Neuropeptides in the duodenal mucosa of chronic alcoholic heavy drinkers

T. Hauge,*, J. Persson,1 and K. Sjölund

Division of Gastroenterology, Departments of Internal Medicine, Lund and
1 Malmö University Hospitals, Sweden

Received 26 July 2000; in revised form 3 November 2000; accepted 5 December 2000


    ABSTRACT
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Previous studies have shown that patients with chronic alcohol ingestion may show a variety of morphological and functional alterations in the small intestine. In this study, we have focused on the neuroendocrine system in the duodenal mucosa in chronic alcoholics; an area little studied. Twenty-three defined chronic alcoholics admitted to the hospital for detoxification underwent clinical examination, followed by upper gastrointestinal endoscopy and blood tests on average 4 days after the most recent alcohol intake. Biopsy specimens were taken from the distal part of the descending duodenum for both immunohistochemical and routine histological examination. The control group consisted of 25 patients referred for upper endoscopy mainly because of dyspepsia (ulcer, reflux type), but who were otherwise healthy. A normal carbohydrate-deficient transferrin and a history of low alcohol consumption (<40 g/week) were required for inclusion in the control group. The tissue specimens were studied using antisera for the following neuropeptides: cholecystokinin, galanin, gastric inhibitory peptide (GIP), glucagon, motilin, neuropeptide Y, pituitary adenylyl cyclase activating peptide, secretin, serotonin, somatostatin, substance P, vasoactive intestinal polypeptide and protein gene product, as a general marker for neurones and cells of the diffuse neuroendocrine system. The density of nerve fibres was evaluated semi-quantitatively and the number of endocrine cells per unit length of mucosa was counted in sections cut perpendicularly to the mucosal surface. All the different peptidergic nerve fibres in the alcohol group showed higher densities than the corresponding controls. However, this was not a statistically significant difference. A slightly significant increase (P = 0.02) in the numbers of glucagon and GIP cells was seen in the alcohol group. Gastrointestinal symptoms were frequently present (87%) in chronic alcoholics. We suggest that chronic alcohol consumption in man may have a general effect on the peptidergic nerve system and some endocrine cell types in the duodenal mucosa.


    INTRODUCTION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Excessive use of alcohol is very often related to gastrointestinal (GI) symptoms, such as diarrhoea, dyspepsia and nausea. The pathophysiology behind these symptoms is still unclear. Most of the alcohol consumed is absorbed in the stomach and the upper part of the small intestine. However, upper GI endoscopy in subjects with high alcohol consumption is often normal. Nevertheless, more careful examination of the upper GI tract may disclose a variety of morphological alterations as well as changes in bacterial flora, intestinal permeability, absorption and motility (Ray et al., 1989Go; Persson, 1991Go; Feinman et al., 1992Go; Bode et al., 1993Go; Hauge et al., 1997Go; Toyonaga and Iwao, 1998Go).

The GI tract has a well-developed neuroendocrine system, including both endocrine cells and peptidergic nerve fibres. The normal occurrence and distribution of the different endocrine cells and peptidergic nerve fibres are well described (Sjölund et al., 1983Go; Sundler et al., 1991Go).

The neuropeptides have different regulatory functions related to motility, absorption, secretion and blood flow in the GI tract. It is thought that the nervous system and the endocrine cells are closely integrated. It has been suggested that vasoactive intestinal peptide (VIP) relaxes smooth muscle cells and has a vasodilator effect in the GI tract. Substance P is thought to constrict the vascular smooth muscle and reduce gastric mucosal blood flow (Grønbech and Lacy, 1995Go; Hayashi et al., 1996Go; Rekik et al., 1998Go). Antibodies to protein gene product (PGP) have become established as markers for neurons and cells of the diffuse neuroendocrine system.

Studies of the acute effects of alcohol on the stomach (mostly in experimental animals) indicate an involvement of nerve fibres; both VIP and substance P seem to be involved in ethanol-induced gastric damage (Karmeli et al., 1993Go; Hayashi et al., 1996Go). Another study on gastric blood flow in alcohol-induced mucosal injury has shown vasodilatation with an increased blood flow (Grønbech and Lacy, 1995Go). Galanin has been regarded as a modulator of GI motility (Ulman et al., 1995Go; Katsoulis et al., 1996Go). Pituitary adenylyl cyclase activating peptide (PACAP) occurs in both the central and peripheral nervous systems, but little is known about its functional role in the gut. VIP-like actions have been suggested (Uddman et al., 1991Go).

Very little is known about the effects of alcohol on the neuropeptides in the human small intestine, especially in response to chronic alcohol ingestion.

In experimental animals, a relationship between chronic alcohol ingestion and altered VIP activity in the small intestine has been suggested (Jimenez et al., 1992Go). Karmeli et al. (1994) treated rats with intragastrically supplied 96% alcohol and subcutaneous indomethacin; somatostatin was thought to have a protective effect in both alcohol and indomethacin-induced gastric mucosal injury.

In light of this it was considered of interest to study the neuroendocrine system in chronic alcoholics. We have examined the most frequently occurring peptidergic nerves and some immunoreactive cells in duodenal biopsies from defined chronic alcoholics, and compared the findings with a control group.


    MATERIALS AND METHODS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Subjects
Twenty-three well-defined chronic alcoholics (four women and 19 men), consecutively admitted to hospital for detoxification after drinking sprees, were examined as soon as possible by performing clinical examination, upper GI endoscopy and blood tests. The mean age of these subjects was 44 years (range 33–59). The examination was usually performed 4 days (range 2–7) after the most recent alcohol intake. Ten subjects were examined after 1–3 days, eight after 4–5 days and five subjects after 6–7 days. The patients and their records were carefully examined before inclusion in the study. Mean alcohol consumption in the last weeks before admission was 338 g pure alcohol daily (range 102–680), mainly in the form of spirits and beer. Patients with manifest alcohol-related somatic diseases (liver cirrhosis), peptic ulcers, malnutrition, or abuse of other drugs, were excluded. Five alcoholics had a past medical history of gastric and duodenal ulcer, two of acute haemorrhage. Pancreatitis was reported in the past medical history of two patients. In the study group, smoking prevalence was 100%. At the time of the examination, none was being treated with gastric acid inhibitors.

The control group consisted of 25 individuals (five women, 20 men) with a mean age 45 years (range 23–76) referred consecutively for upper endoscopy mainly because of dyspepsia (ulcer, reflux type). All the endoscopies thus had clinical indications. The control subjects were subjectively healthy apart from dyspepsia, which was the reason for performing the endoscopy. A careful history was taken of their alcohol consumption, which was below 40 g/week in all cases. In addition, blood measurements for serum CDT (carbohydrate-deficient transferrin) and GGT (gamma-glutamyl transferase) were performed, a normal CDT being required for participation in this study.

The mean (± SEM) value for CDT in the alcohol group was 41.15 ± 9.75 U/l, compared with 15.32 ± 0.97 U/l in the control group (P = 0.0008). (CDT was not available for 10 individuals in the alcoholic group.)

The study protocol was approved by the local ethical committee, and informed consent was obtained from all patients.

Study design
Both groups of patients were examined by upper GI endoscopy, and biopsies were taken with an Olympus FB 24 K forceps. All the endoscopies were performed after fasting for more than 8 h, and by the same endoscopist (T.H.).

The biopsies were taken from the distal part of the descending duodenum, two for immunocytochemistry, and two for tissue pathology. The tissue specimens were stained with haematoxylin and eosin for routine examination using light microscopy. The villus index was examined in all alcoholics and 11 controls. Villus index was determined by adding the height of the villi to the depth of the crypts (from the bottom of the crypts up to the tip of the villi) and dividing by the depth of the crypts.

Blood samples were taken and analysed using standard routines and equipment on the same day as the endoscopy.

Immunocytochemistry
Fresh specimens were quickly immersed in ice-cold 4% buffered formaldehyde solution. After fixation from 1 to 2 days, the specimens were rinsed repeatedly in sucrose-containing buffer, frozen, and sectioned at 15 µm in a cryostat and processed for the immunocytochemical demonstration of cholecystokinin (CCK), galanin, gastrointestinal peptide (GIP), glucagon, motilin, neuropeptide Y (NPY), PACAP, PGP, secretin, serotonin, somatostatin, substance P, and VIP using the indirect immunofluorescence technique (Coons et al., 1955Go). The sections were exposed to the peptide or amine antiserum for 3 h at room temperature. After thorough rinsing in phosphate buffer, they were incubated with fluorescein-labelled anti-rabbit IgG for 30 min at room temperature. All the solutions contained 0.25% serum albumin and Triton X-100. After another rinsing in phosphate buffer, the sections were mounted in phosphate-buffered glycine and examined in a fluorescence microscope. Controls in the immunocytochemical procedure were run as recommended by Sternberger (1979), and included incubation of sections with peptide antiserum inactivated by addition of antigen in excess (1–100 µg per ml of diluted antiserum).

The tissue specimens were stained with different antisera (Table 1Go).


View this table:
[in this window]
[in a new window]
 
Table 1. Details of the antisera used for immunohistochemistry
 
Evaluation
All specimens were evaluated blindly on two or three separate occasions by the same investigator under a Zeiss epifluorescence microscope with a 50 W mercury lamp as the light source. Three to six sections from each of two specimens from each location and each patient were examined, without essential individual variability. The density of the peptidergic nerves was evaluated semi-quantitatively and graded as: –, absent; +, very few and not visible on every visual field; 1+, few; 2+, moderate; 3+, rich supply; 4+, very rich supply. Arithmetic means of the density of nerve fibres were calculated for each substance, and separately for mucosa and submucosa. Fig. 1Go shows a schematic representation of the semi-quantitative evaluation.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 1. Schematic evaluation of the occurrence and density of the different peptide-containing nerves.

 
To complete the semi-quantitative evaluation with statistics, calculations were made by setting: – = 0; (+) = 0.5; + = 1; +(+) = 1.5; ++ = 2; ++(+) = 2.5; +++ = 3; +++(+) = 3.5; ++++ = 4. The calculations were performed separately for mucosa and submucosa, except for PGP, which was examined in nine alcoholics and nine controls.

The different endocrine cell types were counted per unit length of mucosa (0.32 mm) in sections cut perpendicularly to the mucosal surface and arranged so that the entire mucosal height was visible in each section. Five to ten visual fields in three to six sections from each of two specimens taken from each location and each patient were examined.

Statistics
The results are presented as arithmetic means, means, and standard errors of the mean (SEM). Tests for statistical significance were made using the non-parametric Mann–Whitney two-sample test.


    RESULTS
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Twenty of 23 alcoholics had GI symptoms such as dyspepsia (15/23), diarrhoea (11/23), and nausea (6/23). None in the control group had diarrhoea, and nausea was reported in only one patient. For further details concerning the two groups see Table 2Go.


View this table:
[in this window]
[in a new window]
 
Table 2. Gastrointestinal symptoms, alcohol consumption and carbohydrate-deficient transferrin (CDT) in the different groups
 
Endoscopy of the duodenum was macroscopically normal in all alcoholics; only in one case was mucosal rubor observed. Routine microscopic examination, however, showed inflammation in eight of 23 alcoholics. This was found to be mild in six patients, mild to moderate in one patient, and moderate in one. They were all classified as non-specific, five of them as chronic, characterized by an increase of lymphocytes and plasma cells. These findings were all considered as variations within the normal. In the patient with endoscopic findings of duodenal mucosal rubor, the histological picture was mild to moderate non-specific chronic inflammation. The villus index was 3.16 ± 0.15 SEM in the alcoholic group compared with 3.32 ± 0.20 in the control group (P = NS). There were no significant endoscopic or histological differences between alcoholics and controls.

The semi-quantitative evaluation of the peptidergic nerves is shown in Table 3Go. There was a moderate to rich supply of both substance P and VIP in the mucosa and submucosa, but no significant statistical difference between the two groups. Few or moderate numbers of PACAP-containing nerve fibres were seen in both groups. Only very few nerve fibres containing galanin were seen in the submucosa in the alcoholic group. As regards NPY, we saw no or very few nerves in the mucosa, but more in the submucosa. All the single different peptidergic nerve fibres in the alcoholic group showed higher mean values, but not statistically significantly different from the controls (Fig. 2Go). For nine performed examinations of PGP in alcoholics and controls, there was no difference between the groups (data not shown). Figure 3Go shows pictures of VIP and PGP peptidergic nerves in the duodenum of an alcoholic and a control patient, respectively.


View this table:
[in this window]
[in a new window]
 
Table 3. Semi-quantitative evaluation of the density of the different peptidergic nerves
 


View larger version (75K):
[in this window]
[in a new window]
 
Fig. 2. The mean values for the different peptidergic nerves in alcoholics and controls. The mean values in the alcohol group are all higher than the corresponding controls. m/sm, mucosa/submucosa; NPY, neuropeptide Y; PACAP, pituitary adenylyl cyclase activating peptide; SP, substance P; VIP, vasoactive intestinal peptide.

 


View larger version (91K):
[in this window]
[in a new window]
 
Fig. 3. Protein gene product (PGP) and vasoactive intestinal peptide (VIP) peptidergic nerves in the duodenal mucosa in an alcoholic patient (left side) and in a control patient (right side). The picture does not show a difference between the two groups.

 
In the alcoholic group, the density of VIP was compared in those with diarrhoea (11 patients) and those without (12 patients). The mean value of VIP was higher in the group with diarrhoea, both in the mucosa and submucosa. However, there was no statistically significant difference (data not shown).

If the density for all the different peptidergic nerves in the mucosa and submucosa respectively are seen together, there were obvious statistical differences between the alcoholics and controls (P < 0.0001), with a higher density of peptidergic nerves in both mucosa and submucosa in the alcoholic group (Fig. 2Go).

A similar examination of all the different endocrine cells (seen together) showed a mean value in the alcohol group of 15.5 ± 1.6 cells/unit length compared with 13.9 ± 1.6 cells/unit length in the controls (P = NS). However, there was an increased number of GIP and glucagon immunoreactive cells in the alcohol group (P = 0.02 for both) (see Table 4Go). The mean values for CCK, motilin and somatostatin were slightly but not significantly higher in the alcohol group.


View this table:
[in this window]
[in a new window]
 
Table 4. Number (mean values) of the different endocrine cells/unit length mucosa
 

    DISCUSSION
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This descriptive study of the most frequently occurring peptidergic nerves in the duodenal mucosa shows higher mean values for all the different nerves in alcoholics, compared to controls, both in the mucosa and submucosa. This may indicate a general response to chronic high alcohol consumption. There were no significant differences when evaluating the individual peptidergic nerves separately, but the study was perhaps too small to demonstrate statistically significant differences.

With regard to endocrine cells, there was a significantly increased number of GIP and glucagon cells per unit length of mucosa, which may indicate that alcohol interacts with the glucose–insulin axis in the gut.

In a previous small study from our department, some of the peptidergic nerves and immunoreactive cells were examined in biopsies obtained with a Watson capsule from the distal duodenum and proximal jejunum. In this study, the villus height was increased in five of six patients who accepted a second biopsy after 6 weeks of abstinence. With regard to peptidergic nerve fibres and immunoreactive cells, there was no obvious difference (Persson et al., 1990Go). Gut neuropeptides seem to be involved in different morphological alterations in the small intestine, which can also be related to alcohol, for example an altered microcirculation and motility.

Alcohol abuse leads to altered motility in the small intestine (Persson, 1991Go; Feinman et al., 1992Go; Papa et al., 1998Go). Regulatory functions in intestinal motility have been proposed for at least VIP, galanin, and substance P, and possibly also PACAP (Sundler et al., 1991Go; Rekik et al., 1998Go). A mucosal microvascular stasis is considered to be central in the pathogenesis of alcohol-related morphological alterations in the small intestine (Beck et al., 1986Go; Ray et al., 1989Go). Different pathogenetic mechanisms leading to a disturbance in the villus drainage have been described (Beck and Dinda, 1981Go; Dinda et al., 1988Go). An initial action of alcohol in the disruption of the normal drainage can be related to a direct effect on the microcirculation or by inducing contraction on the villus core (Beck and Dinda, 1981Go). In the case of neuropeptides, substance P and VIP may affect local mucosal blood flow via mast cell mediators (Karmeli et al., 1993Go; Hayashi et al., 1996Go). Increased mucosal levels of both VIP and substance P were also found in the stomach after alcohol ingestion. Somatostatin was able to prevent both an alteration in these neuropeptides and the alcohol-related mucosal injury (Karmeli et al., 1994Go). The present study did not show any difference between alcoholics and controls concerning the number of somatostatin immunoreactive cells.

It is possible that the effects related to chronic alcohol ingestion differ from those found after acute ingestion, and that the small intestine adapts to noxious substances. In our study, upper GI endoscopy was performed 4 days after admission. However, a more recently published clinical study using intragastric application of alcohol and alcoholic beverages did not show any macroscopic lesions in the duodenum on endoscopy (Knoll et al., 1998Go).

In conclusion, we have performed an extensive evaluation of the neuroendocrine system in the duodenal mucosa in a group of chronic alcoholics having high alcohol consumption. The findings suggest that chronic alcohol consumption in man may have a general effect on the peptidergic nerve system and on some endocrine cell types in the duodenal mucosa.


    FOOTNOTES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
* Author to whom correspondence should be addressed at: Division of Gastroenterology, Department of Internal Medicine, Østfold Central Hospital, N-1603 Fredrikstad, Norway. Back


    REFERENCES
 TOP
 FOOTNOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Beck, I. T. and Dinda, P. K. (1981) Acute exposure of small intestine to ethanol. Effects on morphology and function. Digestive Diseases and Sciences 26, 817–838.[ISI][Medline]

Beck, I. T., Morris, G. P. and Buell, M. G. (1986) Ethanol-induced vascular permeability changes in the jejunal mucosa of the dog. Gastroenterology 90, 1137–1145.[ISI][Medline]

Bode, C., Kolepke, R., Schäfer, K. and Bode, J. C. (1993) Breath hydrogen excretion in patients with alcoholic liver disease — evidence of small intestinal bacterial overgrowth. Zeitschrift für Gastroenterologie 31, 3–7.[ISI][Medline]

Coons, A. H., Leduc, E. H. and Conolly, J. M. (1955) Studies on antibody production: a method for the histochemical demonstration of specific antibody and its application to a study of the hyperimmune rabbit. Journal of Experimental Medicine 102, 49–60.

Dinda, P. K., Leddin, D. J. and Beck, I. T. (1988) Histamine is involved in ethanol-induced jejunal microvascular injury in rabbits. Gastroenterology 95, 1227–1233.[ISI][Medline]

Feinman, L., Korsten, M. A. and Lieber, C. S. (1992) Alcohol and the digestive tract. In Medical and Nutritional Complications of Alcoholism, Lieber, C. S. ed., pp. 307–340. Plenum, New York.

Grønbech, J. E. and Lacy, E. R. (1995) Role of gastric blood flow in impaired defence and repair of aged rat stomachs. American Journal of Physiology 269, G737–744.[Abstract/Free Full Text]

Hauge, T., Persson, J. and Nilsson, Å. (1997) Gastroduodenal morphology and related symptoms in chronic alcoholics. Diagnostic and Therapeutic Endoscopy 4, 29–33.

Hayashi, H., Nishiyama, K., Majima, M., Katori, M. and Saigenji, K. (1996) Role of endogenous substance P in ethanol-induced mucosal damage in the rat stomach. Journal of Gastroenterology 31, 314–322.[ISI][Medline]

Jimenez, J., Calvo, J. R., Molinero, P., Goberna, R. and Guerrero, J. M. (1992) Chronic ethanol intake inhibits both the vasoactive intestinal peptide binding and the associated cyclic AMP production in rat enterocytes. General Pharmacology 23, 607–611.[Medline]

Karmeli, F., Eliakim, R., Elimelech, O. and Rachmilewiz, D. (1993) Role of vasoactive intestinal peptide (VIP) in pathogenesis of ethanol-induced gastric mucosal damage in rats. Digestive Diseases and Sciences 7, 1210–1219.

Karmeli, F., Eliakim, R., Elimelech, O. and Rachmilewiz, D. (1994) Somatostatin effectively prevents ethanol- and NSAID-induced gastric mucosal damage in rats. Digestive Diseases and Sciences 39, 617–625.[ISI][Medline]

Katsoulis, S., Clemens, A., Morys-Wortmann, C., Schwörer, H., Schaube, H., Klomp, H.-J., Fölsch, U. R. and Schmidt W. E. (1996) Human galanin modulates human colonic motility in vitro. Characterization of structural requirements. Scandinavian Journal of Gastroenterology 31, 446–451.[ISI][Medline]

Knoll, M. R., Köbel, C. B., Teyssen, S. and Singer, M. V. (1998) Action of pure ethanol and some alcoholic beverages on the gastric mucosa in healthy humans: A descriptive endoscopy study. Endoscopy 30, 293–301.[ISI][Medline]

Papa, A., Tursi, A., Cammarota, G., Certo, M., Cuoco, L., Montalto, M. et al. (1998) Effects of moderate and heavy alcohol consumption on intestinal transit time. Panminerva Medica 40, 183–185.[ISI][Medline]

Persson, J. (1991) Alcohol and the small intestine. Review. Scandinavian Journal of Gastroenterology 26, 3–15.[ISI][Medline]

Persson, J., Berg, N. O., Sjölund, K., Stenling, R. and Magnusson, P. H. (1990) Morphological changes in the small intestine after chronic alcohol consumption. Scandinavian Journal of Gastroenterology 25, 173–184.[ISI][Medline]

Ray, M., Dinda, P. K. and Beck, I. T. (1989) Mechanism of ethanol-induced jejunal microvascular and morphologic changes in the dog. Gastroenterology 96, 345–354.[ISI][Medline]

Rekik, M., Delvaux, M., Frexinos, J. and Bueno, L. (1998) Role of vasoactive intestinal polypeptide in the adaptation of intestinal smooth muscle cells to mechanical distension. Journal of Pharmacology and Experimental Therapeutics 287, 832–838.[Abstract/Free Full Text]

Sjölund, K., Sandén, G., Håkanson, R. and Sundler, F. (1983) Endocrine cells in human intestine: an immunocytochemical study. Gastroenterology 85, 1120–1130.[ISI][Medline]

Sternberger, L. A. (1979) In Immunocytochemistry 2nd edn. New York, Toronto: John Wiley and Sons.

Sundler, F., Ekblad, E. and Håkanson, R. (1991) The neuroendocrine system of the gut—an update. Acta Oncologica 30, 419–427.[ISI][Medline]

Toyonaga, A. and Iwao, T. (1998) Portal–hypertensive gastropathy. Journal of Gastroenterology and Hepatology 13, 865–877.[ISI][Medline]

Uddman, R., Luts, A., Absood, A., Arimura, A., Ekelund, M., Desai, H., Hakanson, R., Hambreaus, G. and Sundler, F. (1991) PACAP, a VIP-like peptide, in neurons of the oesophagus. Regulatory Peptides 36, 415–422.[ISI][Medline]

Ulman, L. G., Potter, E. K. and McCloskey, D. I. (1995) Inhibition of vagally induced gastric contractions by sympathetic stimulation, neuropeptide Y and galanin. Journal of Autonomic Nervous System 55, 193–197.[ISI][Medline]





This Article
Abstract
FREE Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Hauge, T.
Articles by Sjölund, K.
PubMed
PubMed Citation
Articles by Hauge, T.
Articles by Sjölund, K.