; NIE, natural indirect effects.

Figure 2.  Mean Arterial Stiffness Index
Mean Arterial Stiffness Index

Sex-specific mean arterial stiffness index values. Boxes contain 50% of the data, and the error bars contain the remainder. The horizontal lines indicate medians.

Table 1.  Participant Characteristics at the UK Biobank Assessment
Participant Characteristics at the UK Biobank Assessment
Table 2.  Adjusted Direct and Indirect Associations of Depression With Arterial Stiffness Mediated via Metabolic Syndrome (MetS)a
Adjusted Direct and Indirect Associations of Depression With Arterial Stiffness Mediated via Metabolic Syndrome (MetS)a
Table 3.  Adjusted Direct and Indirect Associations of Depression With Arterial Stiffness Mediated via Subcomponents of Metabolic Syndromea
Adjusted Direct and Indirect Associations of Depression With Arterial Stiffness Mediated via Subcomponents of Metabolic Syndromea
1.
Yusuf  S, Hawken  S, Ounpuu  S,  et al; INTERHEART Study Investigators.  Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study.   Lancet. 2004;364(9438):937-952. doi:10.1016/S0140-6736(04)17018-9PubMedGoogle ScholarCrossref
2.
Hare  DL, Toukhsati  SR, Johansson  P, Jaarsma  T.  Depression and cardiovascular disease: a clinical review.   Eur Heart J. 2014;35(21):1365-1372. doi:10.1093/eurheartj/eht462PubMedGoogle ScholarCrossref
3.
Khambaty  T, Stewart  JC, Gupta  SK,  et al.  Association between depressive disorders and incident acute myocardial infarction in human immunodeficiency virus-infected adults: Veterans Aging Cohort Study.   JAMA Cardiol. 2016;1(8):929-937. doi:10.1001/jamacardio.2016.2716PubMedGoogle ScholarCrossref
4.
Whooley  MA.  Depression and cardiovascular disease: healing the broken-hearted.   JAMA. 2006;295(24):2874-2881. doi:10.1001/jama.295.24.2874PubMedGoogle ScholarCrossref
5.
Chirinos  JA, Segers  P, Hughes  T, Townsend  R.  Large-artery stiffness in health and disease: JACC State-of-the-Art Review.   J Am Coll Cardiol. 2019;74(9):1237-1263. doi:10.1016/j.jacc.2019.07.012PubMedGoogle ScholarCrossref
6.
Seldenrijk  A, van Hout  HP, van Marwijk  HW,  et al.  Carotid atherosclerosis in depression and anxiety: associations for age of depression onset.   World J Biol Psychiatry. 2011;12(7):549-558. doi:10.3109/15622975.2011.583942PubMedGoogle ScholarCrossref
7.
Vlachopoulos  C, Aznaouridis  K, Stefanadis  C.  Prediction of cardiovascular events and all-cause mortality with arterial stiffness: a systematic review and meta-analysis.   J Am Coll Cardiol. 2010;55(13):1318-1327. doi:10.1016/j.jacc.2009.10.061PubMedGoogle ScholarCrossref
8.
Ben-Shlomo  Y, Spears  M, Boustred  C,  et al.  Aortic pulse wave velocity improves cardiovascular event prediction: an individual participant meta-analysis of prospective observational data from 17,635 subjects.   J Am Coll Cardiol. 2014;63(7):636-646. doi:10.1016/j.jacc.2013.09.063PubMedGoogle ScholarCrossref
9.
Chowienczyk  PJ, Kelly  RP, MacCallum  H,  et al.  Photoplethysmographic assessment of pulse wave reflection: blunted response to endothelium-dependent beta2-adrenergic vasodilation in type II diabetes mellitus.   J Am Coll Cardiol. 1999;34(7):2007-2014. doi:10.1016/S0735-1097(99)00441-6PubMedGoogle ScholarCrossref
10.
Perrault  R, Omelchenko  A, Taylor  CG, Zahradka  P.  Establishing the interchangeability of arterial stiffness but not endothelial function parameters in healthy individuals.   BMC Cardiovasc Disord. 2019;19(1):190. doi:10.1186/s12872-019-1167-3PubMedGoogle ScholarCrossref
11.
Ohkuma  T, Ninomiya  T, Tomiyama  H,  et al; Collaborative Group for J-BAVEL (Japan Brachial-Ankle Pulse Wave Velocity Individual Participant Data Meta-Analysis of Prospective Studies).  Brachial-ankle pulse wave velocity and the risk prediction of cardiovascular disease: an individual participant data meta-analysis.   Hypertension. 2017;69(6):1045-1052. doi:10.1161/HYPERTENSIONAHA.117.09097PubMedGoogle ScholarCrossref
12.
Tąpolska  M, Spałek  M, Szybowicz  U,  et al.  Arterial stiffness parameters correlate with estimated cardiovascular risk in humans: a clinical study.   Int J Environ Res Public Health. 2019;16(14):E2547. doi:10.3390/ijerph16142547PubMedGoogle Scholar
13.
Aguilar  M, Bhuket  T, Torres  S, Liu  B, Wong  RJ.  Prevalence of the metabolic syndrome in the United States, 2003-2012.   JAMA. 2015;313(19):1973-1974. doi:10.1001/jama.2015.4260PubMedGoogle ScholarCrossref
14.
Puymirat  E, Simon  T, Steg  PG,  et al; USIK USIC 2000 Investigators; FAST MI Investigators.  Association of changes in clinical characteristics and management with improvement in survival among patients with ST-elevation myocardial infarction.   JAMA. 2012;308(10):998-1006. doi:10.1001/2012.jama.11348PubMedGoogle ScholarCrossref
15.
Mozumdar  A, Liguori  G.  Persistent increase of prevalence of metabolic syndrome among US adults: NHANES III to NHANES 1999-2006.   Diabetes Care. 2011;34(1):216-219. doi:10.2337/dc10-0879PubMedGoogle ScholarCrossref
16.
Fry  A, Littlejohns  TJ, Sudlow  C,  et al.  Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population.   Am J Epidemiol. 2017;186(9):1026-1034. doi:10.1093/aje/kwx246PubMedGoogle ScholarCrossref
17.
Sudlow  C, Gallacher  J, Allen  N,  et al.  UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age.   PLoS Med. 2015;12(3):e1001779. doi:10.1371/journal.pmed.1001779PubMedGoogle Scholar
18.
Dregan  A.  Arterial stiffness association with chronic inflammatory disorders in the UK Biobank study.   Heart. 2018;104(15):1257-1262. doi:10.1136/heartjnl-2017-312610PubMedGoogle ScholarCrossref
19.
UK Biobank.  UK Biobank: Protocol for a Large-scale Prospective Epidemiological Resource. London, United Kingdom: UK Biobank; 2007.
20.
Millasseau  SCKR, Kelly  RP, Ritter  JM, Chowienczyk  PJ.  Determination of age-related increases in large artery stiffness by digital pulse contour analysis.   Clin Sci (Lond). 2002;103(4):371-377. doi:10.1042/cs1030371PubMedGoogle ScholarCrossref
21.
Gunarathne  A, Patel  JV, Gammon  B, Hughes  EA, Lip  GY.  Impact of mean arterial blood pressure on higher arterial stiffness indices in South Asians compared to white Europeans.   J Hypertens. 2008;26(7):1420-1426. doi:10.1097/HJH.0b013e3282ffb42aPubMedGoogle ScholarCrossref
22.
Kollias  A, Stergiou  GS, Dolan  E, O’Brien  E.  Ambulatory arterial stiffness index: a systematic review and meta-analysis.   Atherosclerosis. 2012;224(2):291-301. doi:10.1016/j.atherosclerosis.2012.03.039PubMedGoogle ScholarCrossref
23.
Sollinger  D, Mohaupt  MG, Wilhelm  A, Uehlinger  D, Frey  FJ, Eisenberger  U.  Arterial stiffness assessed by digital volume pulse correlates with comorbidity in patients with ESRD.   Am J Kidney Dis. 2006;48(3):456-463. doi:10.1053/j.ajkd.2006.05.014PubMedGoogle ScholarCrossref
24.
Kroenke  K, Spitzer  RL, Williams  JB, Löwe  B.  An ultra-brief screening scale for anxiety and depression: the PHQ-4.   Psychosomatics. 2009;50(6):613-621.PubMedGoogle Scholar
25.
Löwe  B, Wahl  I, Rose  M,  et al.  A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population.   J Affect Disord. 2010;122(1-2):86-95. doi:10.1016/j.jad.2009.06.019PubMedGoogle ScholarCrossref
26.
Löwe  B, Kroenke  K, Gräfe  K.  Detecting and monitoring depression with a two-item questionnaire (PHQ-2).   J Psychosom Res. 2005;58(2):163-171. doi:10.1016/j.jpsychores.2004.09.006PubMedGoogle ScholarCrossref
27.
Alberti  KG, Eckel  RH, Grundy  SM,  et al; International Diabetes Federation Task Force on Epidemiology and Prevention; Hational Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; International Association for the Study of Obesity.  Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity.   Circulation. 2009;120(16):1640-1645. doi:10.1161/CIRCULATIONAHA.109.192644PubMedGoogle ScholarCrossref
28.
van der Berg  JD, Stehouwer  CD, Bosma  H,  et al.  Associations of total amount and patterns of sedentary behaviour with type 2 diabetes and the metabolic syndrome: the Maastricht Study.   Diabetologia. 2016;59(4):709-718. doi:10.1007/s00125-015-3861-8PubMedGoogle ScholarCrossref
29.
White  WB, Anwar  YA.  Evaluation of the overall efficacy of the Omron office digital blood pressure HEM-907 monitor in adults.   Blood Press Monit. 2001;6(2):107-110. doi:10.1097/00126097-200104000-00007PubMedGoogle ScholarCrossref
30.
Ridker  PM, Buring  JE, Cook  NR, Rifai  N.  C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially healthy American women.   Circulation. 2003;107(3):391-397. doi:10.1161/01.CIR.0000055014.62083.05PubMedGoogle ScholarCrossref
31.
Sesso  HD, Buring  JE, Rifai  N, Blake  GJ, Gaziano  JM, Ridker  PM.  C-reactive protein and the risk of developing hypertension.   JAMA. 2003;290(22):2945-2951. doi:10.1001/jama.290.22.2945PubMedGoogle ScholarCrossref
32.
Gafoor  R, Booth  HP, Gulliford  MC.  Antidepressant utilisation and incidence of weight gain during 10 years’ follow-up: population based cohort study.   BMJ. 2018;361:k1951. doi:10.1136/bmj.k1951PubMedGoogle ScholarCrossref
33.
Stein  A, Pearson  RM, Goodman  SH,  et al.  Effects of perinatal mental disorders on the fetus and child.   Lancet. 2014;384(9956):1800-1819. doi:10.1016/S0140-6736(14)61277-0PubMedGoogle ScholarCrossref
34.
Dregan  A, Matcham  F, Harber-Aschan  L,  et al.  Common mental disorders within chronic inflammatory disorders: a primary care database prospective investigation.   Ann Rheum Dis. 2019;78(5):688-695. doi:10.1136/annrheumdis-2018-214676PubMedGoogle ScholarCrossref
35.
Pitman  A, Suleman  S, Hyde  N, Hodgkiss  A.  Depression and anxiety in patients with cancer.   BMJ. 2018;361:k1415. doi:10.1136/bmj.k1415PubMedGoogle ScholarCrossref
36.
Valeri  L, Vanderweele  TJ.  Mediation analysis allowing for exposure-mediator interactions and causal interpretation: theoretical assumptions and implementation with SAS and SPSS macros.   Psychol Methods. 2013;18(2):137-150. doi:10.1037/a0031034PubMedGoogle ScholarCrossref
37.
Steen  J, Loeys  T, Moerkerke  B, Vansteelandt  S.  Flexible mediation analysis with multiple mediators.   Am J Epidemiol. 2017;186(2):184-193. doi:10.1093/aje/kwx051PubMedGoogle ScholarCrossref
38.
Rybakowski  JK, Wykretowicz  A, Heymann-Szlachcinska  A, Wysocki  H.  Impairment of endothelial function in unipolar and bipolar depression.   Biol Psychiatry. 2006;60(8):889-891. doi:10.1016/j.biopsych.2006.03.025PubMedGoogle ScholarCrossref
39.
Després  JP, Lemieux  I, Bergeron  J,  et al.  Abdominal obesity and the metabolic syndrome: contribution to global cardiometabolic risk.   Arterioscler Thromb Vasc Biol. 2008;28(6):1039-1049. doi:10.1161/ATVBAHA.107.159228PubMedGoogle ScholarCrossref
40.
Ritchie  SA, Connell  JM.  The link between abdominal obesity, metabolic syndrome and cardiovascular disease.   Nutr Metab Cardiovasc Dis. 2007;17(4):319-326. doi:10.1016/j.numecd.2006.07.005PubMedGoogle ScholarCrossref
41.
Veith  RC, Lewis  N, Linares  OA,  et al.  Sympathetic nervous system activity in major depression: basal and desipramine-induced alterations in plasma norepinephrine kinetics.   Arch Gen Psychiatry. 1994;51(5):411-422. doi:10.1001/archpsyc.1994.03950050071008PubMedGoogle ScholarCrossref
42.
Spieker  LE, Hürlimann  D, Ruschitzka  F,  et al.  Mental stress induces prolonged endothelial dysfunction via endothelin-A receptors.   Circulation. 2002;105(24):2817-2820. doi:10.1161/01.CIR.0000021598.15895.34PubMedGoogle ScholarCrossref
43.
Wittstein  IS, Thiemann  DR, Lima  JA,  et al.  Neurohumoral features of myocardial stunning due to sudden emotional stress.   N Engl J Med. 2005;352(6):539-548. doi:10.1056/NEJMoa043046PubMedGoogle ScholarCrossref
44.
Mill  J, Petronis  A.  Molecular studies of major depressive disorder: the epigenetic perspective.   Mol Psychiatry. 2007;12(9):799-814. doi:10.1038/sj.mp.4001992PubMedGoogle ScholarCrossref
45.
Lind  L, Ingelsson  E, Sundström  J, Siegbahn  A, Lampa  E.  Methylation-based estimated biological age and cardiovascular disease.   Eur J Clin Invest. 2018;48(2). doi:10.1111/eci.12872PubMedGoogle Scholar
46.
Tomfohr  LM, Murphy  ML, Miller  GE, Puterman  E.  Multiwave associations between depressive symptoms and endothelial function in adolescent and young adult females.   Psychosom Med. 2011;73(6):456-461. doi:10.1097/PSY.0b013e3182228644PubMedGoogle ScholarCrossref
47.
Seldenrijk  A, Vogelzangs  N, van Hout  HP, van Marwijk  HW, Diamant  M, Penninx  BW.  Depressive and anxiety disorders and risk of subclinical atherosclerosis: findings from the Netherlands Study of Depression and Anxiety (NESDA).   J Psychosom Res. 2010;69(2):203-210. doi:10.1016/j.jpsychores.2010.01.005PubMedGoogle ScholarCrossref
48.
Greaney  JL, Koffer  RE, Saunders  EFH, Almeida  DM, Alexander  LM.  Self-reported everyday psychosocial stressors are associated with greater impairments in endothelial function in young adults with major depressive disorder.   J Am Heart Assoc. 2019;8(4):e010825. doi:10.1161/JAHA.118.010825PubMedGoogle Scholar
49.
Lewis  TT, Sutton-Tyrrell  K, Penninx  BW,  et al.  Race, psychosocial factors, and aortic pulse wave velocity: the Health, Aging, and Body Composition Study.   J Gerontol A Biol Sci Med Sci. 2010;65(10):1079-1085. doi:10.1093/gerona/glq089PubMedGoogle ScholarCrossref
50.
Tiemeier  H, Breteler  MM, van Popele  NM, Hofman  A, Witteman  JC.  Late-life depression is associated with arterial stiffness: a population-based study.   J Am Geriatr Soc. 2003;51(8):1105-1110. doi:10.1046/j.1532-5415.2003.51359.xPubMedGoogle ScholarCrossref
51.
Camacho  Á, McClelland  RL, Delaney  JA,  et al.  Antidepressant use and subclinical measures of atherosclerosis: the Multi-Ethnic Study of Atherosclerosis.   J Clin Psychopharmacol. 2016;36(4):340-346. doi:10.1097/JCP.0000000000000518PubMedGoogle ScholarCrossref
52.
Tong  B, Abosi  O, Schmitz  S, Myers  J, Pierce  GL, Fiedorowicz  JG.  Bipolar disorder and related mood states are not associated with endothelial function of small arteries in adults without heart disease.   Gen Hosp Psychiatry. 2018;51:36-40. doi:10.1016/j.genhosppsych.2017.12.005PubMedGoogle ScholarCrossref
53.
Park  YW, Zhu  S, Palaniappan  L, Heshka  S, Carnethon  MR, Heymsfield  SB.  The metabolic syndrome: prevalence and associated risk factor findings in the US population from the Third National Health and Nutrition Examination Survey, 1988-1994.   Arch Intern Med. 2003;163(4):427-436. doi:10.1001/archinte.163.4.427PubMedGoogle ScholarCrossref
54.
Bayturan  O, Tuzcu  EM, Lavoie  A,  et al.  The metabolic syndrome, its component risk factors, and progression of coronary atherosclerosis.   Arch Intern Med. 2010;170(5):478-484. doi:10.1001/archinternmed.2009.551PubMedGoogle ScholarCrossref
55.
Groenwold  RH, Sterne  JA, Lawlor  DA, Moons  KG, Hoes  AW, Tilling  K.  Sensitivity analysis for the effects of multiple unmeasured confounders.   Ann Epidemiol. 2016;26(9):605-611. doi:10.1016/j.annepidem.2016.07.009PubMedGoogle ScholarCrossref
56.
Olfson  M, Marcus  SC.  National patterns in antidepressant medication treatment.   Arch Gen Psychiatry. 2009;66(8):848-856. doi:10.1001/archgenpsychiatry.2009.81PubMedGoogle ScholarCrossref
57.
Mason  C, Katzmarzyk  PT.  Variability in waist circumference measurements according to anatomic measurement site.   Obesity (Silver Spring). 2009;17(9):1789-1795. doi:10.1038/oby.2009.87PubMedGoogle ScholarCrossref
58.
Simonetti  GD, Eisenberger  U, Bergmann  IP, Frey  FJ, Mohaupt  MG.  Pulse contour analysis: a valid assessment of central arterial stiffness in children?   Pediatr Nephrol. 2008;23(3):439-444. doi:10.1007/s00467-007-0693-xPubMedGoogle ScholarCrossref
59.
Xi  B, Zhang  L, Mi  J.  Reduced arterial compliance associated with metabolic syndrome in Chinese children and adolescents.   Biomed Environ Sci. 2010;23(2):102-107. doi:10.1016/S0895-3988(10)60038-4PubMedGoogle ScholarCrossref
60.
Broadley  AJ, Korszun  A, Jones  CJ, Frenneaux  MP.  Arterial endothelial function is impaired in treated depression.   Heart. 2002;88(5):521-523. doi:10.1136/heart.88.5.521PubMedGoogle ScholarCrossref
61.
Piepoli  MF, Hoes  AW, Agewall  S,  et al; ESC Scientific Document Group.  2016 European Guidelines on cardiovascular disease prevention in clinical practice: the Sixth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of 10 societies and by invited experts): developed with the special contribution of the European Association for Cardiovascular Prevention & Rehabilitation (EACPR).   Eur Heart J. 2016;37(29):2315-2381. doi:10.1093/eurheartj/ehw106PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    1 Comment for this article
    EXPAND ALL
    Smoking cessation for prevention of cardiovascular diseases and depression: killing two birds with one stone.
    alain braillon, MD, PhD | University Hospital, 80000 Amiens, France. braillon.alain@gmail.com
    The conclusion of the retrospective study in an healthy and wealthy English population about associations between depression, arterial stiffness index (ASI), and metabolic syndrome deserved comment.(1)

    Firstly, its introductory warning “systematic assessment of ASI in the general population is not currently recommended.” must be applauded. Indeed, early detection of subclinical atherosclerosis in asymptomatic subjects has been an old concept to target people at high risk for cardiovascular disease for individualizing the decision to treat. However, since 1993 it remains a concept, no clinical trial has shown yet that screening with coronary artery calcium and carotid artery intima-media thickness improved
    clinical outcomes in those at high risk.(2)

    Secondly, the association between depression and ASI, in one-third of cases, deserves scrutiny:
    a) retrospective studies only generate hypotheses which must be confirmed prospectively. Did Dregan and colleagues pre-registered their hypotheses (e.g. using the Open Science Framework) to avoid p-hacking?
    b) adjustment of confounding variables, always a challenge, was hardy possible. Smoking, highly prevalent in the series (37% in men with depression vs 33% without,(1) not only causes cardiovascular diseases but it also generates and aggravates depression while cessation reduces it.(3) Moreover, the binary classification, current smokers/non-smokers, cannot account for dose-duration effects. Last, alcohol use, a most depressing substance, was overlooked. Why biobanks put the cart before the horse, being concerned by biomarkers or genomics but overlooking obvious major issues?

    Thirdly, why pledging for “preventive priorities” targeting waist circumference in women and hypertriglyceridemia in men? In 1984, Marmot showed that individuals in the lowest, vs highest, socio-economic status group, were 3 times more likely to die of cardiovascular diseases.(2) Tobacco, alcohol and processed foods are industrial products causing the modern epidemics. “Industry-born pandemics” as host-reservoirs are industries. Social determinants (poverty, education...) are environmental factors, or risk factors, influencing their spread as public policies protecting vested industrial interests rather than citizens’ health.(2)

    Last, regarding depression:
    a) in 2012 Frances, warned that DSM-5 “will medicalise normality and result in a glut of unnecessary and harmful drug prescription.”
    b) most cases of depression, even major or persisting, are successfully treated with psychosocial interventions, the patients’ preference. Psychosocial interventions are beneficial for self-esteem and social functioning, without adverse effects. Among antidepressants, escitalopram/citalopram ranks first for QT prolongation, deadly torsades de pointe and, other cardiovascular complications including hypertension;(4)
    c) the CODIACS-QoL Randomized Clinical Trial showed depression screening after acute coronary syndromes is useless.(5)

    1 Dregan A, Rayner L, Davis KAS, et al. JAMA Psychiatry 2020. Online Jan 29. doi:10.1001/jamapsychiatry.2019.4712
    2 Braillon A. Lancet. 2019;393:1805–1806.
    3 Cinciripini PM, Robinson JD, Karam-Hage M, et al. JAMA Psychiatry. 2013;70:522–533.
    4 Ojero-Senard A, Benevent J, Bondon-Guitton E, et al.. Psychopharmacology (Berl) 2017;234:3075–3081.
    5 Kronish IM, Moise N, Cheung YKet al. JAMA Intern Med 2019. Online Oct 21. doi: 10.1001/jamainternmed.2019.4518
    CONFLICT OF INTEREST: None Reported
    READ MORE
    Original Investigation
    January 29, 2020

    Associations Between Depression, Arterial Stiffness, and Metabolic Syndrome Among Adults in the UK Biobank Population Study: A Mediation Analysis

    Author Affiliations
    • 1Department of Psychological Medicine, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London, London, United Kingdom
    • 2Biostatistics and Health Informatics, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London, London, United Kingdom
    • 3Centre for Implementation Science, Health Services and Population Research Department, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London, London, United Kingdom
    • 4South London and Maudsley NHS Foundation Trust, London, United Kingdom
    JAMA Psychiatry. 2020;77(6):598-606. doi:10.1001/jamapsychiatry.2019.4712
    Key Points

    Question  To what extent is the association of depression with peripherally assessed arterial stiffness mediated by metabolic syndrome?

    Findings  In this large population-based cohort study of 124 445 adults aged 40 to 69 years, up to one-third of the association of depression with arterial stiffness was mediated by metabolic syndrome. The addition of inflammation to metabolic syndrome further increased the proportion of the mediated association of depression with arterials stiffness.

    Meaning  Combined data on metabolic syndrome and inflammation may improve the early identification of future cardiovascular risk among adult patients with depression.

    Abstract

    Importance  Previous research has linked a history of depression with arterial stiffness (AS) during midlife.

    Objective  To assess the association of depression with elevated midlife AS and to investigate the extent to which this association is mediated via metabolic syndrome (MetS).

    Design, Settings, and Participants  This population-based retrospective cohort study analyzed data collected between March 2006 and December 2010 from 124 445 participants aged 40 to 69 years from the UK Biobank. Participants without data on AS at baseline (n = 332 780) or who reported a previous diagnosis of cardiovascular disease (n = 45 374) were not eligible. Data analysis was performed from May to August 2019.

    Exposures  Lifetime history of depression was assessed via verbal interview and linked hospital-based clinical depression diagnosis. Metabolic syndrome was defined as the presence of 3 or more of hypertension, dyslipidemia, hyperglycemia, hypertriglyceridemia, and unhealthy waist circumference.

    Main Outcomes and Measures  Peripherally assessed AS index (ASI) using digital photoplethysmography.

    Results  Of 124 445 included participants with ASI assessed, 71 799 (57.7%) were women, and the mean (SD) age was 56 (8) years. A total of 10 304 participants (8.3%) reported a history of depression. Study findings indicated a significant direct association between depression and ASI levels (β = 0.25; 95% CI, 0.17-0.32). A significant indirect association was also observed between depression and ASI levels (β = 0.10; 95% CI, 0.07-0.13), indicating that 29% of the association of depression with ASI was mediated by MetS. The proportion of mediation increased to 37% when C-reactive protein was added to the MetS criteria (direct association: β = 0.21; 95% CI, 0.15-0.28; indirect association: β = 0.13; 95% CI, 0.10-0.17). Concerning components of MetS, the strongest indirect association was for waist circumference, accounting for 25% of the association between depression and ASI levels (direct association: β = 0.26; 95% CI, 0.18-0.34; indirect association: β = 0.09; 95% CI, 0.06-0.11). Among men, hypertriglyceridemia accounted for 19% of the association between depression and ASI (direct association: β = 0.22; 95% CI, 0.05-0.40; indirect association: β = 0.05; 95% CI, 0.02-0.08).

    Conclusions and Relevance  One-third of the association of depression with elevated ASI levels during midlife may be accounted for by combined MetS and inflammatory processes. Unhealthy waist circumference and hypertriglyceridemia emerged as the most important potential targets for preventive interventions within women and men, respectively.

    Introduction

    People diagnosed with depression are at increased risk of major cardiovascular events (MACEs), including coronary heart disease and stroke.1-4 Arterial stiffness (AS), a hardening of the artery wall because of aging and diverse pathologic states, is a key mediator of MACE,5 and growing evidence has found an association of depression with AS.6 Components of metabolic syndrome (MetS), such as hypertension, dyslipidemia, obesity, and hyperglycemia, represent leading promoters of AS.5 Consistent with the key role of AS in predicting MACE incidence, carotid pulse wave velocity represents the criterion-standard measure for central AS,7,8 and several other noninvasive techniques (eg, photoplethysmography) are available as indicators of specific AS parameters (eg, wave reflections).5,9 While the interchangeability of these measures needs further research,10 they were found to be independently associated with incident MACE across different subgroups.9,11,12 Therefore, a key unanswered question for public health is the extent to which the adverse effect of depression on MACE might be mitigated by targeting modifiable metabolic mediators. To address this concern, we need to better understand the role of MetS in mediating the association of depression with AS. The extent to which this mediation differs by sex is critical given that the prevalence of MACE, MetS, and depression is often greater among women compared with men.13,14 Sex-specific mediating effects of MetS are generally underresearched but are a growing issue given the increasing prevalence of MetS in women.15 The present study adopted a mediation analysis approach to quantify to what extent MetS and its individual components mediated the association of depression with AS. The study also aimed to investigate the extent to which the hypothesized mediation varied by sex.

    Methods
    Data

    A population-based, retrospective cohort study was implemented within the 502 299 adults aged 40 to 70 years in the UK Biobank. These participants were recruited from 23 centers between March 2006 and December 2010 throughout the United Kingdom. Participants in the UK Biobank tended to be healthier than the general population, more likely to be women, and less likely to live in socioeconomically deprived areas relative to nonparticipants.16 The study collected detailed phenotype and genotype data, including sociodemographic, lifestyle, clinical diagnosis, treatment, genetic, imaging, and physiological parameters. A detailed description of the UK Biobank data was provided elsewhere.17 The North West Multi-centre Ethics Committee granted ethical approval to UK Biobank, and all participants provided written informed consent.

    This study was restricted to a subset of participants with AS data available at the initial assessment (March 2006 to December 2010). Participants diagnosed with MACE (eg, angina, myocardial infarction, stroke, peripheral artery disease) or type 2 diabetes prior to AS assessment and those with treated hypertension or dyslipidemia prior to depression were excluded. Participants with a depression history prior to AS assessment represented the exposed group (n = 10 304), and those without a history of depression made up the comparison group (n = 114 141). A detailed description of the selection process for the study analytic sample is included in the (eFigure in the Supplement).

    Outcome

    The primary outcome measure was the AS index (ASI) derived from the analysis of digital volume pulse, an indirect method to assess AS peripherally.18 Arterial stiffness index data were collected at baseline via the PulseTrace PCA 2 (careFusion), which uses finger photoplethysmography to assess the pulse waveform using an infrared sensor placed on the index finger of the participant’s dominant hand.19 The pulse waveform comprises a systolic peak and second diastolic peak, and the transit time (peak-to-peak time [PPT]) between the 2 peaks is related to the time it takes for the pulse wave to travel through the peripheral arterial tree.20,21 This path length is proportional to a person’s height (h), enabling the calculation of an index of large artery stiffness using the formula ASI = h / PPT.21 The ASI was predictive of future risk of MACE22 and was correlated with the pulse wave velocity.20,23

    Exposure

    The study used multiple sources to determine depression caseness. Self-reported data relied on the question, “In the touch screen you selected that you have been told by a doctor that you have other (non-cancer) serious illnesses or disabilities, could you now tell me what they are?” asked by a trained nurse during the verbal interview stage of data collection (at the time of AS assessment). The nurse used a tree structure organized by system and loosely based on International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), codes to record a depression diagnosis (UK Biobank field: 20002) using given codes (UK Biobank codes: 1286 and 1531). Linked hospital admission records were also used to identify a primary or secondary diagnosis of depression (ICD-10 codes: F32 [single episode depression], F33 [recurrent depression], F34 [persistent mood disorders], F38 [other mood disorders], and F39 [unspecified mood disorders]). At baseline, participants also completed the 4-item Patient Health Questionnaire (PHQ-4), a brief screening tool that consists of a 2-item depression scale (PHQ-2) and the 2-item Generalized Anxiety Disorder scale.24,25 Thus, caseness for depression was determined by a positive answer to the self-reported question, a primary or secondary diagnosis of clinical depression prior to AS assessment, or a score of 3 or more (cutoff score for depressive disorder26) on the PHQ-2 tool at the baseline assessment visit. Self-reported data on the year of the first depression diagnosis and the start date of a depression episode recorded in the hospital records were used to define a depression event prior to ASI assessment.

    Mediators

    The study’s main mediator variable was MetS defined as the presence of 3 or more of the following: unhealthy waist circumference, hypertension, dyslipidemia, hypertriglyceridemia, and hyperglycemia.27,28 These measures were assessed at the initial visit during which AS was measured. Waist circumference and blood pressure data were assessed by a trained nurse. Waist circumference was measured at the smallest part of the trunk using a 200-cm tape measure (SECA). We used established reference values for women (≥88 cm) and men (≥102 cm) as cutoff points for defining unhealthy waist circumference.27 Two systolic and diastolic blood pressure measurements were taken after participants were at rest for 5 or more minutes using IntelliSense blood pressure monitor model HEM-907XL (Omron).29 The mean of the 2 measures was used, and hypertension was defined as a systolic blood pressure of 130 mm Hg or greater or a diastolic blood pressure of 85 mm Hg or greater or as taking antihypertensive drugs. Dyslipidemia was defined as a high-density lipoprotein cholesterol level less than 40 mg/dL (to convert to millimoles per liter, multiply by 0.0259) among men and less than 50 mg/dL among women or as taking statin drugs. Hypertriglyceridemia was defined as triglyceride levels of 150 mg/dL or greater (to convert to millimoles per liter, multiply by 0.0113) or taking triglyceride-lowering drugs. Hyperglycemia was defined as fasting blood glucose levels greater than 110 mg/dL (to convert to millimoles per liter, multiply by 0.0555) or taking medication for diabetes. To evaluate recent evidence that C-reactive protein (CRP) adds prognostic value to subsequent risk of MACE,30 a second MetS mediator was developed that also included high-sensitivity CRP as part of MetS. In line with previous research,30,31 a cutoff point greater than 3.0 mg/L (to convert to nanomoles per liter, multiply by 9.524) for CRP concentration level was used as an indicator of elevated MACE risk. To ensure comparability, the same definition of MetS was used (ie, 3 or more metrics met). A self-reported antidepressant therapy variable was also developed to evaluate a potential mediating role for antidepressant medication.32 Self-reported generic and trade names of current medications were recorded by a trained nurse. The British National Formulary was used as a guide to identify antidepressant drugs prescribed within the United Kingdom.

    Covariates

    Factors known to be associated with depression, ASI, and MetS were included in the analyses as confounders. These variables included self-reported age (linear and quadratic terms), sex (female vs male), race/ethnicity (recoded as white, black, Asian, mixed, and other), living in a socioeconomically deprived area (quintiles based on Townsend deprivation indices, an official measure for small-area [approximately 1500 residents] relative material deprivation), and parental history of severe depression as a marker of biological vulnerability.33 Self-reported and Hospital Episode Statistics clinical data were used to identify predepression diagnoses of hypertension and high cholesterol as well as common long-term disorders associated with depression and AS, including cancer, chronic kidney disease, inflammatory disorders (eg, rheumatoid arthritis, psoriasis), liver disease, and chronic obstructive pulmonary disease.18,34,35 Self-reported data on the age at which current smokers started to smoke or the age ex-smokers stopped smoking were used to group participants into current smokers or nonsmokers prior to first depression episode.

    Statistical Analysis

    Descriptive statistics (eg, frequency, percentages, means, SDs) were used to contrast baseline differences among participants with depression and their comparison group without depression. To explicate the association of depression with AS, indirect associations acting through MetS as a mediating variable and direct associations not mediated by MetS and their SEs were quantified (Figure 1). A parametric regression approach (paramed package in Stata version 15 [StataCorp]) was used to estimate the total effect, the natural indirect effects (NIE), and natural direct effects (NDE) of depression on ASI.36,37 Two models were estimated: a multivariate logistic regression model for MetS (mediator) conditional on depression (exposure) and all study confounders and a multivariate linear regression model for the ASI (outcome) conditional on depression, MetS, and all study confounders. The NDE represented the effect of depression on ASI that was independent of MetS. An NIE represented the proportion of depression that could be explained by its association with changes in MetS over time. The paramed approach was selected because it tested for the interaction between depression and MetS (or its individual subcomponents) in the ASI regression model using counterfactual definitions of direct and indirect effects.36 To quantify the magnitude of mediation, the study estimated the proportion of the association mediated by MetS (NIE/[NDE + NIE]). Tests for interaction (testparm package) between sex and depression and between sex and MetS were statistically significant, and both pooled and sex-stratified mediation were estimated. All analyses were estimated using bootstrapping (500 replications) to recover the correct SEs for direct and indirect effects. Missing data on study variables ranged from 0% to 15% (eg, dyslipidemia), and multiple imputation using chained equations with 10 imputations was performed to optimize the validity of the findings. All P values were 2-tailed, and statistical significance was set at a P value less than .05.

    To quantify the independent contribution of the individual components of MetS, the study also estimated mediation by these components in separate analyses. Planned additional sensitivity analyses were performed to examine the robustness of study results. First, the analyses were repeated using a lower threshold (2 or more factors) for MetS as well as defining MetS as a continuous variable to account for misclassification bias. Second, to account for protopathic bias, the analyses excluded all participants diagnosed with depression in the 12 months prior to the AS assessment. Third, we assessed a potential mediating role of current use of antidepressant medication. Since previous studies found no association of the type of antidepressant drugs with AS,38 all antidepressant drugs were grouped together. To validate this procedure, additional analyses were performed using selective serotonin reuptake inhibitors and tricyclic drugs as independent mediators. Fourth, we estimated whether the mediation of MetS varied according to depression severity as assessed by the PHQ-4. Fifth, for comparative purposes, we excluded people with treated individual components of MetS.

    Results

    Of 124 445 included participants with ASI assessed, 71 799 (57.7%) were women, and the mean (SD) age was 56 (8) years. A total of 10 304 participants (8.3%) reported a history of depression, and 114 141 participants (91.7%) were without depression (eFigure in the Supplement). The median (interquartile range) time period from depression diagnosis to AS measurement was 12 (5-24) years. Table 1 shows sex-specific baseline characteristics for the depression and comparator groups. Relative to their comparison group, a higher proportion of men diagnosed with depression (30.4% [1051 of 3453] vs 20.1% [9910 of 49 193]) and women diagnosed with depression (26.2% [1796 of 6851] vs 19.3% [12 537 of 64 948]) lived in the most socioeconomically deprived areas by quintile. A total of 3681 women (53.7%) and 1715 men (49.7%) with depression reported current use of antidepressant drugs compared with 2114 women (3.3%) and 705 men (1.4%) without depression. Metabolic syndrome was twice as common among women with depression (31.9% [2188 of 6851] vs 16.9% [10 976 of 64 948]) and men with depression (41.9% [1447 of 3453] vs 21.7% [10 666 of 49 193]) compared with those without depression. Figure 2 indicates that women and men with a history of depression presented with higher median values for ASI compared with the comparison group.

    Table 2 presents the total, direct associations, and indirect associations of depression with ASI assessed peripherally. The indirect associations via MetS implied that we would, on average, observe a 0.10-point (95% CI, 0.07-0.13; P = .001) increase in ASI levels among participants with a history of depression. The proportion of the association between depression and ASI mediated by MetS was 29%. The direct association of depression indicated that we would, on average, observe a 0.25-point (95% CI, 0.17-0.32; P < .001) increase in ASI levels if all study participants were free of MetS. Similar trends emerged within sex-stratified analyses. Notably, the inclusion of CRP as an extra component of MetS increased the magnitude of MetS-mediated association to 37%, more so for men (to 52%) compared with women (to 34%).

    Next, the study calculated the estimates of direct and indirect associations for specific subcomponents of MetS. As documented in Table 3, the MetS subcomponent that accounted for the largest proportion of the mediated association of depression with ASI was unhealthy waist circumference (25%), followed by CRP level (15%). Among men, hypertriglyceridemia was the subcomponent that accounted for the largest proportion (19%) of the indirect association between MetS and ASI (β = 0.05; 95% CI, 0.02-0.08; P = .002).

    Additional Analyses

    Analyses presented in eTable 1 in the Supplement revealed only a modest mediation of antidepressant medication use on the association of depression with peripheral ASI. Estimation models that adjusted for protopathic bias and those that excluded people with treated MetS components (eTable 4 in the Supplement) validated the study findings. A lower cutoff point (2 or more factors) for the MetS criteria resulted in an increase in the proportion of the association between depression and ASI mediated by MetS (44% vs 26%). Defining MetS as a continuous indicator modestly increased the estimates. No variation in estimates emerged according to the severity of depression or type of antidepressant drugs (eTable 2 in the Supplement).

    Discussion

    In a large population of middle-aged adults, a history of depression was associated with an increase in the risk of AS assessed peripherally, and this association was partly mediated through MetS. Specifically, 29% of the association of depression with ASI appeared to be mediated through MetS. The addition of high-sensitivity CRP to the MetS definition increased the proportion of the association between depression and ASI mediated by MetS, particularly among men (from 26% to 52%). Regarding individual components of MetS, unhealthy waist circumference mediated the highest proportion of the association of depression with ASI among women, while hypertriglyceridemia mediated the largest proportion of the association of depression with ASI within men.

    Our study contributes to ongoing research efforts to better understand the risk of MACE associated with depression. Specifically, we identified combined MetS and inflammatory processes as a potentially influential indirect pathway from depression to ASI in a middle-aged adult population. This finding reinforces the value of developing effective complex interventions that target both MetS and inflammatory processes as means to prevent MACE among people diagnosed with depression. Abdominal obesity has been proposed to be the most prevalent form of MetS39 and its magnitude to be directly related to MetS prognosis.40 These suggestions may account for waist circumference emerging as the strongest independent mediator in this study. Triglycerides are considered a clinical marker of excess abdominal obesity,39 which may partly explain our study evidence around hypertriglyceridemia. Several explanations are possible for the unexplained proportion of depression associated with AS. For instance, psychological stress may cause deregulation in the sympathetic nervous system and hypothalamic-pituitary-adrenal axis,41 which in turn can lead to AS42 and elevated risk of MACE.43 Relatedly, epigenetic factors linking depression and MACE through pathways not captured by this study could mediate AS. Early-life traumatic events (underlying depression) have lasting epigenetic effects via DNA methylation,44 a process associated with incident MACE.45 Our understanding of the genetic basis of depression is incomplete, and some of the unexplained risk of AS may also be due to unidentified genetic causes. The study attempted to account for possible biological influences by adjusting for parental depression.

    To our knowledge, this is the first study to document that the association between depression and increased AS46-48 may be partly mediated via MetS. This estimation strategy may account for some of the inconsistent evidence about the association of depression with ASI in previous research.49 As well as methodological concerns (eg, small samples, diverse depression criteria), prior studies have focused solely on the main effect of depression on AS and have overlooked the possibility of multiple pathways and processes (eg, MetS, inflammation) linking them. The difference in the strengths of the association of depression with AS across studies47,49-51 implies the existence of several mechanisms mediating this association, as reported here. Further, our findings reinforce earlier evidence from different populations about the added prognostic information provided by CRP when incorporated into the MetS measure.30 Our finding that the association of depression with ASI did not appear to be mediated by antidepressant medication use is in line with studies that found no link between antidepressant medication and ASI.51,52 All MetS components (including CRP) are well-established risk factors for AS,13,53 and our findings endorse these as important mediators for the association of depression with AS. A 2010 systematic review54 identified obesity and hypertriglyceridemia as the strongest MetS individual predictors of atherosclerosis, which is in line with our findings.

    Strengths and Limitations

    In addition to the large population, the rich phenotypic data available within the UK Biobank allowed for data integration from multiple sources to define depression caseness. Additionally, the availability of objective measures of MetS and AS enhanced the robustness of our study findings.

    Several limitations need consideration. A critical assumption of mediation analysis models is that of no exposure-induced mediator-outcome confounding, and any violation of this assumption would have implications for the identification of natural associations. If unmeasured confounders (eg, biological predisposition) are associated with the measured confounders (eg, parental depression) in a study, any potential influence of the former is at least partly removed via adjustment for the latter.55 Another limitation is the self-reported data on depression that may have led to misclassification of depression caseness, which may not be random, although the inclusion of multiple data sources was adopted to attenuate misclassification bias. The study sample comprised middle-aged participants, which may limit the generalizability of the study findings to other population subgroups, especially in light of likely higher comedication rates in the US population.56 The measurement of waist circumference in our study was found to result in lower mean values than more common measures based on umbilicus, iliac crest, or midpoint.57 It is possible that defining hypertension, dyslipidemia, hypertriglyceridemia, and hyperglycemia to include those taking pharmacotherapy for these conditions may explain in part why waist circumference and triglycerides (less frequently treated pharmacologically compared with hypertension and hyperglycemia) prevailed as the most important contributors to the association of depression with AS. However, sensitivity analyses that excluded participants with treated MetS components validated the study main findings. The UK Biobank used photoplethysmography as a measure of AS because of its ease of use and to mitigate operator-dependent variability. Photoplethysmography has been only moderately correlated with carotid pulse wave velocity,20 as it relates best to peripheral AS rather than central AS, possibly reflecting different pathophysiological states. There is support for photoplethysmography repeatability,20,58 and MetS was associated with increased photoplethysmography-assessed ASI.59 Our research documented the prognostic value of ASI across multiple long-term disorders.18 The study found no variation in MetS-mediated associations according to depression severity or type of antidepressant drug. Since only current use of antidepressant drugs was reported, we cannot exclude the possibility of an antidepressant treatment time lag associated with AS. However, the study findings are in line with evidence suggesting no variation in AS according to the type of antidepressant drug.60

    Conclusions

    While the systematic assessment of ASI in the general population is not currently recommended, it represents a promising screening tool to discriminate among subgroups at mild to moderate risk of future MACE.61 Our study findings identified that about one-third of the proportion of the association between depression and ASI, and consequently, the risk of MACE, may be potentially prevented by addressing the combined effects of MetS and inflammation. This suggestion is contingent on the effective implementation of complex interventions (with several interacting components) targeting multiple MetS and inflammatory processes at the population level. The study also endorsed sex-specific preventive priorities for subcomponents of MetS, including waist circumference (in women) and hypertriglyceridemia (in men).

    Back to top
    Article Information

    Accepted for Publication: December 1, 2019.

    Corresponding Author: Alex Dregan, PhD, Department of Psychological Medicine, Institute of Psychiatry, Psychology, and Neuroscience, King’s College London, De Crespigny Park, London SE5 8AF, United Kingdom (alexandru.dregan@kcl.ac.uk).

    Published Online: January 29, 2020. doi:10.1001/jamapsychiatry.2019.4712

    Author Contributions: Dr Dregan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

    Study concept and design: Dregan, Bakolis, Arias de la Torre, Hatch, Hotopf.

    Acquisition, analysis, or interpretation of data: Dregan, Rayner, Davis, Bakolis, Arias de la Torre, Das-Munshi, Stewart, Hotopf.

    Drafting of the manuscript: Dregan, Das-Munshi, Hotopf.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Dregan, Bakolis, Arias de la Torre, Hatch.

    Obtained funding: Hotopf.

    Administrative, technical, or material support: Dregan, Rayner, Hotopf.

    Study supervision: Dregan, Hotopf.

    Conflict of Interest Disclosures: Dr Das-Munshi has received grants from the Health Foundation and Academy of Medical Sciences. Dr Hatch has received grants from the Economic and Social Research Council, Guy’s and St Thomas’ Charity, National Institute for Health Research, and the Wellcome Trust and is a member of the Mental Health Delivery Group: Workforce Equalities Subgroup, Health Education England, and the Workforce Race and Equality Standard of the NHS England Strategic Advisory Group. Dr Stewart has received grants from Janssen Pharmaceuticals and has cosupervised PhD students for GlaxoSmithKline and Takeda Pharmaceuticals. Dr Hotopf has received grants from King’s College London, Innovative Medicines Initiative, and European Federation of Pharmaceutical Industries and Associations. No other disclosures were reported.

    Funding/Support: This article was partially funded by the National Institute for Health Research Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. Drs Dregan and Arias de la Torre are supported by grant MR/S028188/1 from the Medical Research Council. Dr Bakolis is funded by the National Institute for Health Research Collaboration for Leadership in Applied Health Research and Care South London at King’s College London NHS Foundation Trust.

    Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    Disclaimer: The views expressed are those of the authors and not necessarily those of the NHS England, the National Institute for Health Research, or the Department of Health and Social Care.

    Additional Information: The study was conducted under UK Biobank project number 34554.

    References
    1.
    Yusuf  S, Hawken  S, Ounpuu  S,  et al; INTERHEART Study Investigators.  Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study.   Lancet. 2004;364(9438):937-952. doi:10.1016/S0140-6736(04)17018-9PubMedGoogle ScholarCrossref
    2.
    Hare  DL, Toukhsati  SR, Johansson  P, Jaarsma  T.  Depression and cardiovascular disease: a clinical review.   Eur Heart J. 2014;35(21):1365-1372. doi:10.1093/eurheartj/eht462PubMedGoogle ScholarCrossref
    3.
    Khambaty  T, Stewart  JC, Gupta  SK,  et al.  Association between depressive disorders and incident acute myocardial infarction in human immunodeficiency virus-infected adults: Veterans Aging Cohort Study.   JAMA Cardiol. 2016;1(8):929-937. doi:10.1001/jamacardio.2016.2716PubMedGoogle ScholarCrossref
    4.
    Whooley  MA.  Depression and cardiovascular disease: healing the broken-hearted.   JAMA. 2006;295(24):2874-2881. doi:10.1001/jama.295.24.2874PubMedGoogle ScholarCrossref
    5.
    Chirinos  JA, Segers  P, Hughes  T, Townsend  R.  Large-artery stiffness in health and disease: JACC State-of-the-Art Review.   J Am Coll Cardiol. 2019;74(9):1237-1263. doi:10.1016/j.jacc.2019.07.012PubMedGoogle ScholarCrossref
    6.
    Seldenrijk  A, van Hout  HP, van Marwijk  HW,  et al.  Carotid atherosclerosis in depression and anxiety: associations for age of depression onset.   World J Biol Psychiatry. 2011;12(7):549-558. doi:10.3109/15622975.2011.583942PubMedGoogle ScholarCrossref
    7.
    Vlachopoulos  C, Aznaouridis  K, Stefanadis  C.  Prediction of cardiovascular events and all-cause mortality with arterial stiffness: a systematic review and meta-analysis.   J Am Coll Cardiol. 2010;55(13):1318-1327. doi:10.1016/j.jacc.2009.10.061PubMedGoogle ScholarCrossref
    8.
    Ben-Shlomo  Y, Spears  M, Boustred  C,  et al.  Aortic pulse wave velocity improves cardiovascular event prediction: an individual participant meta-analysis of prospective observational data from 17,635 subjects.   J Am Coll Cardiol. 2014;63(7):636-646. doi:10.1016/j.jacc.2013.09.063PubMedGoogle ScholarCrossref
    9.
    Chowienczyk  PJ, Kelly  RP, MacCallum  H,  et al.  Photoplethysmographic assessment of pulse wave reflection: blunted response to endothelium-dependent beta2-adrenergic vasodilation in type II diabetes mellitus.   J Am Coll Cardiol. 1999;34(7):2007-2014. doi:10.1016/S0735-1097(99)00441-6PubMedGoogle ScholarCrossref
    10.
    Perrault  R, Omelchenko  A, Taylor  CG, Zahradka  P.  Establishing the interchangeability of arterial stiffness but not endothelial function parameters in healthy individuals.   BMC Cardiovasc Disord. 2019;19(1):190. doi:10.1186/s12872-019-1167-3PubMedGoogle ScholarCrossref
    11.
    Ohkuma  T, Ninomiya  T, Tomiyama  H,  et al; Collaborative Group for J-BAVEL (Japan Brachial-Ankle Pulse Wave Velocity Individual Participant Data Meta-Analysis of Prospective Studies).  Brachial-ankle pulse wave velocity and the risk prediction of cardiovascular disease: an individual participant data meta-analysis.   Hypertension. 2017;69(6):1045-1052. doi:10.1161/HYPERTENSIONAHA.117.09097PubMedGoogle ScholarCrossref
    12.
    Tąpolska  M, Spałek  M, Szybowicz  U,  et al.  Arterial stiffness parameters correlate with estimated cardiovascular risk in humans: a clinical study.   Int J Environ Res Public Health. 2019;16(14):E2547. doi:10.3390/ijerph16142547PubMedGoogle Scholar
    13.
    Aguilar  M, Bhuket  T, Torres  S, Liu  B, Wong  RJ.  Prevalence of the metabolic syndrome in the United States, 2003-2012.   JAMA. 2015;313(19):1973-1974. doi:10.1001/jama.2015.4260PubMedGoogle ScholarCrossref
    14.
    Puymirat  E, Simon  T, Steg  PG,  et al; USIK USIC 2000 Investigators; FAST MI Investigators.  Association of changes in clinical characteristics and management with improvement in survival among patients with ST-elevation myocardial infarction.   JAMA. 2012;308(10):998-1006. doi:10.1001/2012.jama.11348PubMedGoogle ScholarCrossref
    15.
    Mozumdar  A, Liguori  G.  Persistent increase of prevalence of metabolic syndrome among US adults: NHANES III to NHANES 1999-2006.   Diabetes Care. 2011;34(1):216-219. doi:10.2337/dc10-0879PubMedGoogle ScholarCrossref
    16.
    Fry  A, Littlejohns  TJ, Sudlow  C,  et al.  Comparison of sociodemographic and health-related characteristics of UK Biobank participants with those of the general population.   Am J Epidemiol. 2017;186(9):1026-1034. doi:10.1093/aje/kwx246PubMedGoogle ScholarCrossref
    17.
    Sudlow  C, Gallacher  J, Allen  N,  et al.  UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age.   PLoS Med. 2015;12(3):e1001779. doi:10.1371/journal.pmed.1001779PubMedGoogle Scholar
    18.
    Dregan  A.  Arterial stiffness association with chronic inflammatory disorders in the UK Biobank study.   Heart. 2018;104(15):1257-1262. doi:10.1136/heartjnl-2017-312610PubMedGoogle ScholarCrossref
    19.
    UK Biobank.  UK Biobank: Protocol for a Large-scale Prospective Epidemiological Resource. London, United Kingdom: UK Biobank; 2007.
    20.
    Millasseau  SCKR, Kelly  RP, Ritter  JM, Chowienczyk  PJ.  Determination of age-related increases in large artery stiffness by digital pulse contour analysis.   Clin Sci (Lond). 2002;103(4):371-377. doi:10.1042/cs1030371PubMedGoogle ScholarCrossref
    21.
    Gunarathne  A, Patel  JV, Gammon  B, Hughes  EA, Lip  GY.  Impact of mean arterial blood pressure on higher arterial stiffness indices in South Asians compared to white Europeans.   J Hypertens. 2008;26(7):1420-1426. doi:10.1097/HJH.0b013e3282ffb42aPubMedGoogle ScholarCrossref
    22.
    Kollias  A, Stergiou  GS, Dolan  E, O’Brien  E.  Ambulatory arterial stiffness index: a systematic review and meta-analysis.   Atherosclerosis. 2012;224(2):291-301. doi:10.1016/j.atherosclerosis.2012.03.039PubMedGoogle ScholarCrossref
    23.
    Sollinger  D, Mohaupt  MG, Wilhelm  A, Uehlinger  D, Frey  FJ, Eisenberger  U.  Arterial stiffness assessed by digital volume pulse correlates with comorbidity in patients with ESRD.   Am J Kidney Dis. 2006;48(3):456-463. doi:10.1053/j.ajkd.2006.05.014PubMedGoogle ScholarCrossref
    24.
    Kroenke  K, Spitzer  RL, Williams  JB, Löwe  B.  An ultra-brief screening scale for anxiety and depression: the PHQ-4.   Psychosomatics. 2009;50(6):613-621.PubMedGoogle Scholar
    25.
    Löwe  B, Wahl  I, Rose  M,  et al.  A 4-item measure of depression and anxiety: validation and standardization of the Patient Health Questionnaire-4 (PHQ-4) in the general population.   J Affect Disord. 2010;122(1-2):86-95. doi:10.1016/j.jad.2009.06.019PubMedGoogle ScholarCrossref
    26.
    Löwe  B, Kroenke  K, Gräfe  K.  Detecting and monitoring depression with a two-item questionnaire (PHQ-2).   J Psychosom Res. 2005;58(2):163-171. doi:10.1016/j.jpsychores.2004.09.006PubMedGoogle ScholarCrossref
    27.
    Alberti  KG, Eckel  RH, Grundy  SM,  et al; International Diabetes Federation Task Force on Epidemiology and Prevention; Hational Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; International Association for the Study of Obesity.  Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity.   Circulation. 2009;120(16):1640-1645. doi:10.1161/CIRCULATIONAHA.109.192644PubMedGoogle ScholarCrossref
    28.
    van der Berg  JD, Stehouwer  CD, Bosma  H,  et al.  Associations of total amount and patterns of sedentary behaviour with type 2 diabetes and the metabolic syndrome: the Maastricht Study.   Diabetologia. 2016;59(4):709-718. doi:10.1007/s00125-015-3861-8PubMedGoogle ScholarCrossref
    29.
    White  WB, Anwar  YA.  Evaluation of the overall efficacy of the Omron office digital blood pressure HEM-907 monitor in adults.   Blood Press Monit. 2001;6(2):107-110. doi:10.1097/00126097-200104000-00007PubMedGoogle ScholarCrossref
    30.
    Ridker  PM, Buring  JE, Cook  NR, Rifai  N.  C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially healthy American women.   Circulation. 2003;107(3):391-397. doi:10.1161/01.CIR.0000055014.62083.05PubMedGoogle ScholarCrossref
    31.
    Sesso  HD, Buring  JE, Rifai  N, Blake  GJ, Gaziano  JM, Ridker  PM.  C-reactive protein and the risk of developing hypertension.   JAMA. 2003;290(22):2945-2951. doi:10.1001/jama.290.22.2945PubMedGoogle ScholarCrossref
    32.
    Gafoor  R, Booth  HP, Gulliford  MC.  Antidepressant utilisation and incidence of weight gain during 10 years’ follow-up: population based cohort study.   BMJ. 2018;361:k1951. doi:10.1136/bmj.k1951PubMedGoogle ScholarCrossref
    33.
    Stein  A, Pearson  RM, Goodman  SH,  et al.  Effects of perinatal mental disorders on the fetus and child.   Lancet. 2014;384(9956):1800-1819. doi:10.1016/S0140-6736(14)61277-0PubMedGoogle ScholarCrossref
    34.
    Dregan  A, Matcham  F, Harber-Aschan  L,  et al.  Common mental disorders within chronic inflammatory disorders: a primary care database prospective investigation.   Ann Rheum Dis. 2019;78(5):688-695. doi:10.1136/annrheumdis-2018-214676PubMedGoogle ScholarCrossref
    35.
    Pitman  A, Suleman  S, Hyde  N, Hodgkiss  A.  Depression and anxiety in patients with cancer.   BMJ. 2018;361:k1415. doi:10.1136/bmj.k1415PubMedGoogle ScholarCrossref
    36.
    Valeri  L, Vanderweele  TJ.  Mediation analysis allowing for exposure-mediator interactions and causal interpretation: theoretical assumptions and implementation with SAS and SPSS macros.   Psychol Methods. 2013;18(2):137-150. doi:10.1037/a0031034PubMedGoogle ScholarCrossref
    37.
    Steen  J, Loeys  T, Moerkerke  B, Vansteelandt  S.  Flexible mediation analysis with multiple mediators.   Am J Epidemiol. 2017;186(2):184-193. doi:10.1093/aje/kwx051PubMedGoogle ScholarCrossref
    38.
    Rybakowski  JK, Wykretowicz  A, Heymann-Szlachcinska  A, Wysocki  H.  Impairment of endothelial function in unipolar and bipolar depression.   Biol Psychiatry. 2006;60(8):889-891. doi:10.1016/j.biopsych.2006.03.025PubMedGoogle ScholarCrossref
    39.
    Després  JP, Lemieux  I, Bergeron  J,  et al.  Abdominal obesity and the metabolic syndrome: contribution to global cardiometabolic risk.   Arterioscler Thromb Vasc Biol. 2008;28(6):1039-1049. doi:10.1161/ATVBAHA.107.159228PubMedGoogle ScholarCrossref
    40.
    Ritchie  SA, Connell  JM.  The link between abdominal obesity, metabolic syndrome and cardiovascular disease.   Nutr Metab Cardiovasc Dis. 2007;17(4):319-326. doi:10.1016/j.numecd.2006.07.005PubMedGoogle ScholarCrossref
    41.
    Veith  RC, Lewis  N, Linares  OA,  et al.  Sympathetic nervous system activity in major depression: basal and desipramine-induced alterations in plasma norepinephrine kinetics.   Arch Gen Psychiatry. 1994;51(5):411-422. doi:10.1001/archpsyc.1994.03950050071008PubMedGoogle ScholarCrossref
    42.
    Spieker  LE, Hürlimann  D, Ruschitzka  F,  et al.  Mental stress induces prolonged endothelial dysfunction via endothelin-A receptors.   Circulation. 2002;105(24):2817-2820. doi:10.1161/01.CIR.0000021598.15895.34PubMedGoogle ScholarCrossref
    43.
    Wittstein  IS, Thiemann  DR, Lima  JA,  et al.  Neurohumoral features of myocardial stunning due to sudden emotional stress.   N Engl J Med. 2005;352(6):539-548. doi:10.1056/NEJMoa043046PubMedGoogle ScholarCrossref
    44.
    Mill  J, Petronis  A.  Molecular studies of major depressive disorder: the epigenetic perspective.   Mol Psychiatry. 2007;12(9):799-814. doi:10.1038/sj.mp.4001992PubMedGoogle ScholarCrossref
    45.
    Lind  L, Ingelsson  E, Sundström  J, Siegbahn  A, Lampa  E.  Methylation-based estimated biological age and cardiovascular disease.   Eur J Clin Invest. 2018;48(2). doi:10.1111/eci.12872PubMedGoogle Scholar
    46.
    Tomfohr  LM, Murphy  ML, Miller  GE, Puterman  E.  Multiwave associations between depressive symptoms and endothelial function in adolescent and young adult females.   Psychosom Med. 2011;73(6):456-461. doi:10.1097/PSY.0b013e3182228644PubMedGoogle ScholarCrossref
    47.
    Seldenrijk  A, Vogelzangs  N, van Hout  HP, van Marwijk  HW, Diamant  M, Penninx  BW.  Depressive and anxiety disorders and risk of subclinical atherosclerosis: findings from the Netherlands Study of Depression and Anxiety (NESDA).   J Psychosom Res. 2010;69(2):203-210. doi:10.1016/j.jpsychores.2010.01.005PubMedGoogle ScholarCrossref
    48.
    Greaney  JL, Koffer  RE, Saunders  EFH, Almeida  DM, Alexander  LM.  Self-reported everyday psychosocial stressors are associated with greater impairments in endothelial function in young adults with major depressive disorder.   J Am Heart Assoc. 2019;8(4):e010825. doi:10.1161/JAHA.118.010825PubMedGoogle Scholar
    49.
    Lewis  TT, Sutton-Tyrrell  K, Penninx  BW,  et al.  Race, psychosocial factors, and aortic pulse wave velocity: the Health, Aging, and Body Composition Study.   J Gerontol A Biol Sci Med Sci. 2010;65(10):1079-1085. doi:10.1093/gerona/glq089PubMedGoogle ScholarCrossref
    50.
    Tiemeier  H, Breteler  MM, van Popele  NM, Hofman  A, Witteman  JC.  Late-life depression is associated with arterial stiffness: a population-based study.   J Am Geriatr Soc. 2003;51(8):1105-1110. doi:10.1046/j.1532-5415.2003.51359.xPubMedGoogle ScholarCrossref
    51.
    Camacho  Á, McClelland  RL, Delaney  JA,  et al.  Antidepressant use and subclinical measures of atherosclerosis: the Multi-Ethnic Study of Atherosclerosis.   J Clin Psychopharmacol. 2016;36(4):340-346. doi:10.1097/JCP.0000000000000518PubMedGoogle ScholarCrossref
    52.
    Tong  B, Abosi  O, Schmitz  S, Myers  J, Pierce  GL, Fiedorowicz  JG.  Bipolar disorder and related mood states are not associated with endothelial function of small arteries in adults without heart disease.   Gen Hosp Psychiatry. 2018;51:36-40. doi:10.1016/j.genhosppsych.2017.12.005PubMedGoogle ScholarCrossref
    53.
    Park  YW, Zhu  S, Palaniappan  L, Heshka  S, Carnethon  MR, Heymsfield  SB.  The metabolic syndrome: prevalence and associated risk factor findings in the US population from the Third National Health and Nutrition Examination Survey, 1988-1994.   Arch Intern Med. 2003;163(4):427-436. doi:10.1001/archinte.163.4.427PubMedGoogle ScholarCrossref
    54.
    Bayturan  O, Tuzcu  EM, Lavoie  A,  et al.  The metabolic syndrome, its component risk factors, and progression of coronary atherosclerosis.   Arch Intern Med. 2010;170(5):478-484. doi:10.1001/archinternmed.2009.551PubMedGoogle ScholarCrossref
    55.
    Groenwold  RH, Sterne  JA, Lawlor  DA, Moons  KG, Hoes  AW, Tilling  K.  Sensitivity analysis for the effects of multiple unmeasured confounders.   Ann Epidemiol. 2016;26(9):605-611. doi:10.1016/j.annepidem.2016.07.009PubMedGoogle ScholarCrossref
    56.
    Olfson  M, Marcus  SC.  National patterns in antidepressant medication treatment.   Arch Gen Psychiatry. 2009;66(8):848-856. doi:10.1001/archgenpsychiatry.2009.81PubMedGoogle ScholarCrossref
    57.
    Mason  C, Katzmarzyk  PT.  Variability in waist circumference measurements according to anatomic measurement site.   Obesity (Silver Spring). 2009;17(9):1789-1795. doi:10.1038/oby.2009.87PubMedGoogle ScholarCrossref
    58.
    Simonetti  GD, Eisenberger  U, Bergmann  IP, Frey  FJ, Mohaupt  MG.  Pulse contour analysis: a valid assessment of central arterial stiffness in children?   Pediatr Nephrol. 2008;23(3):439-444. doi:10.1007/s00467-007-0693-xPubMedGoogle ScholarCrossref
    59.
    Xi  B, Zhang  L, Mi  J.  Reduced arterial compliance associated with metabolic syndrome in Chinese children and adolescents.   Biomed Environ Sci. 2010;23(2):102-107. doi:10.1016/S0895-3988(10)60038-4PubMedGoogle ScholarCrossref
    60.
    Broadley  AJ, Korszun  A, Jones  CJ, Frenneaux  MP.  Arterial endothelial function is impaired in treated depression.   Heart. 2002;88(5):521-523. doi:10.1136/heart.88.5.521PubMedGoogle ScholarCrossref
    61.
    Piepoli  MF, Hoes  AW, Agewall  S,  et al; ESC Scientific Document Group.  2016 European Guidelines on cardiovascular disease prevention in clinical practice: the Sixth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice (constituted by representatives of 10 societies and by invited experts): developed with the special contribution of the European Association for Cardiovascular Prevention & Rehabilitation (EACPR).   Eur Heart J. 2016;37(29):2315-2381. doi:10.1093/eurheartj/ehw106PubMedGoogle ScholarCrossref
    ×