Psychoneuroendocrinology (2009) 34, 353—357

Angiogenic factors in patients with current majordepressive disorder comorbid with borderlinepersonality disorder

Kai G. Kahl a,b,*, Susanne Bens b, Kristin Ziegler b, Sebastian Rudolf b,Andreas Kordon b, Leif Dibbelt c, Ulrich Schweiger b

ava i lab le at www.sc ienced i rect .com

journa l homepage: www.e l sev ie r.com/locate/psyneuen

aDepartment of Psychiatry, Social Psychiatry and Psychotherapy, Hannover Medical School, Carl-Neuberg-Strasse 1,30625 Hannover, GermanybDepartment of Psychiatry and Psychotherapy, University of Luebeck, Ratzeburger Allee 160, 23538 Luebeck, Germanyc Institute for Clinical Chemistry, University of Luebeck, Ratzeburger Allee 160, 23538 Luebeck, Germany

Received 3 October 2007; received in revised form 23 September 2008; accepted 24 September 2008

KEYWORDSMajor depressivedisorder;Borderline personalitydisorder;VEGF;FGF-2;Angiogenesis;Comorbidity

Summary

Background: Major depression has been associated with endocrine and immune alterations, inparticular a dysregulation of the hypothalamus—pituitary—adrenal system with subsequent hyper-cortisolism and an imbalance of pro- and anti-inflammatory cytokines. Recent studies suggest thatvascular endothelial growth factor (VEGF), a cytokine involved in angiogenesis and neurogenesis,may also be dysregulated during stress and depression. These observations prompted us to examineVEGF and other angiogenic factors in patients with major depressive disorder.Methods: Twelve medication-free female patients with a major depressive episode in the contextof borderline personality disorder (MDD/BPD) and twelve healthy women were included. Con-centrations of VEGF, VEGF receptors 1 and 2, basic fibroblast growth factor-2 (FGF-2), hepatocytegrowth factor (HGF), angiopoetin-2, interleukin-8 (IL-8) and transforming growth factor-b1 (TGF-b1) were determined from serum profiles.Results: Increased concentrations of VEGF and FGF-2 were found in MDD/BPD patients comparedto the healthy comparator group. No group differences were found concerning the otherangiogenic factors examined.Conclusion: Depressive episodes in the context of borderline personality disorder may be accom-panied by increased serum concentrations of VEGF and FGF-2. Similar findings have been observedin patients with major depression without a borderline personality disorder. A dysregulation ofangiogenic factors may be another facet of the endocrine and immunologic disturbances fre-quently seen in patients with depressive episodes.# 2008 Elsevier Ltd. All rights reserved.

* Corresponding author at: Department of Psychiatry, Social Psychiatry and Psychotherapy, Hannover Medical School, Carl-Neuberg-Strasse 1,30625 Hannover, Germany. Tel.: +49 511 532 2495; fax: +49 511 532 2415.

E-mail address: kahl.kai@mh-hannover.de (K.G. Kahl).

0306-4530/$ — see front matter # 2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.psyneuen.2008.09.016

354 K.G. Kahl et al.

1. Introduction

A dysregulation of the hypothalamic—pituitary—adrenal sys-tem (HPAS) and increased concentrations of pro-inflamma-tory cytokines (such as interleukin-6 and tumor-necrosisfactor-a) have often been described in major depressivedisorder (MDD) (reviews in Heuser, 1998; Schiepers et al.,2005; Dantzer and Kelley, 2007). Similar findings have beenobserved in patients suffering from major depression comor-bid with borderline personality disorder (Kahl et al., 2006).The above-mentioned alterations have been implicated inthe development of type 2 diabetes mellitus (T2DM) andcardiovascular disorders, which are more prevalent indepressed patients when compared to non-depressed healthysubjects (Katon, 2003; Musselman et al., 2003).

Recent reports suggest that vascular endothelial growthfactor (VEGF)may also be dysregulated during states of stressand depression. VEGF is an angiogenic cytokine able to inducevascular endothelial cell proliferation, migration and vaso-permeability in different cell types (Ferrara et al., 2003).VEGF has also been demonstrated to increase the prolifera-tion of neurons in the adult hippocampus, and has beenshown to be upregulated after electroconvulsive seizuresin rat brains (Newton et al., 2003). The neurogenic/neuro-trophic hypothesis of depression states that depressionresults at least in part from the decreased neurogenesisand/or depletion of neurotrophic factor support, and thatthis loss may eventually lead to structural abnormalities andcompromised neuronal function (Duman et al., 1997). Takentogether, this neurotrophic hypothesis of depression led tothe assumption that low concentrations of VEGF and othertrophic factors, in particular brain derived neurotrophicfactor, may be involved in depression.

In in vitro experiments, in contrast to the above men-tioned hypothesis, an increase in VEGF mRNA expression andhigher VEGF protein concentrations were found after stimu-lation with cortisol or norepinephrine in two ovarian cancercell lines and in brown adipocytes (Fredriksson et al., 2000;Lutgendorf et al., 2003). In women suffering form ovariancarcinoma an association between increased VEGF concen-trations, feelings of helplessness, and low social well-beingwas observed. Women who reported higher levels of socialwell-being had lower VEGF concentrations (Lutgendorf et al.,2002). Depression and cancer related concerns have alsobeen associated with high pre- and postoperative VEGF con-centrations in patients suffering from colorectal carcinoma(Sharma et al., 2008). Very recently, Iga and colleaguesreported higher VEGF mRNA expression in peripheral leuco-cytes from drug-naive depressed patients compared tohealthy subjects (Iga et al., 2007).

The allocation system comprises all neuroendocrine andvegetative functions that control the distribution of meta-bolic energy between the brain and the periphery (Peterset al., 2004). Increased VEGF is part of the defensiveresponse of the allocation system to hypoglycaemia (Dantzet al., 2002). This activation of the allocation system issimilar to the endocrine alterations seen in depression. Takentogether, the published data point towards a dysregulation ofVEGF secretion in depression. However, there is uncertaintyabout the direction of this alteration.

Other angiogenic factors have not yet been studied sys-tematically in depressed patients. As such, we tested our

hypothesis of altered concentrations of angiogenic factors bydetermining the concentrations of VEGF, VEGF receptors 1and 2, FGF-2, HGF, angiopoetin-2, IL-8 and TGF-b1 in serumprofiles of physically healthy, young depressed women.

Earlier studies revealed that ‘‘pure depression’’ can befound only in about 20% of the affected population (Melartinet al., 2002). Up to 70% of major depressive episodes in youngwomen occur in the context of personality disorders, andabout 30% of depressed patients may suffer from borderlinepersonality disorder (BPD; Corruble et al., 1996; Rossi et al.,2001). Patients with both BPD and Major Depression arecharacterised by high disease severity and chronicity andreport high levels of exposure to childhood and life adver-sities (Bellino et al., 2005). For this reason we chose toexamine young depressed women with BPD in order to studya groupwith high disease severity and homogeneity regardingthe comorbid axis-II diagnosis.

2. Methods

The studywas approved by the local ethics committee, and allwomen in the patient and the comparator group gave theirwritten informed consent prior to the beginning of the study.Twelve unmedicated female patients consecutively admittedto our hospital who met DSM-IV diagnostic criteria for currentmajor depressive episode (MDD) and borderline personalitydisorder (BPD) were included. All patients were drug-free atthe time of testing, and had not received antidepressant,neuroleptic or other medication during the previous 8 weeks.Exclusion criteria were anorexia nervosa, substance-relateddisorders, schizophrenia, mental retardation, pregnancy,amenorrhea, medical illness (including (auto-) inflammatorydisease and cancer), and an age of 17 years or younger.Diagnosiswasmadeusing theGermanversionof theStructuredClinical Interview for DSM-IV (SCID I and II). The comparatorgroup consisted of twelve healthy womenwho also underwentdiagnosis by SCID I and II. Subjects with a personal or familyhistory (first degree relatives) of psychiatric illness, personalhistory of substance abuse or personality disorder according toSCID II were excluded. None of the subjects in the comparatorgroup had a history of single or repeated trauma.

Further psychological examinations for the whole studygroup included the German version of the Symptom Checklist(SCL-90-R), and the German version of the Beck DepressionInventory (BDI). For data on the psychometric properties ofthe German version of these scales see Hautzinger et al.(1995) and Schmitz et al. (2000). Physical activity was deter-mined using a 6-point Likert scale (Cuppett and Latin, 2002).Smoking habits were expressed as packyears (cigarettes perday � years of smoking/20). In all study subjects, menstrua-tion was reported to be regular with intervals ranging from 24to 35 days during the previous year. None of the studysubjects received oral contraception.

The serum of each participant was collected during theearlymenstrual phase (betweenday3andday5after theonsetofmenstruation) between 15:00 h and 19:00 h at 10 min inter-vals through an intravenous catheter. Patientswere sitting in aquiet room in a semireclined positionduring the procedure andabstained from eating, drinking, smoking, watching TV orhearingmusic (Kahl et al., 2006). The serum samples obtainedfrom each participant were then stored at�408 until analysis.Vascular endothelial growth factor (VEGF), VEGF receptors 1

Table 1 Sociodemographic and psychological data.

n Age (year) BMI (kg/m2) Packyears Activity BDI GSI (SCL-90-R)

Control 12 25.6 � 3.9 21.8 � 3.2 5.1 � 4.9 4.2 � 1.3 2.6 � 2.5 0.2 � 0.2MDD/BPD 12 26.3 � 5.1 25.9 � 8.7 5.3 � 4.2 3.4 � 1.2 34.9 � 8.3 * 1.9 � 0.8 *

MDD/BPD: depressed patients with comorbid borderline personality disorder; BDI: sum score of the Beck’s Depression Inventory; BMI: bodymassindex; GSI: general severity index; SCL-90-R: Symptom Checklist-90, revised form.* p-value <0.05.

Angiogenic factors in depressed patients comorbid with borderline personality disorder 355

and2 (VEGF-R1;VEGF-R2), basicfibroblast growth factor (FGF-2), hepatocyte growth factor (HGF), angiopoetin-2, interleu-kin-8 (IL-8) and transforming growth factor-b1(TGF-b1) weredetermined at the beginning (T0), after 120 min (T120; 120 minafter the first sample was obtained), and at the end of serumsampling (T240; 240 min after the first sample was obtained).We chose the interval of 120 min because we did not expectrapid diurnal changes in the concentrations of the measuredcytokines, and we also wanted to increase the validity of theexperiment. Angiogenic factors (VEGF, VEGF-R1 and 2, HGF,FGF-2, IL-8, angiopoetin-2, TGF-b1) were determined usingcommercially available ELISA kits according to the manufac-turers instructions (all from R&D Systems; Germany). Theintra- and inter-assay coefficients of variation for all assaysused were less than 7%.

Data were analysed using SPSS (version 15.0). Groupswere compared by t-test (two-sided) and multivariate ana-lysis (MANOVA). Further analysis was performed usingrepeated measures ANOVA. Pearson’s coefficients of correla-tion were calculated. A p-value below 0.05 was consideredto be significant. All values are given as means � S.D.

3. Results

Patients were comparable for age ( p = .72) and smokinghabits ( p = .89) although they did tend to be heavier( p = .14) when compared to the comparator group

Table 2 Repeated measures ANOVA revealed increased concentrfibroblast growth factor-2 (FGF-2) in serum profiles of depressed pat

VEGF(pg/ml) (T0)

VEGF(pg/ml) (T120)

VEGF(pg/ml) (T2

Control 283 � 154 267 � 148 254 � 146MDD/BPD 567 � 389* 502 � 332 * 475 � 342

TGF-b1(ng/ml) (T0)

TGF-b1(ng/ml) (T120)

TGF-b1(ng/ml) (T24

Control 39.7 � 3.8 35.6 � 3.3 35.8 � 2.5MDD/BPD 41 � 7.3 39.6 � 6.7 38.1 � 6.5

HGF(pg/ml) (T0)

HGF(pg/ml) (T120)

HGF(pg/ml) (T2

Control 868 � 185 832 � 185 795 � 155MDD/BPD 1093 � 350 969 � 311 914 � 342

VEGF-R1(pg/ml) (T0)

VEGF-R1(pg/ml) (T120)

VEGF-R1(pg/ml) (T2

Control 36.4 � 6.9 39.1 � 8.9 38.5 � 7.9MDD/BPD 36 � 10.9 37 � 11.5 40.4 � 16.8* Indicates significant differences between the groups obtained by pai

(Table 1). Repeated measures ANOVA using the covariatebody mass index (BMI) was therefore performed for groupcomparisons. Multivariate analysis showed higher concentra-tions of VEGF (F = 6.3; d.f. = 2,20; p = 0.01) in MDD/BPD thanin the healthy subjects (Table 2). There was a significanteffect of time (F = 5.5; d.f. = 2,42; p < 0.01) and a group bytime interaction (F = 4.3; d.f. = 2,42; p < 0.05) on the VEGFconcentrations. A multivariate analysis with the mean con-centrations of the measured angiogenic factors and with BMIas the covariate revealed a significant effect of group(F = 3.3; d.f. = 4; p = 0.03), as well as significant group dif-ferences for VEGF (F = 4.9; d.f. = 1; p = 0.04). For FGF noeffect of time was seen, although a group by time interaction(F = 4.4 d.f. = 2,42; p < 0.05 after Greenhouse—Geisser cor-rection) could be identified. A multivariate analysis with themean concentrations of the measured angiogenic factors andwith BMI as the covariate revealed a significant effect ofgroup (F = 3.3; d.f. = 4; p = 0.03) as well as significant groupdifferences for FGF-2 (F = 5.1; d.f. = 1; p = 0.035). No groupdifferences were found for any of the other angiogenicfactors measured (Table 2). Mauchly’s test supported theassumption of sphericity for VEGF but not for FGF (Mauchly-W = 0.468; p < 0.01), indicating a normal distribution for theVEGF but not the FGF-2 concentrations.

BDI and GSI were both significantly higher in patients thanin controls ( p < 0.001) (Table 1). Correlations were foundfor the BDI sum score and VEGF at the three time-points

ations of vascular endothelial growth factor (VEGF) and basicients comorbid with borderline personality disorder (MDD/BPD).

40)FGF-2(pg/ml) (T0)

FGF-2(pg/ml) (T120)

FGF-2(pg/ml) (T240)

7.1 � 4.6 7.0 � 6.3 5.1 � 6.321.5 � 21.6 * 12.4 � 14.0 9.1 � 10.1

0)Angiopoetin-2(ng/ml) (T0)

Angiopoetin-2(ng/ml) (T120)

Angiopoetin-2(ng/ml) (T240)

2.6 � 1 2.3 � 0.9 2.3 � 12.2 � 1.7 2.1 � 1.7 2.1 � 1.3

40)IL-8(pg/ml) (T0)

IL-8(pg/ml) (T120)

IL-8(pg/ml) (T240)

8.2 � 2.6 8.3 � 3.6 7.8 � 2.812.6 � 1.6 10.7 � 1.8 7.2 � 4.2

40)VEGF-R2(ng/ml) (T0)

VEGF-R2(ng/ml) (T120)

VEGF-R2(ng/ml) (T240)

8.5 � 1.1 8.2 � 1.1 8.1 � 0.98.5 � 2.3 8.2 � 2 7.9 � 1.9

red t-test (p < .05).

356 K.G. Kahl et al.

(T0: r = 0.49; p = 0.01; T120: r = 0.51; p = 0.01; T240: r = 0.45;p = 0.03). No other correlations were found, and in particularno correlations were observed between age or BMI and theexamined endocrine and immunological parameters (datanot shown).

4. Discussion

We found elevated concentrations of VEGF and FGF-2 butsimilar concentrations of angiopoetin-2, IL-8, TGF-b1 andHGF in young depressed patients with borderline personalitydisorder when compared to a healthy comparator group. Thedemonstration of increased VEGF protein concentrations issimilar to the results of others who found an increase in VEGFmRNA expression in peripheral leucocytes of patients suffer-ing from depression without a comorbid personality disorder(Iga et al., 2007). However, protein concentrations of VEGF orother angiogenic factors were not determined in the studycited above. In contrast to our expectations (which werebased on the findings of Dantz et al. during a euglycemiccondition), VEGF decreased slightly but significantly overtime (Dantz et al., 2002). Whether this was related tocircadian variation or to the effects of resting or bloodsampling remains unclear (Hettland et al., 2008).

Interestingly, FGF-2 concentrations were initially high inthe patient group before they dropped off again, while nosuch difference was found in the comparator group. Psycho-logical stress and glucocorticoids have been shown to upre-gulate FGF-2, and very recently glucocorticoids were showntomediate the effects of stress on FGF-2 concentrations in ananimal model (Frank et al., 2007). In our patients, cortisolconcentrations were also increased, as was also reportedelsewhere (Kahl et al., 2006). However, whether theobserved group differences in the variation of FGF-2 con-centrations over time were due to an exaggerated responseto stress in our patients could not be assessed from our data.

VEGF and FGF-2 have been shown to play key roles in bothneurogenesis and angiogenesis, with both having also beenimplicated in major depressive disorder (Newton and Duman,2004; Greenberg and Jin, 2005). The findings of the increasedVEGF and FGF-2 are at odds with the neurotrophic theory ofdepression. Pathological angiogenesis is observed in a varietyof diseases including cancer and complications secondary todiabetes mellitus (Paques et al., 1997; Folkman, 2001; Grantet al., 2004; Presta et al., 2005). Progression of cancer andcomplications of diabetes mellitus have been suggested tooccur more frequently in patients with major depression (deGroot et al., 2001; Spiegel and Giese-Davis, 2003), althoughno patient in our study group was afflicted by one of thesedisorders. Elevated VEGF and FGF-2 concentrations havebeen described as possible factors contributing towards aproliferative diabetic retinopathy. Diabetic retinopathy isfound at a higher frequency amongst patients with T2DMand depression compared to non-depressed patients withT2DM (de Groot et al., 2001; Grant et al., 2004). Whenconsidering these results together with our findings it hasto be remembered that the question of cause and effect inthe interaction between depression, VEGF and diabetic reti-nopathy cannot be addressed by looking at cross-sectionaldata. However it does seem unlikely that diabetic retino-pathy should influence VEGF concentrations.

Angiogenic factors are produced by a variety of cells,including macrophages, T-cells, smooth muscle cells, kera-tinocytes, kidney cells, tumour cells and adipose-tissuederived stem cells (Berse et al., 1992; Freeman et al.,1995; Kilroy et al., 2007). However, the source of increasedVEGF and FGF-2 in our depressed patients remains unclearand was not the focus of our study.

A physical and laboratory examination revealed no evi-dence for physical disease in either study group. Furthermore,smoking habits and physical activity were similar between thegroups. As such, the observed differences in VEGF and FGF-2serum concentrations are unlikely to have been due to differ-ences in behaviour or any underlying physical disease. Inter-estingly, we did find a correlation between the BDI sum scoreandVEGF indicating that the severity of depressionmayalsobeassociated with the observed alterations.

One limitation of our study was the small number ofpatients. According to the pilot character of the study thepower to detect the observed group difference was only 61%for VEGFand only 46% for FGF-2 (a = 0.05;b = 0.20). This limitsthe interpretation of negative results in our study in particular.Another is that BMIwas slightly althoughnot significantly lowerin the group of healthy women, so that the observed differ-ences may therefore have been accounted for by this differ-ence.However, statistical analysiswasperformedwithBMIasacovariate, indicating that theobservedgroupdifferenceswerenotdependent ondifferences in theBMI. Our study includednocomparator group with current MDD and without BPD. Thismakes it difficult todifferentiatewhether theobservedaltera-tions are aneffect of depressionalone, orwhether theyare theresult of an interaction between depression and borderlinepersonality disorder. Earlier observations showed that BPD inthe absence of comorbidity is not associated with majoralterations of HPAS or the cytokine system (Kahl et al.,2005; Jogems-Kosterman et al., 2007).

In summary, we found increased concentrations of VEGFand basic FGF-2 in patients comorbid for MDD and BPD. Theobserved alterations in angiogenic factors may be anotherfacet of the endocrine and immunological disturbancesalready reported in patients with depressive disorders.

Role of funding source

The study was supported by a grant of the Universitaet zuLuebeck (FUL13/03).

Conflict of interest

None of the authors reported financial interests or potentialconflicts of interest.

Acknowledgement

We thank Ms. J. Marple who assisted with the preparation andthe proof reading of the manuscript.

References

Bellino, S., Patria, L., Paradiso, E., Di Lorenzo, R., Zanon, C., Zizza,M., Bogetto, F., 2005. Major depression in patients with border-

Angiogenic factors in depressed patients comorbid with borderline personality disorder 357

line personality disorder: a clinical investigation. Can. J. Psy-chiatry 50, 234—238.

Berse, B., Brown, L.F., Van de Water, L., Dvorak, H.F., Senger, D.R.,1992. Vascular permeability factor (vascular endothelial growthfactor) gene is expressed differentially in normal tissues, macro-phages, and tumors. Mol. Biol. Cell 3, 211—220.

Corruble, E., Ginestet, D., Guelfi, J.D., 1996. Comorbidity of per-sonality disorders and unipolar major depression: a review. J.Affect. Disord. 37, 157—170.

Cuppett, M., Latin, R.W., 2002. A survey of physical activity levels ofcertified athletic trainers. J. Athl. Train. 37, 281—285.

Dantz, D., Bewersdorf, J., Fruehwald-Schultes, B., Kern, W., Jelk-mann, W., Born, J., Fehm, H.L., Peters, A., 2002. Vascularendothelial growth factor: an novel endocrine defensive responseto hypoglycemia. J. Clin. Endocrinol. Metab. 87, 835—840.

Dantzer, R., Kelley, K.W., 2007. Twenty years of research on cytokine-induced sickness behavior. Brain Behav. Immun. 21, 153—160.

de Groot, M., Anderson, R., Freedland, K.E., Clouse, R.E., Lustman,P.J., 2001. Association of depression and diabetes complications:a meta-analysis. Psychosom. Med. 63, 619—630.

Duman, R.S., Heninger, G.R., Nestler, E.J., 1997. A molecular andcellular theory of depression. Arch. Gen. Psychiatry 54, 597—606.

Ferrara, N., Gerber, H.P., LeCouter, J., 2003. The biology of VEGF andits receptors. Nat. Med. 9, 669—676.

Folkman, J., 2001. Angiogenesis-dependent diseases. Semin. Oncol.28, 24—31.

Frank, M.G., Der-Avakian, A., Bland, S.T., Watkins, L.R., Maier, S.F.,2007. Stress-induced glucocorticoids suppress the antisensemolecular regulation of FGF-2 expression. Psychoneuroendocri-nology 32, 376—384.

Fredriksson, J.M., Lindquist, J.M., Bronnikov, G., Nedergaard, J.,2000. Norepinephrine induces vascular endothelial growth factorgene expression in brown adipocytes through a beta-adrenocep-tor cAMP/protein kinase A pathway involving Src but indepen-dently of Erk1/2. J. Biol. Chem. 275, 13802—13811.

Freeman, M.R., Schneck, F.X., Gagnon, M.L., Corless, C., Soker, S.,Niknejad, K., Peoples, G.E., Klagsbrun, M., 1995. Peripheralblood T lymphocytes infiltrating human cancers express vascularendothelial growth factor: a potential role for Tcells in angiogen-esis. Cancer Res. 55, 4140—4145.

Grant, M.B., Afzal, A., Spoerri, P., Pan, H., Shaw, L.C., Mames, R.N.,2004. The role of growth factors in the pathogenesis of diabeticretinopathy. Expert. Opin. Invest. Drugs 13, 1275—1293.

Greenberg, D.A., Jin, K., 2005. From angiogenesis to neuropathology.Nature 438, 954—959.

Hautzinger, M., Bailer, M., Worall, H., Keller, F., 1995. Beck-Depres-sions-Inventar (BDI). Testhandbuch (2. Auflage) Hans-Huber-Ver-lag, Bern.

Hettland, M.L., Christensen, I.J., Lottenburger, T., Johansen, J.S.,Svendsen, M.N., Hørslev-Petersen, K., Nielsen, L., Jørgen Niel-sen, H., 2008. Circulating VEGF as a biological marker in patientswith rheumatoid arthritis? Preanalytical and biological variabilityin healthy persons and in patients. Dis. Markers 24, 1—10.

Heuser, I., 1998. The hypothalamic—pituitary—adrenal system indepression. Pharmacopsychiatry 31, 10—13.

Iga, J., Ueno, S., Yamauchi, K., Numata, S., Tayoshi-Shibuya, S.,Kinouchi, S., Nakataki, M., Song, H., Hokoishi, K., Tanabe, H.,Sano, A., Ohmori, T., 2007. Gene expression and associationanalysis of vascular endothelial growth factor in major depres-sive disorder. Prog. Neuropsychopharmacol. Biol. Psychiatry 31,658—663.

Jogems-Kosterman, B.J., de Knijff, D.W., Kusters, R., van Hoof, J.J.,2007. Basal cortisol and DHEA levels in women with borderlinepersonality disorder. J. Psychiatr. Res. 41, 1019—1026.

Kahl, K.G., Rudolf, S., Stoeckelhuber, B.M., Dibbelt, L., Gehl, H.B.,Markhof, K., Hohagen, F., Schweiger, U., 2005. Bone mineral

density, markers of bone turnover, and cytokines in youngwomen with borderline personality disorder with and with-out comorbid major depressive disorder. Am. J. Psychiatry 162,168—174.

Kahl, K.G., Bens, S., Ziegler, K., Rudolf, S., Dibbelt, L., Kordon, A.,Schweiger, U., 2006. Cortisol, the cortisol-dehydroepiandroster-one ratio, and pro-inflammatory cytokines in patients with cur-rent major depressive disorder comorbid with borderlinepersonality disorder. Biol. Psychiatry 59, 667—671.

Katon, W.J., 2003. Clinical and health services relationships betweenmajor depression, depressive symptoms, and general medicalillness. Biol. Psychiatry 54, 216—226.

Kilroy, G.E., Foster, S.J., Wu, X., Ruiz, J., Sherwood, S., Heifetz, A.,Ludlow, J.W., Stricker, D.M., Potiny, S., Green, P., Halvorsen,Y.D., Cheatham, B., Storms, R.W., Gimble, J.M., 2007. Cytokineprofile of human adipose-derived stem cells: expression of angio-genic, hematopoetic, and pro-inflammatory cytokines. J. Cell.Physiol. 212, 702—709.

Lutgendorf, S.K., Johnson, E.L., Cooper, B., Anderson, B., Sorosky,J.I., Buller, R.E., Sood, A.K., 2002. Vascular endothelial growthfactor and social support in patients with ovarian carcinoma.Cancer 95, 808—815.

Lutgendorf, S.K., Cole, S., Costanzo, E., Bradley, S., Coffin, J.,Jabbari, S., Rainwater, K., Ritchie, J.M., Yang, M., Sood, A.K.,2003. Stress-related mediators stimulate vascular endothelialgrowth factor secretion by two ovarian cancer cell lines. Clin.Cancer Res. 9, 4514—4521.

Melartin, T.K., Rytsala, H.J., Leskela, U.S., Lestela-Mielonen, P.S.,Sokero, T.P., Isometsa, E.T., 2002. Current comorbidity of psy-chiatric disorders among DSM-IV major depressive disorderpatients in psychiatric care in the Vantaa Depression Study. J.Clin. Psychiatry 63, 126—134.

Musselman, D.L., Betan, E., Larsen, H., Phillips, L.S., 2003. Relation-ship of depression to diabetes types 1 and 2: epidemiology,biology, and treatment. Biol. Psychiatry 54, 317—329.

Newton, S.S., Collier, E.F., Hunsberger, J., Adams, D., Terwilliger, R.,Selvanayagam, E., Duman, R.S., 2003. Gene profile of electro-convulsive seizures: induction of neurotrophic and angiogenicfactors. J. Neurosci. 23, 10841—10851.

Newton, S.S., Duman, R.S., 2004. Regulation of neurogenesis andangiogenesis in depression. Curr. Neurovasc. Res. 1, 261—267.

Paques, M., Massin, P., Gaudric, A., 1997. Growth factors and dia-betic retinopathy. Diabetes Metab. 23, 125—130.

Peters, A., Schweiger, U., Pellerin, L., Hubold, C., Oltmanns, K.M.,Conrad, M., Schultes, B., Born, J., Fehm, H.L., 2004. The selfishbrain: competition for energy resources. Neurosci. Biobehav. Rev.28, 143—180.

Presta, M., Dell’Era, P., Mitola, S., Moroni, E., Ronca, R., Rusnati, M.,2005. Fibroblast growth factor/fibroblast growth factor receptorsystem in angiogenesis. Cytokine Growth Factor Rev. 16, 159—178.

Rossi, A., Marinangeli, M.G., Butti, G., Scinto, A., Di Cicco, L.,Kalyvoka, A., Petruzzi, C., 2001. Personality disorders in bipolarand depressive illness. J. Affect. Disord. 65, 3—8.

Schiepers, O.J., Wichers, M.C., Maes, M., 2005. Cytokines and majordepression. Prog. Neuropsychopharmacol. Biol. Psychiatry 29,201—217.

Schmitz, N., Hartkamp, N., Kiuse, J., Franke, G.H., Reister, G., Tress,W., 2000. The Symptom Check-List-90-R (SCL-90-R): a Germanvalidation study. Qual. Life Res. 9, 185—193.

Sharma, A., Greenman, J., Sharp, D.M.,Walker, L.G., Monson, J.R.T.,2008. Vascular endothelial growth factor and psychosocial factorsin colorectal carcinoma. Psychooncology 17, 66—73.

Spiegel, D., Giese-Davis, J., 2003. Depression and cancer:mechanisms and disease progression. Biol. Psychiatry 54,269—328.