- Email: [email protected]
Copeptin during rest and exercise in major depression
Journal of Affective Disorders 151 (2013) 284–290
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
Copeptin during rest and exercise in major depression Jesper Krogh a,c,n, Jens Peter Gøtze b, Martin Balslev Jørgensen a, Lars Østergaard Kristensen c, Caroline Kistorp c, Merete Nordentoft d a
Mental Health Centre Copenhagen, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark Department of Clinical Biochemistry, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark c Department of Endocrinology, Herlev University Hospital, University of Copenhagen, Copenhagen, Denmark d Mental Health Centre Copenhagen, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark b
art ic l e i nf o
a b s t r a c t
Article history: Received 19 April 2013 Received in revised form 10 June 2013 Accepted 10 June 2013 Available online 13 July 2013
Background: High vasopressin levels and a correlation between vasopressin and cortisol has been observed in patients with depression. The aim was to assess copeptin, the c-terminal of provasopressin, and the association between cortisol, adrenocorticotropic hormone (ACTH) and copeptin in patients with depression. Secondly, to examine the copeptin response to acute exercise and aerobic training. Methods: Copeptin, ACTH, and cortisol were measured in 111 patients with depression and 57 controls at rest. Copeptin was also measured during exercise. The depressed patients were subsequently randomized to an aerobic training intervention or an exercise control intervention. Results: The plasma level of copeptin in depressed subjects was 5.14 pg/ml (IQR 3.4–8.4) and 4.82 pg/ml (IQR 2.8–7.5) in healthy controls (p¼ .66). The association between copeptin and cortisol was.02 (95% CI –.44 to.48; p¼ .93) and the association between copeptin and ACTH was –.06 (95% CI –.17 to.05; p¼ .27). All associations were independent of depression status (p¼ .15). Aerobic exercise training did not influence copeptin levels at rest (p ¼.09) or the response to acute exercise (p¼ .574). Copeptin decreased at rest in response to aerobic training in participants with high compliance to the exercise intervention (p¼ .04). Limitations: We did not measure plasma osmolality, which is a possible confounder in this study. Conclusions: Copeptin levels are not elevated or associated to ACTH or cortisol in depressed patients. Aerobic exercise training decreased copeptin levels in high attenders only. This study does not support a role of copeptin or vasopressin in depression. & 2013 Elsevier B.V. All rights reserved.
Keywords: Copeptin Vasopressin Depression Exercise test Exercise training ACTH
1. Introduction It is widely recognized that the observed hypercortisolism in a subgroup of patients with depression is due to centrally mediated hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis. This hypothesis has been strengthened by the finding of raised corticotrophin releasing hormone (CRH) concentrations in cerebrospinal fluid (Scott and Dinan, 1998). CRH is secreted from the paraventricular nucleus (PVN) in hypothalamic region and acts on the corticotrope cells of the pituitary to release adrenocorticotropic hormone (ACTH). Although CRH is considered the major ACTH secretagogue, the release of ACTH occurs in concert with arginine vasopressin (AVP) also released from the PVN. Apart from mediating ACTH release, AVP induces water retention by the kidney, being the main hormone in the regulation of osmotic homeostasis. n Correspondence to: Mental Health Centre Copenhagen, Bispebjerg Bakke 23, DK-2400, Denmark. Tel. : +45 2553 5485; fax: +45 3531 3953. E-mail address: [email protected] (J. Krogh).
0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.06.007
Both chronic and acute stress appears to induce increased CRH expression in the PVN (Liu et al., 2011; Wang et al., 2010; Zhang et al., 2012; Zhao and Ai, 2011). At the pituitary level chronic stress is associated with down regulation of CRH receptors (De Goeij et al., 1992a; De Goeij et al., 1992b; Rabadan-Diehl et al., 1995) while the expression of vasopressin (V1b) receptors increases. Interestingly, the AVP induced ACTH release has been found to be independent of the glucocorticoid feedback (Makino et al., 1995). A gradual shift towards AVP regulated HPA axis in response to chronic stress at the pituitary level, and glucocorticoid mediated suppression of CRH stress response, could suggest an interesting role for vasopressin in affective disorders. Elevated plasma AVP levels in patients with unipolar depression have been reported in one larger study (van Londen et al., 1997) but not in others (Gjerris et al., 1985; Inder et al., 1997). However, a recent study found a strong positive correlation between cortisol and AVP in patients without antidepressant treatment which was interpreted as an indication of high vasopressinergic activity in this patient group (Goekoop et al., 2011). In support of this, a number
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
of studies have found AVP-cortisol correlations in patients with anxious-retarded depression (Goekoop et al., 2011) as well as in suicide victims (Inder et al., 1997). Concern exist about the methodological reliability of plasma AVP measurements because AVP is known to be unstable, attached to platelets and a plasma half-life of approximately 10 min (Molnar et al., 2007; Morgenthaler et al., 2006). AVP is derived from a larger precursor peptide named provasopressin. Copeptin is a 39-amino acid glycopeptide constituting the C-terminal part of provasopressin. Because copeptin and AVP are released in equimolar ratio, copeptin levels potentially reflect the levels of AVP, which has been confirmed in intensive care unit patients (Jochberger et al., 2006). Copeptin has longer plasma half-life compared to vasopressin (∼24 min), and has been found stable in both serum and heparin plasma at room temperature for at least a week (Morgenthaler et al., 2006). Furthermore, due to the small size of AVP it cannot be measured by sandwich immunoassays but needs competitive assays, which generally are less sensitive. We hypothesized that plasma copeptin would be higher in depressed patients at rest, and that we would observe an augmented copeptin response to HPA-axis stimulation during an acute exercise test in depressed patients compared to healthy controls. We also hypothesized that the association between cortisol and copeptin and ACTH and copeptin would be higher in patients with depression as an indication of increased vasopressinergic drive in this patient group. Secondly, we expected that a three months aerobic exercise program would provide lower plasma copeptin levels at rest and during acute exercise, and an attenuated association between cortisol and copeptin and ACTH and copeptin.
2. Methods 2.1. Study population All 115 participants from a previous trial (the DEMO-II trial), a randomized clinical trial evaluating the antidepressant effect of aerobic exercise (Krogh et al., 2012) were eligible for the current study. Eligible participants were men and women between 18 and 60 years of age, referred from a clinical setting by a physician or a psychologist, and a diagnose of major depression (DSM-IV) based on the Danish version of the Mini International Neuropsychiatric Interview (MINI) (Bech et al., 1999). The participants scored above 12 on the Hamilton Depression Rating Scale (HAM-D17) (Hamilton, 1960) and were able to comprehend the informed consent statement. Exclusion criteria were current drug abuse, any antidepressant medication within the last two months, current psychotherapeutic treatment, contraindications to physical exercise, regular recreational exercise over 1 h per week, suicidal behavior according to the HAM-D17 (item 342), pregnancy, or current/previous psychotic or manic symptoms. Four participants from the original trial did not supply blood samples for the current study, thus 111 participants constitute the group of depressed patients for the current study. Healthy controls were recruited through the media and group matched to the patients with regard to sex, age, and body mass index (n¼ 57). The healthy controls were free of any current or previous psychiatric diseases assessed by the MINI (Bech et al., 1999). Evaluation of patients and healthy controls commenced in parallel. Fig. 1 shows patient flow. 2.2. Physical examination For the physical examination, the participants were requested to meet at the research department between 8:00 and 10:00 a.m. The participants were instructed not to take any food or liquids except for water beginning from midnight prior to the examination and abstain from strenuous physical activity prior to the
285
examination. Weight was measured using an electronic weight (Sohnle Medicals, Type 7700, Backnang, Germany). Blood pressure was measured after five minutes rest with the participant in a sitting position using a certified digital blood pressure monitor (Omron M6, Omron Healthcare co. LTD, Kyoto, Japan). The average of three measurements on the right arm is reported. VO2max was estimated using a bicycle cardiopulmonary exercise test (Ergomedic 839 e, Monark, Vansbro, Sweden) based on a standardized exercise protocol (Andersen, 1995). An indwelling venous catheter was inserted in an antecubital vein and blood samples collected after 5 min rest in a sitting position. Samples were collected 10–15 min prior to the exercise test, at VO2max, and 15, 30, and 60 min after the test. The samples were immediately centrifuged at room temperature and stored at −80 1C till analysis. 2.3. Hormone analyses Copeptin was analyzed for all data points and ACTH and cortisol only during rest. Copeptin was measured by the US Kryptor assay as previously described (Terzic et al., 2012). The intra-assay CVmax was estimated to be 18% and 6.3% for plasma copeptin concentrations of 3.0 pmol/l and 7.9 pmol/l, respectively. ACTH was analyzed using an IRMA from BRAHMS (Klose et al., 2002). The intra-assay coefficient of variation was estimated to be 7.9% at the concentration of interest. Cortisol was analyzed using an automated competitive electrochemiluminiscens-immunoassay (ECLIA) from Roche (Modular E) with a CVmax of 8% in the relevant range of analysis. 2.4. Exercise intervention After baseline assessment the depressed patients were randomised to either an aerobic exercise intervention or an exercise control intervention as described in detail elsewhere (Krogh et al., 2012). In short, both groups were scheduled to participate in 45 min supervised sessions three times per week in a 3 months program. In the aerobic training group the participants exercised on stationary bikes at approximately 80% of their maximal heart rate, while the training in the exercise control group was predominantly stretching exercises and low impact exercise such as throwing and catching balls. 2.5. Approvals The protocol was approved by the local ethics committee (H-A2008-046) and the Danish Data Protection Agency (J.nr.2008-412354). Verbal and written informed consent was obtained from all participants involved in the trial. 2.6. Statistics Differences between depressed patients and healthy controls were evaluated using chi-square test and independent-samples t-test. The association between copeptin, ACTH, and cortisol was analyzed using linear regression analysis. The effect of disease status (depressed versus healthy) was examined using an interaction term. In all cortisol analysis we excluded participants using contraceptive pills or hormone replacement therapy. Except where specified, all analyses of copeptin, ACTH, and cortisol were performed after natural log (ln) transformation. The effect of acute exercise was estimated by analyzing copeptin values at VO2max in a linear regression model with depression status as the main dependent variable and sex, age, maximal oxygen uptake, duration of test, systolic blood pressure, and baseline copeptin as independent variables. We used stepwise backwards regression analysis and kept all independent variables with p4 0.1 in the model to avoid overfitting. Secondly, we used a
286
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
Healthy subjects
Depressed subjects
Recruited from mass media:
Recruited from the DEMO-II trial:
N = 57
N = 115
Not available for biomarker assessment du too technical problems
Not available for biomarker assessment du too technical problems
N=4
N=0
Assessment
Assessment
Plasma measurement of copeptin, ACTH, and cortisol N = 111
Randomization
PART 1 Depressed vs. healthy controls
Plasma measurement of copeptin, ACTH, and cortisol N = 57
Part 2. Effect of an exercise intervention in depressed patients
Aerobic exercise training
Exercise control
N = 53
N = 58
Follow-up
Follow-up
Available for follow-up: n = 39 Lost to follow-up: n = 14
Available for follow-up: n = 44 Lost to follow-up: n = 14
Included in analysis: n = 53
Included in analysis: n = 58
Part 1 refers to the assessement of copeptin, ACTH, and cortisol in depressed patients and healthy controls. Part 2 refers to the effect of an exercise intervention on copeptin, ACTH, and cortisol in depressed patients. Fig. 1. Flowchart of participant flow.
The 111 patients and 57 healthy controls were comparable with respect to baseline demographic and clinical characteristics as shown in Table 1.
healthy controls (Mann-Whitney U; p ¼.66). There was no difference between plasma levels of ACTH and cortisol comparing depressed subjects with healthy controls (p 4.40) as shown in Table 1. The univariate association between ACTH and copeptin was –.06 (95% CI –.17 to.05; p ¼.27) and between cortisol and copeptin.02 (95% CI –.44 to.48; p ¼.93) suggesting no association between cortisol and copeptin or ACTH and copeptin in the whole population. The interaction term disease status copeptin was not significant (p ¼.29) indicating that the association between cortisol and copeptin was not different in depressed patients compared to healthy controls as also illustrated in Fig. 2. Neither did the data suggest that the association between ACTH and copeptin was different in depressed patients compared to healthy controls (disease status copeptin level; p ¼.43). The association between cortisol and ACTH was.47 (95% CI.17 to.33; po.001) corresponding to a Pearson's correlation of.40 (po.001), irrespective of disease status (disease status ACTH; p¼.82). Introducing generalized anxiety as an independent variable the association between cortisol and copeptin was not affected (p ¼.77). Copeptin levels at baseline did not predict remission in the patient group at three months follow-up.
3.2. Copeptin during rest
3.3. Copeptin during exercise
The median plasma level of copeptin in depressed subjects was 5.2 pg/ml (IQR 3.4–8.3) and 4.82 pg/ml (IQR 2.8–7.5) in
As illustrated in Table 2 maximal oxygen uptake was 3.88 ml/kg/ min (95% CI 1.7–6.1; p¼.001) lower in depressed patients. The
repeated measurement mixed model with unstructured variance including all five measurements points. In a similar way of analyses, we compared baseline differences and the effect of the randomized exercise intervention in the depressed patients. For missing data, post-intervention on copeptin, ACTH, and cortisol at rest and during VO2max we used multiple imputation applying linear regression using 100 imputations and a maximum of 20 iterations. Group allocation, baseline resting levels, and baseline copeptin levels at VO2max was used as predictor variables. SPSS version 19.0 was used for all analysis. All statistical tests were two-sided and p-values below 0.05 were considered significant.
3. Results 3.1. Participants
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
duration of the test was on average 98.9 s (95% CI 43–154; p¼.001) shorter in depressed patients and the mean maximal heart rate was 6.0 (95% CI –.3 to 12.3; p¼.06) lower compared to the healthy controls. As showed in Fig. 3 there was a significant main effect of an incremental bicycle test on copeptin levels (time: F4, 154 ¼ 70.1; p o.001). However, variations in copeptin response to acute exercise were not affected by depression (disease status time: F4,154 ¼1.7; p ¼.15). The highest level of plasma copeptin was observed at VO2max and the median increase from baseline was
Table 1 Clinical and demographical characteristics. Major depression N ¼111
Healthy controls N ¼ 57
P value
Age, years (SD) Female, n (%)
41.9 (11.2) 75 (68.2%)
40.6 (13.1) 37 (64.9%)
0.50 0.73
Psychometrics HAM-D17 (SD) HAM-A14 (SD) Generalized anxiety, n (%) Recurrent depression, n (%) Age at first depression (SD)
18.9 (3.8) 17.6 (5.5) 65 (59.1%) 58 (52.7%) 36.4 (12.1)
1.1 (1.5) 1.2 (1.7) N/A N/A N/A
– – – – –
66.2 (12.3) 122.8 (17.0)
67.6 (12.3) 119.0 (13.3)
0.51 0.16
82.1 (11.8)
79.3 (9.6)
0.14
77.3 (20.4) 26.4 (6.0)
76.8 (18.4) 25.9 (6.3)
0.86 0.66
25.4 (7.0)
29.3 (6.0)
0.001
666.3 (170.9) 163.3 (17.3)
765.2 (162.0) 169.3 (22.2)
0.001 0.06
178.6 (7.8)
179.6 (9.2)
0.50
0.91 (0.09)
0.94 (0.11)
0.09
Somatic Creatinine, mol/L (SD) Systolic blood pressure, mmHg (SD) Diastolic blood pressure, mmHg (SD) Weight, kg (SD) Body mass index (SD) Exercise test Maximal oxygen uptake, ml/kg/min (SD) Exercise time, seconds (SD) Maximum pulse, observed (SD) Maximum pulse, expected (SD) HRmax/HRmax expected (SD) 1
P value is based on log transformed values. All participants using hormone replacement therapy or oral contraceptives were excluded from cortisol analysis. Abbreviations: HAM-D17—Hamilton Depression Scale, 17 items; HAM-A14—Hamilton Anxiety Scale, 14 items; ACTH—adrenocorticotrop hormone; HRmax—highest heart rate observed during test; HRmax expected—expected maximal heart rate was derived from sex and age.
287
3.8 pg/ml (IQR.4–13.9). Using stepwise backwards regression, the final model suggested that copeptin levels at VO2max were explained by duration of test (β ¼.001; 95% CI .000–.002; p ¼.04) and baseline copeptin levels (β¼.672; 95% CI .539–.810 p o.001) and not by depression. This model explained 40.4% (R2) of the variation in copeptin levels at VO2max. 3.4. Effects of exercise training By randomization the 111 patients were allocated to either aerobic exercise training (n ¼53) or an exercise control (n¼ 58). The two groups did not differ significantly with respect to any of the biomarkers or other variables shown in Table 1 (all p values 4.10). At three months follow-up, copeptin assessment was available from 39/53 in the aerobic training group and 44/58 in the exercise control group. Comparing participants who had available copeptin levels at follow-up with participants who were missing we found no differences regarding sex, age, depression severity, or group allocation (all p values 4.2). The median participation in training sessions was 14.0 sessions (IQR 5–22) in the aerobic exercise group and 10.5 session (IQR 4–21) in the exercise control group (p ¼.41). Post intervention, participants in the aerobic exercise group had a 5.2 ml/kg/min (95% CI 1.8–8.7; p¼ .004) higher maximal oxygen uptake and the duration of test was 96.5 s (95% CI 8.0–185.0; p ¼.03) longer compared to participants in the exercise control group. As illustrated in Fig. 4 there was no effect of the aerobic exercise training on copeptin levels at rest or at VO2max. Adjusting VO2max, the duration of test, or the HRmax/HRexpected ratio did not change the results for copeptin at rest or VO2max. We also analyzed Table 2 Hormones in patients with depression and healthy controls.
Copeptina, pmol/L CopeptinVO2maxa,b, pmol/L ACTHa, pmol/L Cortisola,c, nmol/L
Major depression N ¼111
Healthy controls N ¼57
P value
5.2 9.7 5.0 413.0
4.8 13.4 4.9 411.0
0.47 0.16 0.55 0.54
(3.4–8.3) (4.5–26.3) (3.5–7.0) (290–533)
(2.8–7.5) (6.8–31.4) (2.9–7.1) (309–515)
2
Median and interquartile range is presented. a
Estimation of p value is based on log transformed values. Copeptin measured at peak VO2max during exercise test. c All participants using hormonal replacement therapy or oral contraceptives were excluded from cortisol analysis.
ln ACTH
ln cortisol
b
ln copeptin
ln copeptin
Fig. 2. Correlation between ACTH and copeptin and cortisol and copeptin in depressed and healthy subjects.
288
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
35
pmol/L
time: F4, 154 = 70.1; p < 0.001 disease status x time: F4,154 = 1.7;p = 0.15
30 25 20 15 10 5 0 At rest
VO2max
+15 min. Depressed
+30 min
+60 min
Healthy
Fig. 3. Copeptin response to a an ergometer test in depressed and healthy controls. Copeptin levels are presented with median and interquartile range. Blood samples were collected before the test, at VO2max, and 15, 30, and 60 min after termination of the test. The p-values are based on ln transformed values using a mixed model with repeated measurement approach.
30 pmol/L
All participants
High attendance group
25 P = 0.11
P = 0.73
At rest
At VO2max
P = 0.04
P = 0.17
20 15 10 5 0 Exercise
At rest
At VO2max
Exercise control
Fig. 4. Copeptin at rest and at VO2max post-intervention Copeptin is presented with median and errorbars represents interquartile range. The ‘high attendance group’ analysis is restricted to participants that participated in 50% or more of the training sessions. All statistical analysis is performed after ln transformation and imputation of missing values.
the data utilizing all data points from the maximal exercise test, which supported the above finding (time allocation; F2,76 ¼.73; p ¼.57). Adjusting for duration of test did not alter this result (F4,56 ¼.41; p ¼.80). Post intervention, the association between cortisol and copeptin was –.12 (95% CI –.24 to.01; p¼.07), which was independent of group (p¼.28). In other words, aerobic exercise training was not able to induce changes in the association between copeptin and cortisol in this study. Post intervention the median plasma ACTH at rest was 4.2 pmol/L (IQR 3.6–6.9) in the aerobic exercise group and 5.4 pmol/L (IQR 3.0–6.4) in the exercise control group (p ¼.82). Plasma cortisol was 361 nmol/L (IQR 293–456) in the aerobic exercise group and 383 nmol/L (IQR 280–491) in the exercise control group (p¼ .54).
3.5. Participants with high attendance Post-hoc we restricted our analysis to the 20 participants in the aerobic exercise group and 22 participants in the exercise control group that attended 50% or more of the training sessions. In this subgroup the median participation in the aerobic exercise group was 24 sessions (IQR 20–27) versus 23 sessions (IQR 20–26) in the exercise control group of a total of 36 training sessions (p ¼.75). As illustrated in Fig. 4, at rest plasma copeptin was significantly lower in the aerobic exercise group post intervention compared to the exercise control group (p ¼.04) but not at VO2max (p ¼.17). Change in copeptin from baseline to post intervention was not significantly associated with change in HAM-D17 (p¼ .15), change in VO2max (p ¼ .64), or number of sessions (p ¼.74). The association between cortisol and copeptin remained insignificant (p¼ .62) and independent of allocation (p ¼.62) in this subgroup.
4. Discussion The current study did not find that copeptin plasma levels during rest or in response to acute physical exercise differed between a group of mildly to moderately depressed outpatients and healthy controls. Neither did we find that copeptin was associated with cortisol in healthy controls, depressed patients or in patient subgroups with generalized anxiety. A three months aerobic exercise program did not change copeptin levels during rest or acute exercise, nor the association to cortisol. To our best knowledge copeptin levels in depression have not been reported previously. One large study (57 patients and 37 healthy controls) by van Londen et al. (1997) has reported higher levels of vasopressin in depressed compared to healthy subjects, which is in contradiction to the current results. However, the high vasopressin levels in that study (van Londen et al., 1997) were primarily observed in inpatients whereas the current population only included outpatients. In line with the present study, the results from smaller studies by Inder et al. (1997) and Gjerris et al. (1985) did not find differences in plasma vasopressin levels between depressed outpatients and healthy controls. Concerning the possible association between cortisol and AVP in depressed patients the results are inconclusive. The previous study by Inder et al. (1997) and another recent study (Goekoop et al., 2011) found a positive association between cortisol and AVP. On the other hand a similar study of both in- and outpatients found no association between cortisol and AVP (Goekoop et al., 1997). In the present study we found no correlation between cortisol and copeptin. In favor of the current results is a recent publication found that found no antidepressant effect of a V1b receptor antagonist compared to placebo in two of three trials (Griebel et al., 2012). It should be mentioned, that the ACTH and cortisol levels in the studied population did not suggest a hyperactive HPA axis in this group of depressed patients. It could be argued that our hypothesized association between cortisol and AVP would be the scenario only in individuals with hyperactive HPA-axis. Taken together, it remains unclear if there is an association between cortisol and AVP in depressed patients as a group or in sub-groups of depressed patients. As previously shown, exercise is a potent stimulator of copeptin (Maeder et al., 2010) and AVP release (Takamata et al., 2000) if measured immediately post-exercise (Sanchis-Gomar et al., in press). The primary stimulus of AVP release during exercise is probably increased plasma osmolality due to free water movement out of the vascular space (Takamata et al., 2000). A limitation in the current study was that osmolality was not measured. However, a previous publication found normal to low plasma osmolality in depressed subjects compared to healthy controls (van Londen et al., 1997). Furthermore, none of the participants were diagnosed with renal or cardiac failure known to influence AVP secretion. In contrast to the report by Maeder et al. (2010) investigating a group of patients with known or suspected myocardial ischemia, we did not find that copeptin release during exercise was associated to heart rate, but to resting levels of copeptin and duration of test. The positive association between copeptin levels at VO2max and the duration of the test could reflect increasing osmolality as observed during exercise. Maximal oxygen uptake was not associated to copeptin at VO2max, potentially explained by a poor association between plasma osmolality and VO2max in the current testing situation. At termination of the randomized exercise intervention the participants in the aerobic group had increased their maximal oxygen uptake with 13 percent while the exercise control group had a four percent decline in maximal oxygen uptake (data not shown). It has previously been reported that physical conditioning is associated with a reduction in HPA axis activation in response to a given workload (Luger et al., 1987) and
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
we have previously reported an augmented GH response to acute exercise in depressed subjects after a strength training program (Krogh et al., 2010). In the current study we did not find that copeptin levels at rest or at VO2max were affected by aerobic exercise training. However, restricting our analysis to participants with a high attendance rate we found a significant lowering of plasma copeptin in the aerobic exercise group. This corresponds with findings from exercise studies of heart failure patients (Gademan et al., 2007) and in rat models of heart failure (Lin et al., 2011). The overall participation in training sessions was quite low as could be expected in depressed patients. Therefore, the current study does not allow us to conclude that aerobic exercise training per se is unable to induce changes in the HPA axis, but rather that such potential effect of exercise training is unlikely to occur in a sample of depressed outpatients. This is in line with our previously reported findings that a similar intervention was unable to induce changes in resting levels of growth hormone, prolactin, or cortisol in a similar set-up (Krogh et al., 2010). The ability to draw conclusions from copeptin to AVP relies on the extent plasma copeptin reflects plasma vasopressin. One study have found a correlation of 0.8 between copeptin and vasopressin in healthy volunteers during different states of osmolality, which was supported by a high correlation between copeptin and osmolality (Balanescu et al., 2011). The high copeptin/AVP correlation has previously been found in critically ill patients (Jochberger et al., 2006; Morgenthaler et al., 2006). However, it should be mentioned that at least two studies have found lower copeptin/AVP correlations around 0.3–0.4 in healthy individuals (Jochberger et al., 2006; Morgenthaler et al., 2006). Despite the large number of published studies using copeptin as a marker of vasopressin, an indication of the general acceptability of this approach, the current results should be evaluated in the perspective of the above discussion. To the best of our knowledge this is the first study of copeptin in patients with depression. In conclusion, the present results do not suggest a key-role of copeptin or vasopressin in depression and copeptin is therefore not a candidate as a biomarker of depression.
Role of funding source Funding support for this study was provided by Trygfonden, Nordea Danmark fonden, Aase og Ejnar Danielsens Fond. The study's funding sources had no role in the study design; collection, analysis, or interpretation of data; writing of the report; and decision to submit the paper for publication. The corresponding author had full access to all the data and had final responsibility for the decision to submit for publication.
Conflict of interest All authors report no biomedical financial interest or potential conflicts of interest.
Acknowledgment We are most grateful to the participants and funders of our study. We would also like to thank our interviewers, test staff and the physiotherapists that were responsible for the intervention.
References Andersen, L.B., 1995. A maximal cycle exercise protocol to predict maximal oxygen uptake. Scandinavian Journal of Medicine and Science in Sports 5, 143–146. Balanescu, S., Kopp, P., Gaskill, M.B., Morgenthaler, N.G., Schindler, C., Rutishauser, J., 2011. Correlation of plasma copeptin and vasopressin concentrations in hypo-, iso-, and hyperosmolar States. Journal of Clinical Endocrinology and Metabolism 96 (4), 1046–1052. Bech, P., Bech-Andersen, G., Schütze, T., 1999. Mini Internationalt Neuropsykiatrisk Interview. Dansk version 5.0.0. 〈www.ccmh.dk〉.
289
De Goeij, D.C., Dijkstra, H., Tilders, F.J., 1992a. Chronic psychosocial stress enhances vasopressin, but not corticotropin-releasing factor, in the external zone of the median eminence of male rats: relationship to subordinate status. Endocrinology 131 (2), 847–853. De Goeij, D.C., Jezova, D., Tilders, F.J., 1992b. Repeated stress enhances vasopressin synthesis in corticotropin releasing factor neurons in the paraventricular nucleus. Brain Research 577 (1), 165–168. Gademan, M.G., Swenne, C.A., Verwey, H.F., van der Laarse, A., Maan, A.C., van de Vooren, H., van Pelt, J., van Exel, H.J., Lucas, C.M., Cleuren, G.V., Somer, S., Schalij, M.J., van der Wall, E.E., 2007. Effect of exercise training on autonomic derangement and neurohumoral activation in chronic heart failure. Journal of Cardiac Failure 13 (4), 294–303. Gjerris, A., Hammer, M., Vendsborg, P., Christensen, N.J., Rafaelsen, O.J., 1985. Cerebrospinal fluid vasopressin—changes in depression. British Journal of Psychiatry 147, 696–701. Goekoop, J., de Winter, R., Wolterbeek, R., Wiegant, V., 2011. Support for two increased vasopressinergic activities in depression at large and the differential effect of antidepressant treatment. Journal of Psychopharmacology 25 (10), 1304–1312. Griebel, G., Beeske, S., Stahl, S.M., 2012. The vasopressin V(1b) receptor antagonist SSR149415 in the treatment of major depressive and generalized anxiety disorders: results from 4 randomized, double-blind, placebo-controlled studies. Journal of Clinical Psychiatry 73 (11), 1403–1411. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 23, 56–62. Inder, W.J., Donald, R.A., Prickett, T.C., Frampton, C.M., Sullivan, P.F., Mulder, R.T., Joyce, P.R., 1997. Arginine vasopressin is associated with hypercortisolemia and suicide attempts in depression. Biological Psychiatry 42 (8), 744–747. Jochberger, S., Morgenthaler, N.G., Mayr, V.D., Luckner, G., Wenzel, V., Ulmer, H., Schwarz, S., Hasibeder, W.R., Friesenecker, B.E., Dunser, M.W., 2006. Copeptin and arginine vasopressin concentrations in critically ill patients. Journal of Clinical Endocrinology and Metabolism 91 (11), 4381–4386. Klose, M., Kofoed-Enevoldsen, A., Ostergaard, K.L., 2002. Single determination of plasma ACTH using an immunoradiometric assay with high detectability differentiates between ACTH-dependent and -independent Cushing's syndrome. Scandinavian Journal of Clinical and Laboratory Investigation 62 (1), 33–37. Krogh, J., Nordentoft, M., Mohammad-Nezhad, M., Westrin, Å., 2010. Growth hormone, prolactin and cortisol response to exercise in patients with depression. Journal of Affective Disorders 125, 189–197. Krogh, J., Videbech, P., Thomsen, C., Gluud, C., Nordentoft, M., 2011. DEMO-II trial. Aerobic exercise versus stretching exercise in patients with major depression— a randomised clinical trial. PLoS One 7 (10), e48316. Lin, Q.Q., Lin, R., Ji, Q.L., Zhang, J.Y., Wang, W.R., Yang, L.N., Zhang, K.F., 2011. Effect of exercise training on renal function and renal aquaporin-2 expression in rats with chronic heart failure. Clinical and Experimental Pharmacology and Physiology 38 (11), 179–185. Liu, J., Hu, P., Qi, X.R., Meng, F.T., Kalsbeek, A., Zhou, J.N., 2011. Acute restraint stress increases intrahypothalamic oestradiol concentrations in conjunction with increased hypothalamic oestrogen receptor beta and aromatase mRNA expression in female rats. Journal of Neuroendocrinology 23 (5), 435–443. Luger, A., Deuster, P., Kyle, S., Gallucci, W., Montgomery, L., Gold, P., Loriaux, L., Chrousos, G., 1987. Acute hypothalamic-pituitary-adrenal responses to the stress of treadmill exercise. New England Journal of Medicine 316, 1309–1315. Maeder, M.T., Staub, D., Brutsche, M.H., Arenja, N., Socrates, T., Reiter, M., Meissner, J., Morgenthaler, N.G., Bergmann, A., Struck, J., Mueller, C., 2010. Copeptin response to clinical maximal exercise tests. Clinical Chemistry 56 (4), 674–676. Makino, S., Schulkin, J., Smith, M.A., Pacak, K., Palkovits, M., Gold, P.W., 1995. Regulation of corticotropin-releasing hormone receptor messenger ribonucleic acid in the rat brain and pituitary by glucocorticoids and stress. Endocrinology 136 (10), 4517–4525. Molnar, A.H., Varga, C., Janaky, T., Toth, G., Toth, G., Farkas, J., Laszlo, F., Laszlo, F.A., 2007. Biological half-life and organ distribution of [3H]8-arginine vasopressin following administration of vasopressin receptor antagonist OPC-31260. Regulatory Peptides 141 (1–3), 12–18. Morgenthaler, N.G., Struck, J., Alonso, C., Bergmann, A., 2006. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clinical Chemistry 52 (1), 112–119. Rabadan-Diehl, C., Lolait, S.J., Aguilera, G., 1995. Regulation of pituitary vasopressin V1b receptor mRNA during stress in the rat. Journal of Neuroendocrinology 7 (12), 903–910. Sanchis-Gomar, F., Bonaguri, C., Aloe, R., Pareja-Galeano, H., Martinez-Bello, V., Gomez-Cabrera, M.C., Candel, J., Vina, J., Lippi, G., 2013. Effects of acute exercise and xanthine oxidase inhibition on novel cardiovascular biomarkers. Translational Research, http://dx.doi.org/10.1016/j.trsl.2013.02.006, in press. Scott, L.V., Dinan, T.G., 1998. Vasopressin and the regulation of hypothalamicpituitary-adrenal axis function: implications for the pathophysiology of depression. Life Sciences 62 (22), 1985–1998. Takamata, A., Nose, H., Kinoshita, T., Hirose, M., Itoh, T., Morimoto, T., 2000. Effect of acute hypoxia on vasopressin release and intravascular fluid during dynamic exercise in humans. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 279 (1), R161–R168. Terzic, D., Johansson-Fällgren, A., Ragnarsson, O., Goetze, J.P., Hammarsten, O., 2012. Evaluation of a sensitive copeptin assay for clinical measurement. Open Clinical Chemistry Journal 5, 21–26.
290
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
van Londen, L., Goekoop, J.G., van Kempen, G.M., Frankhuijzen-Sierevogel, A.C., Wiegant, V.M., van der Velde, E.A., De Wied, D., 1997. Plasma levels of arginine vasopressin elevated in patients with major depression. Neuropsychopharmacology 17 (4), 284–292. Wang, S.S., Yan, X.B., Hofman, M.A., Swaab, D.F., Zhou, J.N., 2010. Increased expression level of corticotropin-releasing hormone in the amygdala and in the hypothalamus in rats exposed to chronic unpredictable mild stress. Neuroscience Bulletin 26 (4), 297–303.
Zhang, L., Hernandez, V.S., Liu, B., Medina, M.P., Nava-Kopp, A.T., Irles, C., Morales, M., 2012. Hypothalamic vasopressin system regulation by maternal separation: its impact on anxiety in rats. Neuroscience 215, 135–148. Zhao, D.Q., Ai, H.B., 2011. Oxytocin and vasopressin involved in restraint waterimmersion stress mediated by oxytocin receptor and vasopressin 1b receptor in rat brain. PLoS One 6 (8), e23362.
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research report
Copeptin during rest and exercise in major depression Jesper Krogh a,c,n, Jens Peter Gøtze b, Martin Balslev Jørgensen a, Lars Østergaard Kristensen c, Caroline Kistorp c, Merete Nordentoft d a
Mental Health Centre Copenhagen, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark Department of Clinical Biochemistry, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark c Department of Endocrinology, Herlev University Hospital, University of Copenhagen, Copenhagen, Denmark d Mental Health Centre Copenhagen, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark b
art ic l e i nf o
a b s t r a c t
Article history: Received 19 April 2013 Received in revised form 10 June 2013 Accepted 10 June 2013 Available online 13 July 2013
Background: High vasopressin levels and a correlation between vasopressin and cortisol has been observed in patients with depression. The aim was to assess copeptin, the c-terminal of provasopressin, and the association between cortisol, adrenocorticotropic hormone (ACTH) and copeptin in patients with depression. Secondly, to examine the copeptin response to acute exercise and aerobic training. Methods: Copeptin, ACTH, and cortisol were measured in 111 patients with depression and 57 controls at rest. Copeptin was also measured during exercise. The depressed patients were subsequently randomized to an aerobic training intervention or an exercise control intervention. Results: The plasma level of copeptin in depressed subjects was 5.14 pg/ml (IQR 3.4–8.4) and 4.82 pg/ml (IQR 2.8–7.5) in healthy controls (p¼ .66). The association between copeptin and cortisol was.02 (95% CI –.44 to.48; p¼ .93) and the association between copeptin and ACTH was –.06 (95% CI –.17 to.05; p¼ .27). All associations were independent of depression status (p¼ .15). Aerobic exercise training did not influence copeptin levels at rest (p ¼.09) or the response to acute exercise (p¼ .574). Copeptin decreased at rest in response to aerobic training in participants with high compliance to the exercise intervention (p¼ .04). Limitations: We did not measure plasma osmolality, which is a possible confounder in this study. Conclusions: Copeptin levels are not elevated or associated to ACTH or cortisol in depressed patients. Aerobic exercise training decreased copeptin levels in high attenders only. This study does not support a role of copeptin or vasopressin in depression. & 2013 Elsevier B.V. All rights reserved.
Keywords: Copeptin Vasopressin Depression Exercise test Exercise training ACTH
1. Introduction It is widely recognized that the observed hypercortisolism in a subgroup of patients with depression is due to centrally mediated hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis. This hypothesis has been strengthened by the finding of raised corticotrophin releasing hormone (CRH) concentrations in cerebrospinal fluid (Scott and Dinan, 1998). CRH is secreted from the paraventricular nucleus (PVN) in hypothalamic region and acts on the corticotrope cells of the pituitary to release adrenocorticotropic hormone (ACTH). Although CRH is considered the major ACTH secretagogue, the release of ACTH occurs in concert with arginine vasopressin (AVP) also released from the PVN. Apart from mediating ACTH release, AVP induces water retention by the kidney, being the main hormone in the regulation of osmotic homeostasis. n Correspondence to: Mental Health Centre Copenhagen, Bispebjerg Bakke 23, DK-2400, Denmark. Tel. : +45 2553 5485; fax: +45 3531 3953. E-mail address: [email protected] (J. Krogh).
0165-0327/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2013.06.007
Both chronic and acute stress appears to induce increased CRH expression in the PVN (Liu et al., 2011; Wang et al., 2010; Zhang et al., 2012; Zhao and Ai, 2011). At the pituitary level chronic stress is associated with down regulation of CRH receptors (De Goeij et al., 1992a; De Goeij et al., 1992b; Rabadan-Diehl et al., 1995) while the expression of vasopressin (V1b) receptors increases. Interestingly, the AVP induced ACTH release has been found to be independent of the glucocorticoid feedback (Makino et al., 1995). A gradual shift towards AVP regulated HPA axis in response to chronic stress at the pituitary level, and glucocorticoid mediated suppression of CRH stress response, could suggest an interesting role for vasopressin in affective disorders. Elevated plasma AVP levels in patients with unipolar depression have been reported in one larger study (van Londen et al., 1997) but not in others (Gjerris et al., 1985; Inder et al., 1997). However, a recent study found a strong positive correlation between cortisol and AVP in patients without antidepressant treatment which was interpreted as an indication of high vasopressinergic activity in this patient group (Goekoop et al., 2011). In support of this, a number
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
of studies have found AVP-cortisol correlations in patients with anxious-retarded depression (Goekoop et al., 2011) as well as in suicide victims (Inder et al., 1997). Concern exist about the methodological reliability of plasma AVP measurements because AVP is known to be unstable, attached to platelets and a plasma half-life of approximately 10 min (Molnar et al., 2007; Morgenthaler et al., 2006). AVP is derived from a larger precursor peptide named provasopressin. Copeptin is a 39-amino acid glycopeptide constituting the C-terminal part of provasopressin. Because copeptin and AVP are released in equimolar ratio, copeptin levels potentially reflect the levels of AVP, which has been confirmed in intensive care unit patients (Jochberger et al., 2006). Copeptin has longer plasma half-life compared to vasopressin (∼24 min), and has been found stable in both serum and heparin plasma at room temperature for at least a week (Morgenthaler et al., 2006). Furthermore, due to the small size of AVP it cannot be measured by sandwich immunoassays but needs competitive assays, which generally are less sensitive. We hypothesized that plasma copeptin would be higher in depressed patients at rest, and that we would observe an augmented copeptin response to HPA-axis stimulation during an acute exercise test in depressed patients compared to healthy controls. We also hypothesized that the association between cortisol and copeptin and ACTH and copeptin would be higher in patients with depression as an indication of increased vasopressinergic drive in this patient group. Secondly, we expected that a three months aerobic exercise program would provide lower plasma copeptin levels at rest and during acute exercise, and an attenuated association between cortisol and copeptin and ACTH and copeptin.
2. Methods 2.1. Study population All 115 participants from a previous trial (the DEMO-II trial), a randomized clinical trial evaluating the antidepressant effect of aerobic exercise (Krogh et al., 2012) were eligible for the current study. Eligible participants were men and women between 18 and 60 years of age, referred from a clinical setting by a physician or a psychologist, and a diagnose of major depression (DSM-IV) based on the Danish version of the Mini International Neuropsychiatric Interview (MINI) (Bech et al., 1999). The participants scored above 12 on the Hamilton Depression Rating Scale (HAM-D17) (Hamilton, 1960) and were able to comprehend the informed consent statement. Exclusion criteria were current drug abuse, any antidepressant medication within the last two months, current psychotherapeutic treatment, contraindications to physical exercise, regular recreational exercise over 1 h per week, suicidal behavior according to the HAM-D17 (item 342), pregnancy, or current/previous psychotic or manic symptoms. Four participants from the original trial did not supply blood samples for the current study, thus 111 participants constitute the group of depressed patients for the current study. Healthy controls were recruited through the media and group matched to the patients with regard to sex, age, and body mass index (n¼ 57). The healthy controls were free of any current or previous psychiatric diseases assessed by the MINI (Bech et al., 1999). Evaluation of patients and healthy controls commenced in parallel. Fig. 1 shows patient flow. 2.2. Physical examination For the physical examination, the participants were requested to meet at the research department between 8:00 and 10:00 a.m. The participants were instructed not to take any food or liquids except for water beginning from midnight prior to the examination and abstain from strenuous physical activity prior to the
285
examination. Weight was measured using an electronic weight (Sohnle Medicals, Type 7700, Backnang, Germany). Blood pressure was measured after five minutes rest with the participant in a sitting position using a certified digital blood pressure monitor (Omron M6, Omron Healthcare co. LTD, Kyoto, Japan). The average of three measurements on the right arm is reported. VO2max was estimated using a bicycle cardiopulmonary exercise test (Ergomedic 839 e, Monark, Vansbro, Sweden) based on a standardized exercise protocol (Andersen, 1995). An indwelling venous catheter was inserted in an antecubital vein and blood samples collected after 5 min rest in a sitting position. Samples were collected 10–15 min prior to the exercise test, at VO2max, and 15, 30, and 60 min after the test. The samples were immediately centrifuged at room temperature and stored at −80 1C till analysis. 2.3. Hormone analyses Copeptin was analyzed for all data points and ACTH and cortisol only during rest. Copeptin was measured by the US Kryptor assay as previously described (Terzic et al., 2012). The intra-assay CVmax was estimated to be 18% and 6.3% for plasma copeptin concentrations of 3.0 pmol/l and 7.9 pmol/l, respectively. ACTH was analyzed using an IRMA from BRAHMS (Klose et al., 2002). The intra-assay coefficient of variation was estimated to be 7.9% at the concentration of interest. Cortisol was analyzed using an automated competitive electrochemiluminiscens-immunoassay (ECLIA) from Roche (Modular E) with a CVmax of 8% in the relevant range of analysis. 2.4. Exercise intervention After baseline assessment the depressed patients were randomised to either an aerobic exercise intervention or an exercise control intervention as described in detail elsewhere (Krogh et al., 2012). In short, both groups were scheduled to participate in 45 min supervised sessions three times per week in a 3 months program. In the aerobic training group the participants exercised on stationary bikes at approximately 80% of their maximal heart rate, while the training in the exercise control group was predominantly stretching exercises and low impact exercise such as throwing and catching balls. 2.5. Approvals The protocol was approved by the local ethics committee (H-A2008-046) and the Danish Data Protection Agency (J.nr.2008-412354). Verbal and written informed consent was obtained from all participants involved in the trial. 2.6. Statistics Differences between depressed patients and healthy controls were evaluated using chi-square test and independent-samples t-test. The association between copeptin, ACTH, and cortisol was analyzed using linear regression analysis. The effect of disease status (depressed versus healthy) was examined using an interaction term. In all cortisol analysis we excluded participants using contraceptive pills or hormone replacement therapy. Except where specified, all analyses of copeptin, ACTH, and cortisol were performed after natural log (ln) transformation. The effect of acute exercise was estimated by analyzing copeptin values at VO2max in a linear regression model with depression status as the main dependent variable and sex, age, maximal oxygen uptake, duration of test, systolic blood pressure, and baseline copeptin as independent variables. We used stepwise backwards regression analysis and kept all independent variables with p4 0.1 in the model to avoid overfitting. Secondly, we used a
286
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
Healthy subjects
Depressed subjects
Recruited from mass media:
Recruited from the DEMO-II trial:
N = 57
N = 115
Not available for biomarker assessment du too technical problems
Not available for biomarker assessment du too technical problems
N=4
N=0
Assessment
Assessment
Plasma measurement of copeptin, ACTH, and cortisol N = 111
Randomization
PART 1 Depressed vs. healthy controls
Plasma measurement of copeptin, ACTH, and cortisol N = 57
Part 2. Effect of an exercise intervention in depressed patients
Aerobic exercise training
Exercise control
N = 53
N = 58
Follow-up
Follow-up
Available for follow-up: n = 39 Lost to follow-up: n = 14
Available for follow-up: n = 44 Lost to follow-up: n = 14
Included in analysis: n = 53
Included in analysis: n = 58
Part 1 refers to the assessement of copeptin, ACTH, and cortisol in depressed patients and healthy controls. Part 2 refers to the effect of an exercise intervention on copeptin, ACTH, and cortisol in depressed patients. Fig. 1. Flowchart of participant flow.
The 111 patients and 57 healthy controls were comparable with respect to baseline demographic and clinical characteristics as shown in Table 1.
healthy controls (Mann-Whitney U; p ¼.66). There was no difference between plasma levels of ACTH and cortisol comparing depressed subjects with healthy controls (p 4.40) as shown in Table 1. The univariate association between ACTH and copeptin was –.06 (95% CI –.17 to.05; p ¼.27) and between cortisol and copeptin.02 (95% CI –.44 to.48; p ¼.93) suggesting no association between cortisol and copeptin or ACTH and copeptin in the whole population. The interaction term disease status copeptin was not significant (p ¼.29) indicating that the association between cortisol and copeptin was not different in depressed patients compared to healthy controls as also illustrated in Fig. 2. Neither did the data suggest that the association between ACTH and copeptin was different in depressed patients compared to healthy controls (disease status copeptin level; p ¼.43). The association between cortisol and ACTH was.47 (95% CI.17 to.33; po.001) corresponding to a Pearson's correlation of.40 (po.001), irrespective of disease status (disease status ACTH; p¼.82). Introducing generalized anxiety as an independent variable the association between cortisol and copeptin was not affected (p ¼.77). Copeptin levels at baseline did not predict remission in the patient group at three months follow-up.
3.2. Copeptin during rest
3.3. Copeptin during exercise
The median plasma level of copeptin in depressed subjects was 5.2 pg/ml (IQR 3.4–8.3) and 4.82 pg/ml (IQR 2.8–7.5) in
As illustrated in Table 2 maximal oxygen uptake was 3.88 ml/kg/ min (95% CI 1.7–6.1; p¼.001) lower in depressed patients. The
repeated measurement mixed model with unstructured variance including all five measurements points. In a similar way of analyses, we compared baseline differences and the effect of the randomized exercise intervention in the depressed patients. For missing data, post-intervention on copeptin, ACTH, and cortisol at rest and during VO2max we used multiple imputation applying linear regression using 100 imputations and a maximum of 20 iterations. Group allocation, baseline resting levels, and baseline copeptin levels at VO2max was used as predictor variables. SPSS version 19.0 was used for all analysis. All statistical tests were two-sided and p-values below 0.05 were considered significant.
3. Results 3.1. Participants
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
duration of the test was on average 98.9 s (95% CI 43–154; p¼.001) shorter in depressed patients and the mean maximal heart rate was 6.0 (95% CI –.3 to 12.3; p¼.06) lower compared to the healthy controls. As showed in Fig. 3 there was a significant main effect of an incremental bicycle test on copeptin levels (time: F4, 154 ¼ 70.1; p o.001). However, variations in copeptin response to acute exercise were not affected by depression (disease status time: F4,154 ¼1.7; p ¼.15). The highest level of plasma copeptin was observed at VO2max and the median increase from baseline was
Table 1 Clinical and demographical characteristics. Major depression N ¼111
Healthy controls N ¼ 57
P value
Age, years (SD) Female, n (%)
41.9 (11.2) 75 (68.2%)
40.6 (13.1) 37 (64.9%)
0.50 0.73
Psychometrics HAM-D17 (SD) HAM-A14 (SD) Generalized anxiety, n (%) Recurrent depression, n (%) Age at first depression (SD)
18.9 (3.8) 17.6 (5.5) 65 (59.1%) 58 (52.7%) 36.4 (12.1)
1.1 (1.5) 1.2 (1.7) N/A N/A N/A
– – – – –
66.2 (12.3) 122.8 (17.0)
67.6 (12.3) 119.0 (13.3)
0.51 0.16
82.1 (11.8)
79.3 (9.6)
0.14
77.3 (20.4) 26.4 (6.0)
76.8 (18.4) 25.9 (6.3)
0.86 0.66
25.4 (7.0)
29.3 (6.0)
0.001
666.3 (170.9) 163.3 (17.3)
765.2 (162.0) 169.3 (22.2)
0.001 0.06
178.6 (7.8)
179.6 (9.2)
0.50
0.91 (0.09)
0.94 (0.11)
0.09
Somatic Creatinine, mol/L (SD) Systolic blood pressure, mmHg (SD) Diastolic blood pressure, mmHg (SD) Weight, kg (SD) Body mass index (SD) Exercise test Maximal oxygen uptake, ml/kg/min (SD) Exercise time, seconds (SD) Maximum pulse, observed (SD) Maximum pulse, expected (SD) HRmax/HRmax expected (SD) 1
P value is based on log transformed values. All participants using hormone replacement therapy or oral contraceptives were excluded from cortisol analysis. Abbreviations: HAM-D17—Hamilton Depression Scale, 17 items; HAM-A14—Hamilton Anxiety Scale, 14 items; ACTH—adrenocorticotrop hormone; HRmax—highest heart rate observed during test; HRmax expected—expected maximal heart rate was derived from sex and age.
287
3.8 pg/ml (IQR.4–13.9). Using stepwise backwards regression, the final model suggested that copeptin levels at VO2max were explained by duration of test (β ¼.001; 95% CI .000–.002; p ¼.04) and baseline copeptin levels (β¼.672; 95% CI .539–.810 p o.001) and not by depression. This model explained 40.4% (R2) of the variation in copeptin levels at VO2max. 3.4. Effects of exercise training By randomization the 111 patients were allocated to either aerobic exercise training (n ¼53) or an exercise control (n¼ 58). The two groups did not differ significantly with respect to any of the biomarkers or other variables shown in Table 1 (all p values 4.10). At three months follow-up, copeptin assessment was available from 39/53 in the aerobic training group and 44/58 in the exercise control group. Comparing participants who had available copeptin levels at follow-up with participants who were missing we found no differences regarding sex, age, depression severity, or group allocation (all p values 4.2). The median participation in training sessions was 14.0 sessions (IQR 5–22) in the aerobic exercise group and 10.5 session (IQR 4–21) in the exercise control group (p ¼.41). Post intervention, participants in the aerobic exercise group had a 5.2 ml/kg/min (95% CI 1.8–8.7; p¼ .004) higher maximal oxygen uptake and the duration of test was 96.5 s (95% CI 8.0–185.0; p ¼.03) longer compared to participants in the exercise control group. As illustrated in Fig. 4 there was no effect of the aerobic exercise training on copeptin levels at rest or at VO2max. Adjusting VO2max, the duration of test, or the HRmax/HRexpected ratio did not change the results for copeptin at rest or VO2max. We also analyzed Table 2 Hormones in patients with depression and healthy controls.
Copeptina, pmol/L CopeptinVO2maxa,b, pmol/L ACTHa, pmol/L Cortisola,c, nmol/L
Major depression N ¼111
Healthy controls N ¼57
P value
5.2 9.7 5.0 413.0
4.8 13.4 4.9 411.0
0.47 0.16 0.55 0.54
(3.4–8.3) (4.5–26.3) (3.5–7.0) (290–533)
(2.8–7.5) (6.8–31.4) (2.9–7.1) (309–515)
2
Median and interquartile range is presented. a
Estimation of p value is based on log transformed values. Copeptin measured at peak VO2max during exercise test. c All participants using hormonal replacement therapy or oral contraceptives were excluded from cortisol analysis.
ln ACTH
ln cortisol
b
ln copeptin
ln copeptin
Fig. 2. Correlation between ACTH and copeptin and cortisol and copeptin in depressed and healthy subjects.
288
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
35
pmol/L
time: F4, 154 = 70.1; p < 0.001 disease status x time: F4,154 = 1.7;p = 0.15
30 25 20 15 10 5 0 At rest
VO2max
+15 min. Depressed
+30 min
+60 min
Healthy
Fig. 3. Copeptin response to a an ergometer test in depressed and healthy controls. Copeptin levels are presented with median and interquartile range. Blood samples were collected before the test, at VO2max, and 15, 30, and 60 min after termination of the test. The p-values are based on ln transformed values using a mixed model with repeated measurement approach.
30 pmol/L
All participants
High attendance group
25 P = 0.11
P = 0.73
At rest
At VO2max
P = 0.04
P = 0.17
20 15 10 5 0 Exercise
At rest
At VO2max
Exercise control
Fig. 4. Copeptin at rest and at VO2max post-intervention Copeptin is presented with median and errorbars represents interquartile range. The ‘high attendance group’ analysis is restricted to participants that participated in 50% or more of the training sessions. All statistical analysis is performed after ln transformation and imputation of missing values.
the data utilizing all data points from the maximal exercise test, which supported the above finding (time allocation; F2,76 ¼.73; p ¼.57). Adjusting for duration of test did not alter this result (F4,56 ¼.41; p ¼.80). Post intervention, the association between cortisol and copeptin was –.12 (95% CI –.24 to.01; p¼.07), which was independent of group (p¼.28). In other words, aerobic exercise training was not able to induce changes in the association between copeptin and cortisol in this study. Post intervention the median plasma ACTH at rest was 4.2 pmol/L (IQR 3.6–6.9) in the aerobic exercise group and 5.4 pmol/L (IQR 3.0–6.4) in the exercise control group (p ¼.82). Plasma cortisol was 361 nmol/L (IQR 293–456) in the aerobic exercise group and 383 nmol/L (IQR 280–491) in the exercise control group (p¼ .54).
3.5. Participants with high attendance Post-hoc we restricted our analysis to the 20 participants in the aerobic exercise group and 22 participants in the exercise control group that attended 50% or more of the training sessions. In this subgroup the median participation in the aerobic exercise group was 24 sessions (IQR 20–27) versus 23 sessions (IQR 20–26) in the exercise control group of a total of 36 training sessions (p ¼.75). As illustrated in Fig. 4, at rest plasma copeptin was significantly lower in the aerobic exercise group post intervention compared to the exercise control group (p ¼.04) but not at VO2max (p ¼.17). Change in copeptin from baseline to post intervention was not significantly associated with change in HAM-D17 (p¼ .15), change in VO2max (p ¼ .64), or number of sessions (p ¼.74). The association between cortisol and copeptin remained insignificant (p¼ .62) and independent of allocation (p ¼.62) in this subgroup.
4. Discussion The current study did not find that copeptin plasma levels during rest or in response to acute physical exercise differed between a group of mildly to moderately depressed outpatients and healthy controls. Neither did we find that copeptin was associated with cortisol in healthy controls, depressed patients or in patient subgroups with generalized anxiety. A three months aerobic exercise program did not change copeptin levels during rest or acute exercise, nor the association to cortisol. To our best knowledge copeptin levels in depression have not been reported previously. One large study (57 patients and 37 healthy controls) by van Londen et al. (1997) has reported higher levels of vasopressin in depressed compared to healthy subjects, which is in contradiction to the current results. However, the high vasopressin levels in that study (van Londen et al., 1997) were primarily observed in inpatients whereas the current population only included outpatients. In line with the present study, the results from smaller studies by Inder et al. (1997) and Gjerris et al. (1985) did not find differences in plasma vasopressin levels between depressed outpatients and healthy controls. Concerning the possible association between cortisol and AVP in depressed patients the results are inconclusive. The previous study by Inder et al. (1997) and another recent study (Goekoop et al., 2011) found a positive association between cortisol and AVP. On the other hand a similar study of both in- and outpatients found no association between cortisol and AVP (Goekoop et al., 1997). In the present study we found no correlation between cortisol and copeptin. In favor of the current results is a recent publication found that found no antidepressant effect of a V1b receptor antagonist compared to placebo in two of three trials (Griebel et al., 2012). It should be mentioned, that the ACTH and cortisol levels in the studied population did not suggest a hyperactive HPA axis in this group of depressed patients. It could be argued that our hypothesized association between cortisol and AVP would be the scenario only in individuals with hyperactive HPA-axis. Taken together, it remains unclear if there is an association between cortisol and AVP in depressed patients as a group or in sub-groups of depressed patients. As previously shown, exercise is a potent stimulator of copeptin (Maeder et al., 2010) and AVP release (Takamata et al., 2000) if measured immediately post-exercise (Sanchis-Gomar et al., in press). The primary stimulus of AVP release during exercise is probably increased plasma osmolality due to free water movement out of the vascular space (Takamata et al., 2000). A limitation in the current study was that osmolality was not measured. However, a previous publication found normal to low plasma osmolality in depressed subjects compared to healthy controls (van Londen et al., 1997). Furthermore, none of the participants were diagnosed with renal or cardiac failure known to influence AVP secretion. In contrast to the report by Maeder et al. (2010) investigating a group of patients with known or suspected myocardial ischemia, we did not find that copeptin release during exercise was associated to heart rate, but to resting levels of copeptin and duration of test. The positive association between copeptin levels at VO2max and the duration of the test could reflect increasing osmolality as observed during exercise. Maximal oxygen uptake was not associated to copeptin at VO2max, potentially explained by a poor association between plasma osmolality and VO2max in the current testing situation. At termination of the randomized exercise intervention the participants in the aerobic group had increased their maximal oxygen uptake with 13 percent while the exercise control group had a four percent decline in maximal oxygen uptake (data not shown). It has previously been reported that physical conditioning is associated with a reduction in HPA axis activation in response to a given workload (Luger et al., 1987) and
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
we have previously reported an augmented GH response to acute exercise in depressed subjects after a strength training program (Krogh et al., 2010). In the current study we did not find that copeptin levels at rest or at VO2max were affected by aerobic exercise training. However, restricting our analysis to participants with a high attendance rate we found a significant lowering of plasma copeptin in the aerobic exercise group. This corresponds with findings from exercise studies of heart failure patients (Gademan et al., 2007) and in rat models of heart failure (Lin et al., 2011). The overall participation in training sessions was quite low as could be expected in depressed patients. Therefore, the current study does not allow us to conclude that aerobic exercise training per se is unable to induce changes in the HPA axis, but rather that such potential effect of exercise training is unlikely to occur in a sample of depressed outpatients. This is in line with our previously reported findings that a similar intervention was unable to induce changes in resting levels of growth hormone, prolactin, or cortisol in a similar set-up (Krogh et al., 2010). The ability to draw conclusions from copeptin to AVP relies on the extent plasma copeptin reflects plasma vasopressin. One study have found a correlation of 0.8 between copeptin and vasopressin in healthy volunteers during different states of osmolality, which was supported by a high correlation between copeptin and osmolality (Balanescu et al., 2011). The high copeptin/AVP correlation has previously been found in critically ill patients (Jochberger et al., 2006; Morgenthaler et al., 2006). However, it should be mentioned that at least two studies have found lower copeptin/AVP correlations around 0.3–0.4 in healthy individuals (Jochberger et al., 2006; Morgenthaler et al., 2006). Despite the large number of published studies using copeptin as a marker of vasopressin, an indication of the general acceptability of this approach, the current results should be evaluated in the perspective of the above discussion. To the best of our knowledge this is the first study of copeptin in patients with depression. In conclusion, the present results do not suggest a key-role of copeptin or vasopressin in depression and copeptin is therefore not a candidate as a biomarker of depression.
Role of funding source Funding support for this study was provided by Trygfonden, Nordea Danmark fonden, Aase og Ejnar Danielsens Fond. The study's funding sources had no role in the study design; collection, analysis, or interpretation of data; writing of the report; and decision to submit the paper for publication. The corresponding author had full access to all the data and had final responsibility for the decision to submit for publication.
Conflict of interest All authors report no biomedical financial interest or potential conflicts of interest.
Acknowledgment We are most grateful to the participants and funders of our study. We would also like to thank our interviewers, test staff and the physiotherapists that were responsible for the intervention.
References Andersen, L.B., 1995. A maximal cycle exercise protocol to predict maximal oxygen uptake. Scandinavian Journal of Medicine and Science in Sports 5, 143–146. Balanescu, S., Kopp, P., Gaskill, M.B., Morgenthaler, N.G., Schindler, C., Rutishauser, J., 2011. Correlation of plasma copeptin and vasopressin concentrations in hypo-, iso-, and hyperosmolar States. Journal of Clinical Endocrinology and Metabolism 96 (4), 1046–1052. Bech, P., Bech-Andersen, G., Schütze, T., 1999. Mini Internationalt Neuropsykiatrisk Interview. Dansk version 5.0.0. 〈www.ccmh.dk〉.
289
De Goeij, D.C., Dijkstra, H., Tilders, F.J., 1992a. Chronic psychosocial stress enhances vasopressin, but not corticotropin-releasing factor, in the external zone of the median eminence of male rats: relationship to subordinate status. Endocrinology 131 (2), 847–853. De Goeij, D.C., Jezova, D., Tilders, F.J., 1992b. Repeated stress enhances vasopressin synthesis in corticotropin releasing factor neurons in the paraventricular nucleus. Brain Research 577 (1), 165–168. Gademan, M.G., Swenne, C.A., Verwey, H.F., van der Laarse, A., Maan, A.C., van de Vooren, H., van Pelt, J., van Exel, H.J., Lucas, C.M., Cleuren, G.V., Somer, S., Schalij, M.J., van der Wall, E.E., 2007. Effect of exercise training on autonomic derangement and neurohumoral activation in chronic heart failure. Journal of Cardiac Failure 13 (4), 294–303. Gjerris, A., Hammer, M., Vendsborg, P., Christensen, N.J., Rafaelsen, O.J., 1985. Cerebrospinal fluid vasopressin—changes in depression. British Journal of Psychiatry 147, 696–701. Goekoop, J., de Winter, R., Wolterbeek, R., Wiegant, V., 2011. Support for two increased vasopressinergic activities in depression at large and the differential effect of antidepressant treatment. Journal of Psychopharmacology 25 (10), 1304–1312. Griebel, G., Beeske, S., Stahl, S.M., 2012. The vasopressin V(1b) receptor antagonist SSR149415 in the treatment of major depressive and generalized anxiety disorders: results from 4 randomized, double-blind, placebo-controlled studies. Journal of Clinical Psychiatry 73 (11), 1403–1411. Hamilton, M., 1960. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 23, 56–62. Inder, W.J., Donald, R.A., Prickett, T.C., Frampton, C.M., Sullivan, P.F., Mulder, R.T., Joyce, P.R., 1997. Arginine vasopressin is associated with hypercortisolemia and suicide attempts in depression. Biological Psychiatry 42 (8), 744–747. Jochberger, S., Morgenthaler, N.G., Mayr, V.D., Luckner, G., Wenzel, V., Ulmer, H., Schwarz, S., Hasibeder, W.R., Friesenecker, B.E., Dunser, M.W., 2006. Copeptin and arginine vasopressin concentrations in critically ill patients. Journal of Clinical Endocrinology and Metabolism 91 (11), 4381–4386. Klose, M., Kofoed-Enevoldsen, A., Ostergaard, K.L., 2002. Single determination of plasma ACTH using an immunoradiometric assay with high detectability differentiates between ACTH-dependent and -independent Cushing's syndrome. Scandinavian Journal of Clinical and Laboratory Investigation 62 (1), 33–37. Krogh, J., Nordentoft, M., Mohammad-Nezhad, M., Westrin, Å., 2010. Growth hormone, prolactin and cortisol response to exercise in patients with depression. Journal of Affective Disorders 125, 189–197. Krogh, J., Videbech, P., Thomsen, C., Gluud, C., Nordentoft, M., 2011. DEMO-II trial. Aerobic exercise versus stretching exercise in patients with major depression— a randomised clinical trial. PLoS One 7 (10), e48316. Lin, Q.Q., Lin, R., Ji, Q.L., Zhang, J.Y., Wang, W.R., Yang, L.N., Zhang, K.F., 2011. Effect of exercise training on renal function and renal aquaporin-2 expression in rats with chronic heart failure. Clinical and Experimental Pharmacology and Physiology 38 (11), 179–185. Liu, J., Hu, P., Qi, X.R., Meng, F.T., Kalsbeek, A., Zhou, J.N., 2011. Acute restraint stress increases intrahypothalamic oestradiol concentrations in conjunction with increased hypothalamic oestrogen receptor beta and aromatase mRNA expression in female rats. Journal of Neuroendocrinology 23 (5), 435–443. Luger, A., Deuster, P., Kyle, S., Gallucci, W., Montgomery, L., Gold, P., Loriaux, L., Chrousos, G., 1987. Acute hypothalamic-pituitary-adrenal responses to the stress of treadmill exercise. New England Journal of Medicine 316, 1309–1315. Maeder, M.T., Staub, D., Brutsche, M.H., Arenja, N., Socrates, T., Reiter, M., Meissner, J., Morgenthaler, N.G., Bergmann, A., Struck, J., Mueller, C., 2010. Copeptin response to clinical maximal exercise tests. Clinical Chemistry 56 (4), 674–676. Makino, S., Schulkin, J., Smith, M.A., Pacak, K., Palkovits, M., Gold, P.W., 1995. Regulation of corticotropin-releasing hormone receptor messenger ribonucleic acid in the rat brain and pituitary by glucocorticoids and stress. Endocrinology 136 (10), 4517–4525. Molnar, A.H., Varga, C., Janaky, T., Toth, G., Toth, G., Farkas, J., Laszlo, F., Laszlo, F.A., 2007. Biological half-life and organ distribution of [3H]8-arginine vasopressin following administration of vasopressin receptor antagonist OPC-31260. Regulatory Peptides 141 (1–3), 12–18. Morgenthaler, N.G., Struck, J., Alonso, C., Bergmann, A., 2006. Assay for the measurement of copeptin, a stable peptide derived from the precursor of vasopressin. Clinical Chemistry 52 (1), 112–119. Rabadan-Diehl, C., Lolait, S.J., Aguilera, G., 1995. Regulation of pituitary vasopressin V1b receptor mRNA during stress in the rat. Journal of Neuroendocrinology 7 (12), 903–910. Sanchis-Gomar, F., Bonaguri, C., Aloe, R., Pareja-Galeano, H., Martinez-Bello, V., Gomez-Cabrera, M.C., Candel, J., Vina, J., Lippi, G., 2013. Effects of acute exercise and xanthine oxidase inhibition on novel cardiovascular biomarkers. Translational Research, http://dx.doi.org/10.1016/j.trsl.2013.02.006, in press. Scott, L.V., Dinan, T.G., 1998. Vasopressin and the regulation of hypothalamicpituitary-adrenal axis function: implications for the pathophysiology of depression. Life Sciences 62 (22), 1985–1998. Takamata, A., Nose, H., Kinoshita, T., Hirose, M., Itoh, T., Morimoto, T., 2000. Effect of acute hypoxia on vasopressin release and intravascular fluid during dynamic exercise in humans. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 279 (1), R161–R168. Terzic, D., Johansson-Fällgren, A., Ragnarsson, O., Goetze, J.P., Hammarsten, O., 2012. Evaluation of a sensitive copeptin assay for clinical measurement. Open Clinical Chemistry Journal 5, 21–26.
290
J. Krogh et al. / Journal of Affective Disorders 151 (2013) 284–290
van Londen, L., Goekoop, J.G., van Kempen, G.M., Frankhuijzen-Sierevogel, A.C., Wiegant, V.M., van der Velde, E.A., De Wied, D., 1997. Plasma levels of arginine vasopressin elevated in patients with major depression. Neuropsychopharmacology 17 (4), 284–292. Wang, S.S., Yan, X.B., Hofman, M.A., Swaab, D.F., Zhou, J.N., 2010. Increased expression level of corticotropin-releasing hormone in the amygdala and in the hypothalamus in rats exposed to chronic unpredictable mild stress. Neuroscience Bulletin 26 (4), 297–303.
Zhang, L., Hernandez, V.S., Liu, B., Medina, M.P., Nava-Kopp, A.T., Irles, C., Morales, M., 2012. Hypothalamic vasopressin system regulation by maternal separation: its impact on anxiety in rats. Neuroscience 215, 135–148. Zhao, D.Q., Ai, H.B., 2011. Oxytocin and vasopressin involved in restraint waterimmersion stress mediated by oxytocin receptor and vasopressin 1b receptor in rat brain. PLoS One 6 (8), e23362.