Depression and atherosclerosis.

The aim of this paper was to examine the interaction between depression and atherosclerosis through a systematic review. A Medline search was performed from 1966 through 2009 using relevant terms such as depression, epidemiology and atherosclerosis. This was supplemented by a thorough manual search using bibliographies of reviews and full articles. The papers were divided and analyzed separately for each vascular bed. Depression is diagnosed usually before atherosclerosis becomes obvious. The contribution of depression in the development of atherosclerosis emerges from various mechanisms, including lack of physical activity, that exist in this illness. Controversies about the etiology and pathogenesis exist. These interactions of all elements and the importance of each one have not been investigated adequately. Repeated objective measurements for atherosclerosis are lacking. There is an association between depression and atherosclerosis, but the strength of this relationship has to be determined. Prospective studies are needed to determine the early and long term effects of their interaction.

Comorbidity between depression and cardiovascular disease.

Morbidity and mortality of cardiovascular disease (CVD) is exceedingly high worldwide. Depressive illness is a serious psychiatric illness that afflicts a significant portion of the population in all countries. Numerous epidemiological studies have confirmed that high comorbidity exists between these two conditions. Apparently healthy individuals with depression have at least a two-fold higher risk of developing CVD. Following myocardial infarction the emergence of clinical depression poses heightened risk of morbidity and mortality. To understand the complex mechanisms accountable for this comorbidity, several factors have been considered. They include pathophysiologic factors, such as sympathoadrenal activation, homeostatic imbalance between the sympathetic and the parasympathetic systems with diminished vagal tone and loss of heart rate variability in depression. Neuroendocrine factors consist mainly of hypothalamic-pituitary-adrenal axis activation resulting in hypercortisolemia with associated sequelae. Platelet activation and hypercoaguability have been demonstrated in depression and appear to normalize with selective serotonin reuptake inhibitor (SSRI) treatment. Inflammatory processes and release of proinflammatory cytokines have also been described whether or not depression is comorbid with another disease entity. Endothelial dysfunction has been detected in depression and may prove to be a trait marker for this illness. Central and peripheral serotonergic transmission may be one common link between the two disease entities. Comorbid depression must be treated vigorously and SSRIs exert beneficial action not only in ameliorating depression but also in reversing platelet activation and inflammation, thereby reducing cardiovascular morbidity and mortality.

Multimodality imaging in a depressed patient with violent behavior and temporal lobe seizures.

Patients suffering from epilepsy commonly experience behavioral symptoms. Behavioral manifestations are especially prevalent in patients with seizures originating in the limbic system. This case report illustrates how an objective, multimodality work-up can guide the clinician in the diagnosis and the treatment of a patient with a complex presentation. After the discontinuation of some medications, the patient underwent a multimodality work-up that consisted of MRI, SPECT, and conventional and quantitative EEG (LORETA). In this case, the functional imaging studies showed a convergence of findings across the three modalities: MRI, SPECT and qEEG. Because of these findings, we supported more aggressive treatment of the seizure disorder. Ultimately this treatment resulted in resolution of the aggression and the depression. In summary, when applied routinely, a comprehensive, systematic, diagnostic approach will minimize treatment false starts and failures, may reduce costs, and also, potentially decrease the severity and the duration of symptoms.

Relevance of imidazoline receptors and agmatine to psychiatry: a decade of progress.

The cardiovascular relevance of imidazoline receptors (IR) has received tremendous attention since their discovery in 1984. However, evidence also has accumulated for the relevance of IR and an endogenous ligand, agmatine, to psychiatric disease. Emphasis has been placed on altered levels of the I(1)-imidazoline site on human platelets and in human postmortem brain tissue from depressed patients. Attempts at exploring the molecular nature of the I(1) protein have led to the cloning of a protein, IRAS. Based on transfection studies, IRAS seems to be involved in neuronal plasticity events. The I(2) site also appears linked to psychiatric research since some of these sites are localized to a specific domain on monoamine oxidases. Different peptides have been identified by means of an imidazoline-receptor-binding-protein (IRBP) antiserum, and these peptides, some of which appear to be fragments derived from IRAS, undergo changes in platelets and brain commensurate with altered mood states of the subject, notably depressive symptomatology. The search for an endogenous ligand for imidazoline receptor(s) also has led to agmatine, a decarboxylated derivative of arginine. Research on agmatine has mushroomed over the past several years and its measurement in the blood and brain has opened new research opportunities. This novel neurotransmitter interacts with a variety of receptors and has been implicated in mediation of stress responses, analgesia, drug addiction and withdrawal, convulsions, and neuroprotection. Given that IR and agmatine appear involved in a multitude of neurophysiologic and pathologic functions, the potential for new drug development is intriguing.

Atypical [(3)H]clonidine binding sites in human caudate and platelets on cryostat-cut sections.

Pharmacological characterization is described for a human imidazoline binding site (I-site) labeled by [(3)H]clonidine using standard autoradiographic method. Under conditions that mask alpha(2)-adrenergic sites, only a single high affinity site was observed in human caudate and blood platelet sections. Affinity constants (K(i)) were highly correlated between the two tissues (r = 0.90, P = 0.0003). This site is dissimilar to classical I(1) and I(2) sites, even though both tissues possess abundant I(1) and I(2) sites by filtration binding methods. It is suggested that the isotonic buffer conditions inherent to the procedure alter drug affinities to the classical I(1) site.

Reductions in neuronal and glial density characterize the dorsolateral prefrontal cortex in bipolar disorder.

BACKGROUND: Bipolar disorder (BPD) is a mental illness in which depression and mania typically alternate, and both phases can present with psychotic features. The symptomatology of BPD, therefore, resembles major depressive disorder (MDD) and schizophrenia (SCHZ), posing diagnostic dilemmas. Distinct alterations in cellular architecture of the dorsolateral prefrontal cortex distinguish SCHZ and MDD, whereas the cellular neuropathology of BPD has not been studied. METHODS: Dorsolateral prefrontal area 9 was analyzed using a three-dimensional morphometric method in postmortem brains from 10 BPD patients and 11 matched nonpsychiatric control subjects. RESULTS: Area 9 in BPD was characterized by reduced neuronal density in layer III (16%-22%) and reduced pyramidal cell density in layers III and V (17%-30%). A 19% reduction in glial density was found in sublayer IIIc coupled with enlargement and changes in shape of glial nuclei spanning multiple layers. CONCLUSIONS: The morphologic signature of BPD, i.e., decreased neuronal and glial density in association with glial hypertrophy, is distinct from previously described elevations in neuronal density in SCHZ, instead resembling the reductions in cell density found in MDD. Thus, the neuropathologic distinctions between BPD and SCHZ are indicative of separate mental illnesses, each with a unique morphologic disturbance of specific neural circuits.

Imidazoline receptor proteins are decreased in the hippocampus of individuals with major depression.

BACKGROUND: A downregulation of I(2)-imidazoline binding sites has been reported in frontal cortices of depressed suicide victims, according to I(2)-radioligand binding and confirmed by Western blotting. We now report Western blots of imidazoline receptor proteins in hippocampi of subjects with and without depression at the time of death. METHODS: Postmortem diagnoses were obtained from 17 cases of Axis I major depressive disorder and 17 cases without Axis I psychopathology. No psychotropic compounds were found in body fluids. Hippocampi were removed, sectioned, and assessed histologically. Throughout the analysis, each major depressive disorder sample was paired with a sample from a psychiatrically healthy subject based on equivalent life spans and postmortem delays. The antiserum was identical to that used in previous studies that reported a downregulation of cortical 29/30-kd imidazoline receptor-binding proteins in depression. RESULTS: A triad of imidazoline receptor-binding protein bands (40-50 kd) was detected in the human hippocampus. Subjects with major depressive disorder had significantly less intensity in each imidazoline receptor-binding proteins band compared with control subjects (p =. 01 for overall bands). CONCLUSIONS: The present results can be aligned with previous reports of downregulation of I(2)-radioligand binding sites in both cortices and platelets of depressed patients.

Autoradiographic comparison of [3H]-clonidine binding to non-adrenergic sites and alpha(2)-adrenergic receptors in human brain.

UNLABELLED: Clonidine is a partial agonist at brain alpha(2)-adrenoceptors (alpha(2)AR), but also has high affinity (K(D) = 51 nM) in homogenate binding assays for non-adrenergic imidazoline-binding sites (I-sites; imidazoline receptors). Herein, an autoradiographic comparison of [3H]-clonidine binding to I-sites and alpha(2)AR in sections of human brain is reported. For I-sites, the adrenergic component of 50 nM [3H]-clonidine binding was masked with either 60 microM norepinephrine (NE; alpha(2)AR agonist) or 12.5 microM methoxy-idazoxan (MIDX; selective alpha(2)AR antagonist), whereas the remaining non-adrenergic sites were studied by displacement with 20 microM cirazoline. Levels of [3H]-clonidine binding to alpha(2)AR and I-sites, determined in adjacent tissue sections, were positively correlated across 27 brain regions (p = 0.0003; r(2) = 0.385). The principal olivary nucleus and the rostral portion of the ventrolateral medulla had highest ratios of I-sites: alpha(2)AR (>4:1). Quantitative transepts drawn across hippocampal images revealed alpha(2)AR enrichments in the CA-1 and inner molecular layers of the dentate gyrus-areas not enriched in I-sites. Competition curves were generated for I-sites in caudate sections using 10 ligands known to distinguish between I(1) and I(2) subtypes. The rank-order of affinities were cirazoline > harmane > BDF6143 > idazoxan = tizanidine (affinities of agmatine, efaroxan, moxonidine, NE, and oxymetazoline were too low to be reliable). Only the endogenous I-site ligand, harmane, had a monophasic displacement curve at the non-adrenergic sites (Ki = 521 +/- 12 nM). IN CONCLUSION: 1) the distribution of non-adrenergic [3H]-clonidine binding sites in human brain sections was correlated with, but distinct from alpha(2)AR; and 2) the affinities of these sites was distinct from alpha(1)AR, alpha(2)AR, I(1) or I(2) sites as previously defined in membrane binding assays. The properties of this non-adrenergic [3H]-clonidine binding site are consistent with I-sites previously labeled by [3H]-cirazoline in rat brain.

Sucrose and quinine intake by maternally-deprived and control rhesus monkeys.

Clinical depression is often characterized by a loss of interest or pleasure in formerly enjoyable activities. Analogs of anhedonia are established in rats, but the generality of this phenomenon to other species is unknown. Maternally-deprived rhesus macaques show a wide range of behavioral abnormalities that are reversed by chronic antidepressant treatment. We tested consumption by maternally deprived versus control macaques of sweetened (seven sucrose concentrations) or bitter water (four quinine concentrations) versus plain water to evaluate a non-human primate model of depression for signs of anhedonia. All monkeys consumed more sweetened than tap water, but maternally-deprived monkeys had a diminished preference for sweetened water than did controls. However, maternally deprived animals consumed more bitter water than did controls. Baseline fluid consumption did not differ. The data suggest that 'anhedonia' in animal models may be secondary to a generally attenuated responsiveness to stimuli, rather than a unitary reduction in responsiveness to the appetitive properties of stimuli. We conclude that maternally-deprived rhesus monkeys do not display gustatory signs of anhedonia, but rather of insensitivity to gustatory stimuli.

Elevated P-selectin on platelets in depression: response to bupropion.

Increased platelet activation has been suggested as a possible reason for the increased vulnerability of depressed patients to ischemic heart disease (IHD). Translocation of P-selectin, an integral alpha-granule membrane protein, to the platelet surface is a measure of platelet activation. Herein, western blots of platelet plasma membranes containing P-selectin were quantified in patients with major depression (n=19; mean age=39 +/- 2 years) and healthy comparison subjects (n=17; mean age=36 +/- 2 years). None evidenced clinical signs of IHD, and only two patients had a lifestyle IHD risk factor (smoking). Blood was obtained from all 19 depressed patients before treatment, and 15 returned after 6-8 weeks of open-label bupropion treatment. Bupropion was chosen as the antidepressant because it did not elevate plasma norepinephrine or serotonin, endogenous agonists that can induce platelet degranulation. Western blotting revealed more P-selectin immunoreactivity (75 kD band) in depressed patients compared to healthy controls (P=0.003). After bupropion treatment, P-selectin remained high in depressed patients. beta3-Integrin, a reference plasma membrane protein that does not translocate during activation, was of equivalent density in depressed patients and healthy control subjects, and was unchanged after treatment with bupropion. P-Selectin failed to correlate with severity of illness based on the Hamilton Depression scale, or with the post-treatment plasma concentration of bupropion. The results suggest an elevation in P-selectin on platelet plasma membranes might be a trait marker for depression.

Psychoendocrine sequelae of chronic testosterone deficiency.

The precise role of testosterone in regulating mood, especially in alleviating depression, remains unclear although converging evidence indicates that androgens may exert antidepressant action. A model that may potentially assist in the clarification of androgen-mediated effects on mood is the study of cryptorchid men who may grow up with varying degrees of testosterone deficiency depending on the time in their life when cryptorchism is corrected. In this report, we describe a rare case of bilateral cryptorchism that did not come to the attention of the physician to implement effective substitution with testosterone until much later in adult life. The patient developed severe and suicidal depression which responded solely to testosterone. In addition, the patient experienced a delayed but accelerated puberty without any adverse events. These observations, although based on a single case, provide strong evidence that testosterone may exert powerful antidepressant action in the absence of concomitant antidepressant agents.

Effect of bupropion on immunodensity of putative imidazoline receptors on platelets of depressed patients.

A substantial number of studies have demonstrated increased imidazoline receptors (I1 binding sites) on platelets of depressed patients and downregulation following antidepressant treatments. Herein, imidazoline receptor binding protein (IRBP) antiserum was used to quantify imidazoline receptors on platelets of depressed patients before and after treatment with the atypical aminoketone antidepressant, bupropion. Western blots revealed an increase in IRBP-immunodensity (p = 0.01, two-tailed) in a 33 kDa protein band in untreated depressed patients (n = 21) as compared with controls (n = 17). This band has been positively correlated with I1 binding sites on platelets. Following 6 weeks' treatment with bupropion, IRBP-immunodensity was downregulated in depressed patients (p = 0.03, paired t-test); predominantly in responders (p = 0.005). Patients non-responsive to bupropion (n = 5) were significantly different from responders (p = 0.05) by exhibiting no elevation in IRBP-immunodensity at pre-treatment and no downregulation of the 33 kDa band after treatment. IRBP-immunodensity was negatively correlated (r = -0.79, p = 0.01) with plasma concentrations of bupropion and its metabolites at week-4 of BUP treatment. Thus, a 33-kDa IRBP on platelet plasma membranes is elevated in depression and normalized in responders to bupropion.

Plasma agmatine and platelet imidazoline receptors in depression.

Plasma agmatine concentrations are elevated significantly in depressed patients compared to healthy controls. Treatment with the antidepressant bupropion normalized plasma agmatine levels. Correlational evidence is presented that a change in plasma agmatine levels may lead to similar changes in platelet I1 imidazoline receptors.

Alpha 2-adrenoceptors and I1-imidazoline binding sites: relationship with catecholamines in women of reproductive age.

A comparison is presented between plasma catecholamine concentrations and platelet [125I]-p-iodoclonidine binding sites in 16 healthy women. Blood samples were obtained at six regularly spaced intervals over two consecutive menstrual cycles from healthy women with regular menstrual periods. Although no cycle-related changes were observed per se, there were significant correlations between the platelet binding sites and plasma norepinephrine and epinephrine concentrations. The densities of platelet alpha 2-adrenoceptors were negatively correlated in an exponential fashion (r2 = 0.694, P = 0.009) with plasma epinephrine concentrations, implying agonist-induced downregulation. On the other hand, platelet I1-imidazoline binding sites were positively correlated with plasma concentrations of norepinephrine in a linear fashion (r2 = 0.326, P = 0.021). This is the first indication that I1 binding sites might be upregulated by a physiological factor. Furthermore, the data suggest that elevations in plasma norepinephrine might explain reports of upregulated I1 binding sites in depressed patients.

Plasma MHPG response to yohimbine treatment in women with hypoactive sexual desire.

The use of yohimbine to treat impotence has suggested that decreased male sexual desire may relate to decreased activity of central noradrenergic neurons. Previous trials of yohimbine to treat female sexual problems are not available. Yohimbine is an alpha 2-adrenergic antagonist that stimulates norepinephrine (NE) release. In the present study, plasma concentrations of 3-methoxy-4-hydroxyphenylglycol (MHPG), the major central nervous system metabolite of NE, were measured in 9 women diagnosed with hypoactive sexual desire. Daily logs of mood and sexual activity, and trimonthly MHPG blood drawings, were obtained over an initial baseline menstrual cycle followed by two subsequent treatment cycles (yohimbine or placebo), in randomized order. Blood samples were obtained at 9:00 a.m. during (a) the early follicular phase of each cycle (24 hr after the onset of each cycle), (b) the ovulatory phase (i.e., within 1 day of an oral temperature rise), and (c) the midluteal phase (i.e., 20-25 days into each cycle). Comparisons were made with a group of 7 healthy female controls. Women with hypoactive sexual desire had slightly lower plasma MHPG values than controls at baseline, although there was only a trend toward significance during the early follicular phase (p = .09). Yohimbine (5.4 mg orally, 3 times daily, beginning at menses) caused a sustained rise in plasma MHPG of similar magnitude to that reported in men. However, in terms of improved sexual desire, yohimbine had no obvious therapeutic effect. Thus, plasma MHPG and the alpha 2-adrenergic response to yohimbine appeared within normal ranges in women with hypoactive sexual desire, with no therapeutic response to yohimbine.

Determination of agmatine in brain and plasma using high-performance liquid chromatography with fluorescence detection.

Decarboxylated arginine, agmatine, is a neurotransmitter candidate for imidazoline receptors. A method is described to measure agmatine in rat brain and human plasma by isocratic high-performance liquid chromatography (HPLC) with fluorescence detection and o-phthalaldehyde derivatization. Quantitation is based on the method of additions of internal agmatine spikes. This assay has sensitivity in the low picomole range and a detection limit of 100 fmol. The correlation coefficient for the agmatine standard curve was 0.999+/-0.001 S.D., and intra- and inter-assay C.V.s were less than 8%. The accuracy of our isocratic method compared favorably with a gradient HPLC protocol, originally developed for bacterial agmatine, which we modified for use with tissues. Agmatine concentrations in rat brain were proportioned similarly to the regional distribution of imidazoline-1 receptors. These methods can be used as reliable research tools in various biological samples.

Chronic imipramine treatment downregulates IR1-imidazoline receptors in rat brainstem.

One subtype of imidazoline receptors (IR1) is similar to alpha 2-adrenoceptors (alpha 2 AR) based on their high affinity for clonidine and related imidazoline compounds. On the other hand, IR1 possess low affinity for norepinephrine (NE) and other catecholamines. Imidazoline receptors have also been found to be over-expressed in plasma membranes from platelets and brain tissues of depressed patients. Over-expression of IR1 in platelet membranes of depressed patients became normalized after various antidepressant treatment to the patients. Herein, the prototypic antidepressant, imipramine (IMI), has been studied in regard to its treatment effects on [125I]p-iodoclonidine binding to both alpha 2 AR and IR1 in rat brainstem membranes. No effects of chronic IMI treatment were found on brainstem alpha 2 AR binding sites (Bmax and/or KD parameters unchanged) after 25 days of daily injections (i.p. IMI 20 mg/kg/day). However, IMI induced a decrease in the density (Bmax measured under NE mask) of brainstem IR1 sites, with no change in KD. Downregulation of IR1 sites was dose-dependent (minimal effective dose of i.p. IMI was 10 mg/kg/day) and time-dependent (> 16 days of treatment). These results implicate brainstem IR1 in the chronic effects of antidepressants.

A 2-year study of sertraline in the treatment of obsessive-compulsive disorder.

The present study investigated the tolerability, safety profile, and anti-obsessional efficacy of sertraline, a selective serotonin reuptake inhibitor, during long-term treatment of patients with obsessive-compulsive disorder (OCD). Fifty-nine OCD patients who had completed a 1 year double-blind, fixed dose study comparing sertraline and placebo subsequently entered a 1-year open extension. Among the 51 patients who had been treated with sertraline during the double-blind phase, the mean total duration of sertraline treatment was 690 days. Only treatment responders who completed the 52-week double-blind treatment phase were permitted to enter the open extension. The higher rate (p < 0.02) of sertraline patients (51 out of 241) than of placebo patients (eight out of 84), who responded to treatment and entered the open-label phase is therefore consistent with the greater mean improvement observed in the sertraline group during double-blind treatment. Placebo responders differed from sertraline responders in that they were less impaired at baseline of the double-blind study [Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) of 18.5 versus 23.4] and they exhibited less improvement during double-blind treatment (-6.1 versus -11.4). In the open-label phase all patients received sertraline at a starting dose of 50 mg once a day, titrated in 50 mg increments to a maximum dose of 200 mg according to clinical response. At end-point the mean Y-BOCS score for all patients decreased by a further 3.6 points. Patients previously treated with placebo showed greater improvement after being switched to sertraline than those who received continued sertraline treatment. Patients who completed the study and received 2 full years of sertraline treatment (n = 38) exhibited a mean improvement of 15.6 points using the Y-BOCS. Sertraline was well tolerated during both the double-blind phase and the open extension, and the incidence of adverse experiences was generally reduced during the second year of treatment. Three patients discontinued open treatment because of adverse experiences. Long-term sertraline treatment did not appear to be associated with the emergence, increased incidence, or increased severity of adverse experiences or clinically significant abnormalities in laboratory tests, vital signs, or the electrocardiogram. The study supports the long-term safety and tolerability of sertraline over a 2-year treatment course and the sustained efficacy of sertraline in patients with OCD.

Platelet I1-imidazoline binding sites are decreased by two dissimilar antidepressant agents in depressed patients.

Previous studies have indicated that there may be a dysregulation of alpha 2-adrenoceptors and imidazoline receptors in depression. This study compares the effects of chronic antidepressant treatment with a serotonin reuptake inhibitor (fluoxetine) versus a noradrenaline reuptake inhibitor (desipramine) on the binding parameters of the platelet imidazoline binding site (subtype I1) and of the platelet alpha 2-adrenoceptor in depressed patients. After 6 weeks of treatment with either antidepressant, platelet I1 binding sites became normalized (i.e. downregulated). A negative correlation was obtained between plasma epinephrine concentrations and platelet alpha 2-adrenoceptor Bmax values within the samples, but no correlation was obtained between any plasma catecholamine and a platelet I1 binding parameter. An additional finding was the increased affinity of alpha 2-adrenoceptors for p125I-clonidine in untreated depressed patients compared to healthy subjects. Because of the density of platelet I1 binding sites was downregulated by both of the antidepressants, we postulate that a decrease in platelet I1 binding site density may be related to an improved state from depression that these antidepressants produce.

Platelet I1-imidazoline binding sites are elevated in depression but not generalized anxiety disorder.

Depressed patients have been reported to have a higher than normal density of platelet binding sites for 3H-clonidine, an alpha 2-adrenoceptor agonist. Paradoxically, other studies using 3H-alpha 2, antagonists have found no differences from controls. Because 3H-clonidine interacts with platelet alpha 2-adrenoceptors to form G-protein complexes, whereas 3H-alpha 2-antagonists bind with uncoupled receptors, an elevation in G-protein coupling might explain this paradox. Another possibility is that depression might be associated with increased non-adrenergic I1-imidazoline binding sites, which are also clonidine sensitive. To distinguish these possibilities, we utilized p125I-clonidine to measure density (Bmax) and affinity (KD) of platelet G-protein coupled alpha 2-adrenoceptors as well as platelet I1 binding sites, and compared diagnostic groups of major depressive disorder (MDD), generalized anxiety disorder (GAD) and healthy subjects. Specific inhibition of binding by norepinephrine (NE = 10 microM) was used to selectively quantify alpha 2-adrenoceptors, whereas inhibition by 10 microM moxonidine (a > 100-fold selective I1 ligand) quantified I1 binding sites under a NE mask. I1 sites were found to be markedly elevated by, on average, +136% in MDD patients (p = .0007), whereas there was only a marginal increase in alpha 2-adrenoceptor Bmax values in MDD patients (p = .08; GAD and healthy subjects did not differ). Treatment of MDD patients for 6-8 weeks with desipramine downregulated I1 sites as well as alpha 2-adrenoceptors. Positive correlations were also noted for both sites: (a) between Bmax values and the severity of depression (using the Hamilton Depression Rating Scale); and (b) between end-of-treatment plasma desipramine concentrations and the extent of downregulation in Bmax values when subject groups were pooled. None of the binding parameters was associated with plasma catecholamine concentrations. The results suggest that an increased density of platelet I1 binding sites may partially explain the utility of radiolabeled clonidine as a potential biological marker for depressive illness, although an additional increase in G-protein coupling cannot be excluded.