Article Text

Download PDFPDF

Presence of the neuropeptide Y1 receptor in tenocytes and blood vessel walls in the human Achilles tendon
  1. D Bjur1,
  2. H Alfredson2,
  3. S Forsgren1
  1. 1
    Department of Integrative Medical Biology, Anatomy Section, Umeå University, Umeå, Sweden
  2. 2
    Department of Surgical and Perioperative Science, Sports Medicine, Umeå University, Umeå, Sweden
  1. Correspondence to Dr S Forsgren, Integrative Medical Biology, Anatomy, Umeå University, SE-901 87 Umeå, Sweden; Sture.Forsgren{at}


Background: Questions remain concerning the mechanisms of the development of chronic pain and impaired function of tendons (tendinosis). Aspects known to occur are cell proliferation, angiogenesis and altered blood flow regulation. Neuropeptide Y (NPY) is widely distributed in the body and has powerful effects in relation to these processes. NPY has its effects via the G-protein-coupled Y receptors. There is no information concerning the presence or absence of NPY receptors in Achilles tendons or other tendons.

Objective: To clarify the expression patterns of the NPY receptors Y1 and Y2 in normal and tendinosis Achilles tendons of humans.

Methods: Immunohistochemical methods were used. Examination on NPY was carried out in parallel.

Results: The tenocytes showed strong immunoreactions for the Y1 receptor. The immunoreactions were more intense in the tenocytes of tendinosis tendons than in non-tendinosis tendons. The rounded/oval tenocytes typically seen in tendinosis tendons exhibited marked Y1 receptor reactions on their exterior. Pronounced Y1 reactions were seen in the smooth muscle of the arterioles of both tendinosis and non-tendinosis tendons. No reactions for the Y2 receptor were noted. NPY was detected in nerve fascicles and in perivascular innervation.

Conclusions: This study shows that there is a morphological correlate for the occurrence of pronounced NPY effects via the Y1 receptor in both tenocytes, especially for tendinosis tendons, and blood vessel walls in the Achilles tendon. The findings are of particular interest as NPY is known to have proliferative, angiogenic and blood vessel-regulating effects. The effects of targeting the Y1 receptor in tendinosis is an interesting task to be evaluated further.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Chronic pain and impaired function in tendons (tendinopathy) is a frequent condition, not least involving the Achilles tendon. When there is thickening and abnormalities in the tendon in parallel, as seen by ultrasound or magnetic resonance imaging (MRI) examination, the condition is referred to as tendinosis.1 The mechanisms behind tendinosis are largely unknown.

Neuropeptide Y (NPY) is a neuropeptide that was characterised in 1982,2 and is mainly present in the neurons of the sympathetic nervous system.3 4 5 However, NPY is also expressed in other cell types such as parasympathetic neurons.6 7 This neuropeptide belongs to the so-called neuropeptide Y family and activates the Y receptors, especially the Y1, Y2 and Y5 receptors,8 9 which are G-protein-coupled receptors. NPY is markedly involved in blood vessel regulation.10 11 The effects of NPY include actions leading to an increase in cell proliferation and angiogenesis.12 13 14 An increase in cell numbers and an increase in the number of blood vessels via angiogenesis15 16 17 18 and disturbed blood vessel regulation1 19 are typical features seen in tendinosis.

There is no information concerning the presence or absence of Y receptors in tendons in either humans or animals. However, given the known effects of NPY, it would be of interest to know whether there are Y receptors in the human Achilles tendon and to what extent they occur in tendinosis. This is of particular interest as targeting Y receptors has been suggested for several conditions.20 21 22 A main feature in most of the conditions for which this has been discussed is the occurrence of large numbers of Y1 receptors in the affected tissue.12 23

In recent studies in our group, the main cells (tenocytes) of the Achilles tendon, especially those in tendinosis Achilles tendons, have been shown to express receptors for nerve signal substances other than NPY, including muscarinic receptors,24 adrenergic receptors25 and the substance P-preferred receptor neurokinin 1.26 27 All these recent observations on receptor expression in tenocytes show that these cells are highly influenced by nerve signal substances. This fact is likely to be of importance for the function of the tendons. It is also possible that NPY is involved in such effects.

One drawback is that there is no information as to whether tenocytes and blood vessel walls in tendons are equipped with NPY receptors. In the present study, therefore, normal and tendinosis Achilles tendons were examined with respect to the existence of NPY receptors, with particular emphasis on the Y1 receptor. Expression patterns of NPY were also evaluated.

Materials and methods


In the present investigation, Achilles tendon (calcaneal tendon) specimens from a total of 45 individuals (mean age 45.4 years, range 21–61) were processed and examined.

The study included two main groups. One consisted of 37 patients (17 men and 20 women; mean age 46.2 years, range 26–61) with a chronic painful condition in the Achilles tendon that was diagnosed clinically, and by ultrasonography or MRI, as tendinosis. All these patients had experienced a long duration of pain symptoms (mean 19 months) and exhibited activity-related pain in a tender, thickened portion of the Achilles tendon, located 2–6 cm cranial (mid-portion) to the tendon insertion into the calcaneal bone. This corresponded to the location of tendon changes verified by ultrasonography (localised tendon widening, irregular structure and focal hypoechoic areas) or MRI (localised widening and increased signal intensity). Sixteen individuals in this group (eight men and eight women; mean age 46.8 years, range 26–61) formed a subgroup. These individuals had previously been treated with sclerosing polidocanol injections three to five times in order to reduce pain. The treatment had had a poor effect and so these particular patients underwent surgery, which was performed several months after the last polidocanol injection.

The other main group consisted of eight individuals (three men and five women; mean age 41.5 years, range 21–47) with pain-free and normal (non-tendinosis) Achilles tendons, as seen by clinical examination and frequently also as verified by ultrasonography.

All the patients included in the study were otherwise healthy, on no medication and were non-smokers.

The study protocol was approved by the Committee of Ethics at the Faculty of Medicine and Odontology, Umeå University, and the Regional Ethical Review Board in Umeå. The experiments were conducted according to the principles expressed in the Declaration of Helsinki.

Sampling, sectioning and fixation

Surgical procedures were performed in strict sterile conditions. A local anaesthetic (Pilokainhydrochloride 4–5 ml, 10 mg/ml; Astra Zeneca, Södertälje, Sweden) was administered to the skin. In both groups a longitudinal incision was then made through the skin and the paratenon. The mid-portion of the tendon was thus visualised. Tissue samples (2 mm in width and 1–5 mm in length) were then taken from the tendon. Sometimes these specimens contained paratendinous loose connective tissue, this tissue often being coalesced with the tendon tissue proper. In the tendinosis group, the tissue samples were taken during surgical treatment. The specimens were taken from the dorsal or ventral side of the tendon. In the non-tendinosis group, the specimens were from the dorsal side of the tendon and were taken with emphasis on minimising tissue trauma.

The specimens were subjected immediately after the surgical procedure to chemical fixation by immersion, overnight at 4°C, in a solution of 4% formaldehyde in 0.1 M phosphate buffer, pH 7.0. Thereafter the specimens were thoroughly washed in Tyrode’s solution, containing 10% sucrose, at 4°C overnight, mounted on thin cardboard in OCT embedding medium (Miles Laboratories, Naperville, Illinois, USA) and frozen at −80°C until sectioning.

Sections, 7 μm thick, were cut using a cryostat, and then mounted on slides pre-coated with crome-alun gelatine. The sections were then dried and processed for immunohistochemistry, using an immunofluorescence technique.

Immunohistochemical procedures

The sections were pretreated with acid potassium permanganate for 2 minutes, a procedure that was used to enhance the specific immunofluorescence reaction sites.28 The samples were then rinsed three times, 5 minutes each, in phosphate-buffered saline (PBS), incubated for 20 minutes in a 1% solution of detergent Triton X-100 (Kebo Lab, Stockholm, Sweden) in 0.01 M PBS, pH 7.2, containing 0.1% sodium azide as a preservative and subsequently rinsed in PBS three times, again 5 minutes each. The sections were then incubated in 5% normal swine serum (for NPY staining) or 5% normal donkey serum (for Y1 and Y2 receptor stainings) in PBS supplemented with 0.1% bovine serum albumin for 15 minutes. This was followed by incubation with the primary antibody diluted in PBS (1 : 500 for NPY antibody and 1 : 100 for the Y1 and Y2 receptor antibodies) in a humid environment, for 60 minutes at 37°C. After this incubation with the specific antiserum, three 5-minute washes in PBS and another incubation in normal swine serum (for NPY staining) or normal donkey serum (for Y1 and Y2 receptor staining) followed, as described above. Incubation then proceeded with the secondary antibody corresponding to fluorescein tetramethylrhodamine isothiocyanate-conjugated swine antirabbit IgG (Dakopatts, Denmark), diluted 1 : 40 in PBS, for 30 minutes at 37°C (for NPY staining) or fluorescein isothiocyanate (FITC)-conjugated AffiniPure donkey antigoat IgG (Jackson ImmunoResearch, Pennsylvania, USA), diluted 1 : 40 in PBS, for 30 minutes at 37°C (for Y1 and Y2 receptor staining). Finally, the sections were mounted in Vectashield microscopy mounting medium after a further three washes in PBS, 5 minutes each.

The sections were examined under a Zeiss Axioscope 2 plus microscope, equipped with epifluorescence optics and an Olympus DP70 digital camera.

Antibodies and control stainings

Rabbit antibody against NPY (PC223L; Oncogene, Boston, Massachusetts, USA) and goat antibodies against Y1 (sc21992; Santa Cruz, Santa Cruz, California, USA) and Y2 (sc14736; Santa Cruz) receptors (R) were used. The antigen for the NPY antibody was a synthetic peptide corresponding to porcine NPY. According to the supplier, the antibody should be reactive in various species including man. The antigen used for the Y1 receptor is a peptide mapping near the C-terminus of the Y1 receptor of human origin, and the antigen for the Y2 receptor is a peptide mapping near the C-terminus of the Y2 receptor of human origin.

For control purposes, the NPY antiserum was preabsorbed with synthetic NPY peptide (sc 115P; Neo MPS, Strasbourg, France; 50 μg/ml antiserum). For the evaluation of reactions obtained with the Y1 and Y2 antibodies blocking peptides were used. Y1 antiserum was thus preabsorbed with Y1 blocking peptide (sc-21992P; Santa Cruz; 20 μg/ml antiserum) and Y2 antiserum was preabsorbed with Y2 blocking peptide (sc-14736P; Santa Cruz; 20 μg/ml antiserum).

Quantitative evaluations and statistics

Semiquantitative evaluations were performed concerning levels of immunoreactions for the Y1 receptor on the tenocytes. Two observers (DB, SF) conducted these assessments independently. The assessments were related to evaluations of the intensities in the immunoreactions. A four-graded (1–4) scale was used, for which grade 1 corresponded to very weak reactions and grade 4 to very strong reactions. Comparisons were thereby made between the entire tendinosis group and the non-tendinosis group, between the two subgroups of tendinosis (patients treated with polidocanol vs patients who had not had that treatment) and between men and women for the entire tendinosis group. Concerning the tendinosis group of patients who had been given polidocanol, specimens of eight patients could be used for evaluations of tenocyte reactions. The other specimens of this group conformed to tissue that mainly consisted of paratendinous connective tissue. Concerning the assessment of the entire tendinosis group, the reactions in 29 tendons could thus be evaluated. Tenocyte reactions in the specimens of all eight tendons of the non-tendinous group could be evaluated.

The Mann–Whitney test was used to compare the groups. The statistical software applied was SPSS 11.0 for Macintosh; p<0.05 was considered significant. The Spearman rank correlation test was applied for age-related correlation analyses.



The specimens were found to consist mainly of tendon tissue proper, the tissue being composed of areas of collagen fibril bundles interspersed by regions of connective tissue. Blood vessels and occasional nerves occurred in these latter regions. Some of the specimens also contained, to varying extents, outer parts of the tendon, ie, paratendinous connective tissue. Large blood vessels and nerve fascicles occurred in these parts.

The tenocytes in the tendon tissue proper of the non-tendinosis tendons showed a typical slender and spindle-shaped form. The tenocytes in the tendinosis tendons frequently exhibited a rounded/oval appearance. There were usually larger numbers of tenocytes in the tendinosis tendons than in the non-tendinosis tendons.

Neuropeptide Y

Immunohistochemical reactions for NPY were seen in the nerve fascicles and to some extent in perivascular nerve fibres (fig 1). Both the immunolabelled nerve fascicles and the immunoreactions in the perivascular innervation were particularly observed in the paratendinous connective tissue. NPY immunoreaction was not detected in tenocytes.

Figure 1

Parts of the wall of an arteriole in a specimen from an Achilles tendinosis tendon processed for neuropeptide Y (NPY) (a). NPY immunoreactions occur in nerve fibres located in the media/adventitia junction of the wall (arrows). The section in (b) is from the same specimen as that in (a) and is a section stained with NPY antiserum that had been preabsorbed with NPY peptide. No specific immunoreactions are seen in (b). Tetramethylrhodamine isothiocyanate ×20. a, bar 25 μm; b, bar 25 μm.

Y1 receptor

Marked immunoreactions for the Y1 receptor were detected within the smooth muscle of the blood vessel walls, but not in the endothelial layer. The same was true for large-sized (figs 2a, 3b and d) as well as small (fig 3a and d) blood vessels. The intensities of the Y1 receptor immunoreactions were equally marked in blood vessels of non-tendinosis tendons as in those of tendinosis tendons, including the tendons that were from patients who had previously been given sclerosing treatment. Y1 receptor immunoreactions were not detected in the nerve fascicles (not shown).

Figure 2

Section of an Achilles tendinosis tendon processed for the Y1 receptor (a). Y1 receptor immunoreactions are seen in the smooth musculature of blood vessel walls (arrows). There are no specific reactions in the endothelial layers (asterisks). No Y1 receptor immunoreactions are observed in (b), which is a section from the same specimen as that in (a) and that had been pre-absorbed with Y1 receptor peptide. FITC ×20. a, bar 100 μm; b, bar 100 μm.

Figure 3

Sections of Achilles tendinosis tendons processed for the Y1 receptor (a, b, d). The section in (c) is a parallel section to that in (b) and had been stained for demonstration of the Y2 receptor. Y1 receptor immunoreactions are seen in the smooth musculature (arrows) but not in endothelial layers (asterisks) of blood vessel walls. No specific immunoreactions are seen in (c). FITC. a, ×10, bar 200 μm; b, c, ×20 bar 60 μm; d, ×20, bar 40 μm.

The tenocytes in the tendon tissue proper showed Y1 receptor immunoreactions (fig 4a) both for non-tendinosis (fig 5a) and tendinosis (figs 4a, 5b and c, 6) tendons. Reactions were seen both in the long and slender tenocytes (fig 5b) and those that showed a rounded/oval appearance (figs 4a, 5c, 6). In these latter tenocytes, which typically were seen in tendinosis tendons but not in non-tendinosis tendons, it was obvious that the reactions occurred at the plasma membrane (figs 5c, 6). The tenocytes thus appeared to be covered with a Y1 receptor-expressing exterior.

Figure 4

Two consecutive sections of an Achilles tendinosis tendon processed for the Y1 receptor (a) and the Y2 receptor (b). In (a), the tenocytes show immunoreactions at cell exteriors (arrows). No specific immunoreactions are seen in (b). FITC ×40. a, bar 50 μm; b, bar 50 μm.

Figure 5

Achilles tendons processed for the Y1 receptor. The section in (a) is from a non-tendinosis tendon, whereas those in (b) and (c) are from tendinosis tendons. Specific immunoreactions are seen in the elongated and slender tenocytes in (a) and (b) (arrows). The reactions are especially marked in (b). In (c), marked immunoreactions are seen in the exteriors of the frequently occurring tenocytes (arrows). Note the pronounced difference in tenocyte appearance in (c) compared with (a) and (b). The tenocytes in (c) show a rounded/widened appearance, which is an appearance frequently seen for tenocytes in tendinosis tendons. FITC ×40. a–c, bar 50 μm.

Figure 6

Section of an Achilles tendinosis tendon processed for the Y1 receptor. This particular tendon was from a patient who had previously been given sclerosing treatment with polidocanol. Tenocytes show marked immunoreactions at their exteriors (arrows). FITC ×63. bar 40 μm.

A semiquantitative evaluation was made concerning the intensities of Y1 receptor immunoreactions for the tenocytes. The immunoreactions for the tenocytes in the tendinosis group were stronger than those in the non-tendinosis group (p<0.01). There was no statistical difference in this respect between tenocytes of men compared with those of women in the tendinosis group. There were no correlations between age and receptor values. A comparison of the intensities of Y1 receptor immunoreactions in tenocytes of the tendinosis patients who had been treated with polidocanol injections with those tendinosis patients who not had been given this treatment was also made semiquantitatively. The intensities of immunoreactions for the Y1 receptor in the tenocytes in the former group (see fig 6) were as marked as those of the non-treated tendinosis patients (see figs 4a, 5b and c). There was no statistically significant difference between the two subgroups.

Y2 receptor

There were no Y2 receptor immunorections in the blood vessel walls (fig 3c), tenocytes (fig 4b) or nerve fascicles.


The Y1 receptor was clearly expressed in the tenocytes in the Achilles tendon. On the other hand, these cells did not show expression of the Y2 receptor. The Y1 receptor immunoreactions were more marked for the tenocytes of tendinosis tendons than for those of non-tendinosis tendons. Expression of the Y1 receptor was also distinctly shown in the cells of the blood vessel walls of all groups.

We did not find immunohistochemical expression of the antigen for the Y1 receptor, ie, NPY, in the tenocytes. The origin of the NPY that affects the tenocytes is thus unclear. However, the fact that we did not find NPY expression in these cells does not imply that absolutely no NPY is produced in the cells. It may be that the production level of the peptide is too low to permit detection with our immunohistochemical methods. In comparison, in a recent study, we did not observe immunohistochemical expression of substance P in the tenocytes of the Achilles tendon, whereas they showed marked expression of the substance P-preferred receptor, the neurokinin 1 receptor.27 In accordance with previous findings for Achilles tendons in humans25 and rats,29 we found NPY immunoreactions in nerve fascicles and in the perivascular innervation. This perivascular innervation was mainly confined to the arteries and arterioles present in the outer parts of the tendon, ie, the paratendinous connective tissue. Similarly, there is scarce sympathetic innervation within the interior of Achilles tendons,25 whereas there is a marked sympathetic innervation in the paratendinous connective tissue.30

The expression pattern of NPY receptors has never previously been investigated in tendons. Nevertheless, the presence of Y1 receptors in blood vessel walls, the receptors being confined to the smooth musculature, is in accordance with the presence of Y1 receptors on the smooth muscle cells in other locations.31 32 Stimulation of the Y1 receptor in blood vessel walls by NPY is known to lead to constriction.14 It is well known that NPY is involved in cardiovascular regulation,10 NPY having a direct vasoconstrictive effect but also potentiating the effects of other substances such as alpha-adrenergic agonists.11

It should be recalled that there is a pathologically high blood flow in Achilles tendinosis, as seen by examination via the colour Doppler technique.1 19 One reason for this may be that the marked presence of adrenoreceptors in the blood vessel walls and the occurrence of a prominent perivascular sympathetic innervation of the vessels seen in tendinosis tendons leads to excessive vasoconstrictive effects and thus the increase in blood flow.33 NPY may also be involved in these effects via a potentiation of the effects of adrenergic agonists. Injections of a sclerosing substance (polidocanol) in the region of the tendon showing this elevated blood flow may have a favourable effect concerning the amelioration of the pain symptoms in Achilles tendinosis.1 The examinations on the specimens of the tendinosis patients who had been given sclerosing treatment several months or more before the operation showed that this treatment had not influenced the Y1 receptor expression pattern in the blood vessel walls. These thus showed equally marked Y1 receptor reactions as the blood vessel walls of the tendons of patients who had not been given sclerosing treatment. Both subgroups of tendinosis patients required operation.

What is already known on this topic

Receptors for certain nerve signal substances are present in the tendon cells (tenocytes) and blood vessel walls of tendons, particularly tendinosis tendons; that includes muscarinic and adrenergic receptors. It is not known if receptors for the neuromodulator NPY are present in these locations.

What this study adds

One of the NPY receptors, the Y1 receptor, was confined to both tenocytes and blood vessel walls. Concerning tenocytes, the presence of the Y1 receptor was particularly seen for tendinosis tendons. In conclusion, there is a morphological correlate for the occurrence of marked NPY effects via the Y1 receptor.

An interesting observation in the present study is the finding of Y1 receptors on the tenocytes. These findings show that there is a morphological correlate for the occurrence of effects of NPY on the Y1 receptors of not only small and large-sized blood vessels but also the Y1 receptors on the tenocytes. As the magnitude of Y1 receptor immunoreactions was especially marked for tenocytes of tendinosis tendons, it is of great interest to note that cardinal features in tendinosis are proliferation of tenocytes and angiogenesis (for references, see the first few paragraphs of this paper). It has thus previously been observed that NPY can have effects on cell proliferation and angiogenesis. The proliferative effects occurring via the Y1 receptor include effects on neuroproliferation.34 35 36 NPY and the Y1 receptor have been found to be of importance in the proliferation process of olfactory precursor cells13 and NPY stimulates retinal neural cell proliferation via the Y1 receptor as well as other NPY receptors.37 Furthermore, NPY promotes proliferation of adipocyte precursor cells38 and prostatic cancer cells.39 The proliferative effects of NPY are known to include effects leading to the stimulation of angiogenesis.12 22 40 Stimulation of the Y1 receptor leads to vascular smooth muscle proliferation.41 42

As the tenocytes are also equipped with adrenergic receptors,25 it can be speculated that a cooperation between NPY and catecholamines occurs concerning tenocyte proliferation. Such cooperation between NPY and catecholamine (noradrenaline) has been shown in the proliferation of cultured human vascular smooth muscle cells.41

There is currently a belief that targeting Y1 receptors may be worthwhile in several conditions. One area in which this is attractive is certain cancers, namely those in which Y1 receptors are overexpressed.23 Y1 receptors are expressed in the cells of a variety of cancers.22 23 The basis for this targeting in tumours is the fact that NPY affects both tumour cell proliferation and angiogenesis.43 Furthermore, Y1 receptor-selective antagonists may be applied to treat vascular remodelling in cardiovascular disease.12 Targeting NPY is also attractive for the potential treatment of obesity,20 as this peptide is highly involved in the regulatory loops in the hypothalamus that control food intake,20 and as treatment with Y1 receptor antagonists reduces obesity experimentally.44

The present study shows that there is a morphological basis for targeting the Y1 receptor in tendons, as both the tenocytes and the blood vessel walls were equipped with this receptor. Nevertheless, there are certain limitations to the present study. Functional studies on the effects of Y1 targeting are thus needed. In this respect, the routes of administration used in other conditions are noteworthy. Previously, intrathecal or intraparenchymal administration into brain regions, as well as oral administration, have been used in experimental studies using Y1 antagonists for obesity treatment.44 In tumour treatment, it has been suggested that the targeting of Y1 receptors should be carried out using NPY analogues coupled with radionuclides or cytotoxic agents for scintigraphic tumour therapy.23 In studies on endothelium-denuded carotid arteries of mice, a Y1 receptor antagonist was applied adventitially, a treatment that reduced the atherosclerotic neointimal area.45

In conclusion, marked reactions for the Y1 receptor were observed in both tendon cells and blood vessel walls. These findings indicate that NPY has pronounced effects concerning both tenocyte function and blood vessel regulation in tendons. The observations imply that tendon function might be influenced via targeting the Y1 receptor.


The authors would like to thank Ulla Hedlund for skilful technical assistance.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 41.
  43. 42.
  44. 43.
  45. 44.


  • Funding Financial support has been given by the Faculty of Medicine at Umeå University, the Swedish National Centre for Research in Sports, the County Council of Västerbotten, the JC Kempe and Seth M Kempe Memorial Foundations, Örnsköldsvik and Magn Bergvalls Stiftelse.

  • Competing interests None.

  • Ethics approval The study protocol was approved by the Committee of Ethics at the Faculty of Medicine and Odontology, Umeå University, and the Regional Ethical Review Board in Umeå.

  • Provenance and Peer review Not commissioned; externally peer reviewed.

Linked Articles