You are here

Article Text

Article menu

Download PDFPDF

Clinical science
Age-related changes of human conjunctiva on in vivo confocal microscopy
  1. Wenqing Zhu,
  2. Jiaxu Hong,
  3. Tianyu Zheng,
  4. Qihua Le,
  5. Jianjiang Xu,
  6. Xinghuai Sun
  1. Department of Ophthalmology, Eye & Ear, Nose Throat Hospital of Fudan University, Shanghai, PR China
  1. Correspondence to Dr Jianjiang Xu, No 83, Fenyang Road, Xuhui District, Shanghai 200031, PR China; xu_heihei{at}hotmail.com

Abstract

Aims To analyse age-related changes of human conjunctiva by in vivo laser scanning confocal microscopy (LSCM).

Methods 80 healthy subjects were enrolled and divided into four groups according to age: A (0–19 years), B (20–39 years), C (40–59 years) and D (over 60 years). In vivo LSCM was performed on the subjects' bulbar conjunctivas, and the images were recorded. The morphology and densities of conjunctival epithelial cells, goblet cells (GCs), dendritic cells (DCs) and the positive rate of conjunctival microcysts were analysed. The diameters of subepithelial fibres in each group were measured as well.

Results The morphology and the densities of both conjunctival epithelial cells and GCs showed no significant age-related differences. The positive rates of conjunctival microcysts were 30% (6/20), 50% (10/20), 60% (12/20) and 75% (15/20) in each group, showing a substantial increase with age (p=0.035). The densities of DCs decreased significantly with ageing (p=0.033). The diameters of subepithelial fibres reduced with age significantly also (p=0.000).

Conclusions The age-related changes of conjunctivas included decreasing densities of DCs as well as the degeneration of subepithelial structures. Even though no significant changes in GCs densities could be seen in the different age groups, the increasing rates of conjunctival microcysts indicated that the cellular function of GCs declined while ageing.

  • Confocal microscopy
  • conjunctiva
  • goblet cells
  • dendritic cells
  • epithelial cells
  • physiology
  • degeneration
  • imaging

http://dx.doi.org/10.1136/bjo.2008.155820

Statistics from Altmetric.com

    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.

    The conjunctiva is a translucent mucous menbrane that covers the inner surface of the eyelids and extends to the limbus on the surface of the globe. It plays a critical role in keeping the integrity of the ocular surface, serving as the source of the tear mucous layer1 and assisting the immunity system.2 Clinical studies have shown that the incidences of some conjunctiva-related diseases such as dry eye,3 conjunctivochalasis4 and Sjögren syndrome5 increase significantly with age. However, whether there are any age-related changes in conjunctiva in its normal physiological state has not been sufficiently proven yet.

    GCs are the primary source of ocular mucin. Any aberrations in the mucin secretion of GCs can disturb the ocular environment.6 Previous results have created opposing views on whether the densities of GCs are actually dependent on age. Marquardt et al7 and Vujkovic et al8 concluded that the number of GCs did not vary significantly with the age, based on their results. On the contrary, Abdel-Khalek et al9 demonstrated that 25% of the subjects who were 79 years and older did show a great reduction in the number of GCs and, at the same time, degenerated GCs or so-called hyaline bodies. These hyaline bodies might represent occluded goblet cells with the unreleased contents as reported by Kessing,10 which were considered to be more common with increasing age. In recent years, more and more attention was paid to certain features that appeared as round or oval empty structures within the conjunctival epithelium under LSCM.11–13 These conjunctival microcysts were similar to the hyaline bodies in morphology and probably contained aqueous humour instead of highly hydrophilic gel-forming mucins.12 However, the findings clarified neither the relationship of conjunctival microcysts and hyaline bodies nor the age-related changes of conjunctival microcysts in normal conjunctivas.

    DCs express lymphocyte costimulatory molecules that migrate to lymphoid organs and secrete cytokines to initiate immune responses.14 As the strongest antigen presenting cell (APCs), DCs play an important role in the immunity of the ocular surface. Dendritic-shaped cells observed by in vivo confocal microscopy are probably APCs, active in the immunity of the ocular surface, and in vivo LSCM is a useful method for evaluating epithelial DCs distribution on the ocular surface.15 The age dependence of DC density has been shown in both the skin16 17 and the tongue.18 Steuhl19 further demonstrated a decline in the density of DCs with ageing in an ex vivo experiment. There has been little literature available, however, on how ageing affects DCs of the conjunctiva in vivo, let alone other portions of the conjunctiva, such as the epithelium and the subepithelial structure.

    Applying LSCM on conjunctivas opened up a new promising method to investigate conjunctiva morphology. In the current study, the conjunctivas of healthy subjects with different age were examined by LSCM to analyse the age-related changes in the morphology and the density of conjunctival epithelial cells, GCs and DCs, and to calculate the positive rates of conjunctival microsysts in the different age groups. The diameters of subepithelial fibres were also measured in each of the subjects.

    Materials and methods

    Study subjects

    Eighty eyes of 80 healthy subjects (35 men and 45 women) were examined in this study. The subjects were divided into four groups according to the following ages: A (0–19 years; 11 men and nine women; average age 14.0±4.1 years), B (20–39 years; eight men and 12 women; average age 30.7±6.6 years), C (40–59 years; nine men and 11 women; average age 47.7±6.3 years) and D (over 60 years; seven men and 13 women; average age 72.5±6.6 years).

    Subject exclusion criteria included a history of ocular trauma or surgery, contact lens wear, current or long-term topical ocular medication, ocular disease or systemic disease that may have affected the conjunctiva. No ocular abnormalities were found after performing routine slit-lamp examinations. Informed consent, according to the tenets of the Declaration of Helsinki, was obtained from all the subjects. This study was approved by the Ethics Committee of Eye & Ear, Nose Throat Hospital of Fudan University.

    In vivo LSCM

    The Heidelberg retina tomograph/Rostock cornea modul (Heidelberg Engineering GmbH, Dossenheim, Germany) was used in this study. It uses a 60× water-immersion objective lens (Olympus Europa GmbH, Hamburg, Germany) and a 670 nm diode laser as a light source, allowing a scanning area of 380 μm×380 μm with lateral and vertical resolutions of both 1 μm and a magnification up to 800 times.

    Before the examination, one drop of 0.4% oxybuprocaine hydrochloride (Benoxil; Santen Pharmaceutical, Japan) and Vidisic gel (0.2% Carbomer 940; Bausch & Lomb, Germany) was applied to the lower conjunctival sac. The patient's head had to be positioned in the headrest, and their eyes had to gaze steadily at the fixation tool. The positions of the superior and inferior bulbar conjunctivae where the images were taken were located 3 mm away from the limbus, while the images of the nasal and temporal bulbar conjunctivae were taken 5 mm away from the limbus. The images were recorded at one point along the z-axis as single scans or in the movie-motion mode.

    Images analysis

    For each conjunctiva, four images were taken at each of the following levels: superficial epithelial cell, intermediate epithelial cell, basal epithelial cell, GC, DC and subepithelial fibre. The first clear images (without motion blur or compression lines) from four observation points were selected to calculate the cellular densities by Cell Count Software (Heidelberg Engineering GmbH) in manual mode. The diameter of the subepithelial fibres was measured at three different points along the same fibres by the Image-J software (http://rsb.info.nih.gov/ij/download.html). The data were expressed as density±SD (cells/mm2). The positive rates of the conjunctival microcysts were then calculated. All images were subsequently randomised and encoded by a single independent observer.

    Statistical analysis

    Statistical analysis was performed in SPSS 11.5 for Windows (SPSS, Chicago, Illinois). Basic descriptive statistics were calculated on all data gathered, and values are reported as mean±SD. One-way analysis of variance (ANOVA) was used to compare the means of four independent groups. The presence of conjunctival microcysts in different age groups was analysed by χ2 test. A p value of less than 0.05 was considered statistically significant.

    Results

    Conjunctiva epithelium and subepithelial structures

    The superficial, intermediate and basal layers of the conjunctival epithelium could be clearly distinguished in vivo by LSCM. Superficial epithelial cells were characterised as large irregular-shaped cells with oval-shaped hyporeflective nuclei (figure 1A), and sometimes hyper-reflective desquamation was visible (figure 1B). The images of intermediate cells were captured with features of small tightly arranged cells with punctiform hyper-reflective nuclei (figure 1C). Basal cells appeared to be polygonal cells within hyper-reflective cell borders with tiny bright nuclei that were sometimes discernible (figure 1D). No significant changes were found in the morphology and the density of conjunctival epithelial cells (table 1). The mean depths of conjunctival superficial, intermediate and basal layers were 5.1 μm, 13.4 μm and 22.5 μm.

    View this table:
    Table 1

    Mean densities of conjunctival epithelial cells

    Figure 1
    Figure 1

    Laser scanning in vivo confocal microscopy of the normal human conjunctiva demonstrating characteristic features. (A) Superficial epithelial cells: large irregular-shaped cells with oval-shaped hyporeflective nuclei (black arrows). (B) Superficial epithelial cells with hyper-reflective-appearing desquamation (black arrows). (C) Intermediate epithelial cells: small tight-arranged cells with punctiform hyper-reflective nuclei. (D) Basal epithelial cells: polygonal cells within hyper-reflective cell borders with tiny bright nuclei discernible or not. (E) Oblique section: conjunctival epithelial cells (black arrow) and conjunctival substantia propria (white arrow) compatible with hyper-reflective basement membrane zone. (F) Conjunctival substantia propria: blood vessels with rich immune cells and hyper-reflective subepithelial fibres. The scale bar indicates 50 μm.

    The basement membrane, a prominent hyper-reflective band, separated epithelial cells from the subepithelial structure (figure 1E). The conjunctival substantia propria, beneath the basement membrane, was mainly composed of irregularly arranged fibres, sharp blood vessels and rich in immune cells (figure 1F). The depth of the conjunctival substantia propria ranged from 25 μm to 136 μm. The mean diameters of the subepithelial fibres were 21±2, 17±2, 14±2 and 8±1 (cells/mm2), respectively, for the four age groups (A, B, C, D), indicating a significant reduction (F=202.224, p=0.000).

    Goblet cell and conjunctival microcysts

    In this study, the GCs were described as large, hyper-reflective, oval-shaped cells with hyporeflective nuclei or with relatively homogeneous brightness, two to three times larger than the surrounding epithelial cells, crowded in groups or mainly dispersed in the layer of epithelial cells. There were no significant morphological changes in GCs in the different groups (figure 2). The mean densities of the GCs did not seem to be age-dependent (p=0.279) (table 2).

    View this table:
    Table 2

    Mean densities of goblet cells and dendritic cells

    Figure 2
    Figure 2

    Laser scanning in vivo confocal microscopy of goblet cells in group A, B, C, D. The scale bar indicates 50 μm.

    Conjunctival microcysts were characterised as giant structures, round to oval in shape, approximately two to three times larger than the normal GCs and containing hyper-reflective contents. Depending on the tangent plane, they could also show either considerable granular brightness or hyper-reflective spots, surrounded by hyporeflective rings (figure 3A–C). The conjunctival microcysts were found in the levels of intermediate epithelial cells, basal epithelial cells or subepithelial structures in 60.47%, 34.88% and 4.65% of the subjects, respectively. The positive rates of conjunctival microcysts significantly increased with ageing (p=0.035) (table 3).

    View this table:
    Table 3

    Positive rate of conjunctival microcysts

    Figure 3
    Figure 3

    Laser scanning in vivo confocal microscopy of conjunctival microcysts and dendritic cells in the normal human conjunctiva. (A) Typical feature, two to three times larger than the normal goblet cells, containing hyper-reflective contents and surrounded by hyporeflective rings. (B) Conjunctival microcysts, which sometimes manifested as a small granular brightness or only hyper-reflective spots, even empty depending on the tangent plane. (C) Goblet cells (black arrow) and conjunctival microcysts (white arrows) in the same image. (D) Dendritic cell bodies lacking dendrites or with small dendrites. (E) Dendritic cell bodies bearing long dendrites. (F) Wire netting pattern of dendritic cells with long entwined dendrites. The scale bar indicates 50 μm.

    Dendritic cell

    The in vivo LSCM images showed that the typical DCs appeared to be hyper-reflective corpuscular particles with dendritic processes, scattered among the conjunctival epithelial cells. However, several kinds of atypical morphologies could be seen also, such as DC bodies lacking dendrites, having small or long dendrites, or the wire netting pattern of DCs with long entangled dendrites (figure 3D–F). The average DC densities indicated a significantly decreasing tendency with increasing age (p=0.033) (table 2).

    Discussion

    Since the research of the conjunctiva morphology using LSCM is still limited, this systemic large-sample study was of great benefit to describing the conjunctiva morphology with LSCM. While age-related changes in cornea20 and limbus21 have been presented in recent years, no reports were available regarding dynamic evolvement of the conjunctiva with age. Therefore, the present study was aimed towards the establishment of baseline normative data of the human conjunctiva and the assessment of age-related changes in human conjunctiva by LSCM.

    Conjunctiva epithelium

    There was no doubt that these features of conjunctival epitheliums were connected to the physiological processes such that the closer epithelial cells were to the surface, the more squamous they were. To a certain extent, the superficial cells rupture and show hyper-reflective desquamation on LSCM. The features of the superficial epithelium and basal epithelium of bulbar conjunctiva were generally accepted.22 There was still controversy about the morphology of the intermediate epithelium, though. Kobayashi et al23 did not describe this layer at all, and Messmer et al22 thought it was the basal epithelium without the rarely visible hyper-reflective nuclei. Our study found that the nuclei of the polygonal cells within hyper-reflective cell borders were hyper-reflective, hyporeflective or invisible. It was thus far-fetched to distinguish between these two layers based on the existence of the cell nuclei. We also demonstrated that cellular features of small tightly arranged cells with punctiform hyper-reflective nuclei were observed frequently by locating the superficial epithelium and hence increasing the scanning depth. The definition of the intermediate epithelium came from the histological classification of the longitudinal section; therefore, it was not completely comparable with the LSCM images, which analyse the transverse section. Based on this study, the intermediate epithelium could be defined as the inward stretch of the superficial epithelium as well as the transitive cellular morphology between the superficial epithelium and the basal epithelium.

    No significant age-related changes were found in the morphology of conjunctival epithelium by LSCM, mainly due to the intercellular space and the microplicae that were considered to have age-related changes by electron microscopy9 being indistinguishable. A significant reduction in the diameters of subepithelial fibres with increasing age was found, corresponding to the clinical phenomenon that conjunctivochalasis occurs in older people more frequently. There were no related results showing how age influenced the conjunctival subepithelial fibre. LSCM provided a new approach to observing the subepithelial fibres and analysing the age-related changes.

    Goblet cell

    Many age-related diseases are closely related to abnormal GC density levels. However, no conclusions were drawn about age-related changes in GC densities in previous studies of subjects below 80 years of age, by either biopsy or impression cytology. Our study indicated that no correlation between the GC density and age using LSCM could be found, which suggested that the differences most likely existed not in the quantity of GCs but in the quality. Kessing10 considered that 50% of the GCs in older people were abnormal and found a phenomenon that frequent occlusion of GCs with retention of their contents increased with age. Abdel-Khalek et al9 reported that hyaline bodies found in 25% of subjects older than 79 years of age were thought to be composed of degenerated GCs, special structures characterised by a central or an eccentric granular mass surrounded by a lucent zone.

    Under LSCM, Labbé et al11 observed conjunctival microcysts in the filtering blebs after glaucoma surgery. Amar et al12 speculated that these microcysts seemed to correspond to goblet cells, mostly containing aqueous humour instead of highly hydrophilic gel-forming mucins. Ciancaglini et al13 did not find conjunctival any microcysts in normal subjects. We found that images of conjunctival microcysts could also be captured by LSCM in the normal bulbar conjunctiva and that the percentage of conjunctival microcysts increased significantly with age, suggesting that the quality of GCs changed with age. Whether conjunctival microcysts are degenerated GCs or normal intermediate products that come with the development and maturation of GCs still remained unclear. However, conjunctival microcysts played a crucial role in clarifying the reason for abnormal tear film function in older people, and it was possible to establish a new criterion for the diagnosis of keratoconjunctivitis sicca.

    Dendritic cell

    The DCs observed by LSCM in the current study were similar to those of previous studies,22 23 which appeared to be hyper-reflective, with either long dendrites or lacking in cell dendrites, presumably indicating mature and immature phenotypes. The morphology of conjunctival DCs observed in the current study provided an insight into immunity reactions in the relationship between the morphology and the function in immune cells. The density of DCs in the conjunctival epithelium ranged from 1 to 300 (cells/mm2), depending on the observation method and the examination area selected.24 Rodrigues et al24 mentioned that fewer DCs could be spotted in older people, while numerous DCs were seen in infant. Steuhl et al19 found a declining tendency in the density of DCs with increasing age by measuring the aforementioned density in human conjunctivae. The current study found significant age-related changes in the density of DCs, which is most likely related to downregulations of the immunity corresponding to advancing age. DCs played a major role in the processing of antigens presentation and carry Class II histocompatibility antigens in the afferent arm of allograft rejection. The observation of the density and distribution of DCs in the conjunctiva enabled an in vivo assessment of cell-induced immunity reactions in ocular disorders. It might also help in revealing the mechanisms of inflammatory diseases of the ocular surface.

    Conclusions

    The DCs had an age-related change in cell morphology and density. The subepithelial fibres were more slender in older people, while increasing numbers of conjunctival microcysts with age may be the critical factor affecting mucin secretion, though there were no significant changes in the density of conjunctival GCs. These data in normal human conjunctivae could serve as a suitable basis for further investigations in ocular pathology and provide a new insight into the pathogenesis of ocular surface diseases.

    References

    1. ↵
      1. Dartt DA
      . Control of mucin production by ocular surface epithelial cells. Exp Eye Res 2004;78:173–85.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Huang LC,
      2. Jean D,
      3. Proske RJ,
      4. et al
      . Innate immunity at the ocular surface: spectrum of antimicrobial peptide expression. Invest Ophthalmol Vis Sci 2005;46(Suppl):883.
      OpenUrl
    3. ↵
      1. Moss SE,
      2. Klein R,
      3. et al
      . Long-term incidence of dry eye in an older population. Optom Vis Sci 2008;85:668–74.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Yumi H,
      2. NorihikoY Aoi K,
      3. et al
      . Age-related changes in the mucocutaneous junction and the conjunctivochalasis in the lower lid margins. Jpn J Ophthalmol 2003;107:363–8.
      OpenUrl
    5. ↵
      1. Whaley K,
      2. Williamson J,
      3. et al
      . Sjögren's syndrome and autoimmunity in a geriatric population. Age Ageing 1972;1:197–205.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Roat MI,
      2. Ohji M,
      3. Hunt LE,
      4. et al
      . Conjunctival epithelial cell hypermitosis and goblet cell hyperplasia in atopic keratoconjunctivitis. Am J Ophthalmol 1993;116:456–63.
      OpenUrlPubMedWeb of Science
    7. ↵
      1. Marquardt R,
      2. Wenz FH
      . Histological studies of goblet cell counts in human conjunctiva. Klin Monbl Augenheilkd 1979;175:692–6.
      OpenUrlPubMed
    8. ↵
      1. Vujkovic V,
      2. Mikac G,
      3. Kozomara R
      . Distribution and density of conjunctival goblet cells. Med Pregl 2002;55:195–200.
      OpenUrlCrossRefPubMed
    9. ↵
      1. Abdel-Khalek LMR,
      2. Williamson J,
      3. Lee WR
      . Morphological changes in the human conjunctival epithelium. I. In the normal elderly population. Br J Ophthalmol 1978;62:792–9.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Kessing SV
      . Mucous gland system of the conjunctiva. A quantitative normal anatomical study. Acta Ophthalmol 1968;95:1–60.
      OpenUrl
    11. ↵
      1. Labbé A,
      2. Dupas B,
      3. Hamard P,
      4. et al
      . In vivo confocal microscopy study of blebs after filtering surgery. Ophthalmology 2005;112:1979.
      OpenUrlPubMedWeb of Science
    12. ↵
      1. Amar N,
      2. Labbe A,
      3. Hamard P,
      4. et al
      . Filtering blebs and aqueous pathway—an immunocytological and in vivo confocal microscopy study. Ophthalmology 2008;115:1154–61.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Ciancaglini M,
      2. Carpineto P,
      3. Agnifili L,
      4. et al
      . Conjunctival modifications in ocular hypertension and primary open angle glaucoma: an in vivo confocal microscopy study. Invest Ophthalmol Vis Sci 2008;49:3042–8.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Zhivov A,
      2. Stave J,
      3. Vollmar B,
      4. et al
      . In vivo confocal microscopic evaluation of langerhans cell density and distribution in the corneal epithelium of healthy volunteers and contact lens wearers. Cornea 2007;26:47–54.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Banchereau J,
      2. Steinman RM
      . Dendritic cells and the control of immunity. Nature 1998;78:173–85.
      OpenUrl
    16. ↵
      1. Gilchrest BA,
      2. Murphy GF,
      3. Soter NA
      . Effect of chronologic aging and ultraviolet irradiation on langerhans cells in human epidermis. J Invest Dermatol 1982;79:85–8.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Thiers BH,
      2. Maize JC,
      3. Spicer SS,
      4. et al
      . The effect of aging and chronic sun exposure on human langerhans cell populations. J Invest Dermatol 1984;82:223–6.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Sasaki M
      . Histomorphometric analysis of age-related-changes in epithelial thickness and langerhans cell-density of the human tongue. Tohoku J Exp Med 1994;173:321–36.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Steuhl KP,
      2. Sitz U,
      3. Knorr M,
      4. et al
      . Age-dependent distribution of Langerhans cells within human conjunctival epithelium. Ophthalmologe 1995;92:21–5.
      OpenUrlPubMed
    20. ↵
      1. Niederer RL,
      2. Perumal D,
      3. Sherwin T,
      4. et al
      . Age-related differences in the normal human cornea: a laser scanning in vivo confocal microscopy study. Br J Ophthalmol 2007;91:1165–9.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Zheng TY,
      2. Xu JJ
      . Age-related changes of human limbus on in vivo confocal microscopy. Cornea 2008;27:782–6.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Messmer EM,
      2. Mackert MJ,
      3. Zapp DM,
      4. et al
      . In vivo confocal microscopy of normal conjunctiva and conjunctivitis. Cornea 2006;25:781–8.
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Kobayashi A,
      2. Yoshita T,
      3. Sugiyama K
      . In vivo findings of the bulbar/palpebral conjunctiva and presumed meibomian glands by laser scanning confocal microscopy. Cornea 2005;24:985–8.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Rodrigues MM,
      2. Rowden G,
      3. Hackett J,
      4. et al
      . Langerhans cells in the normal conjunctiva and peripheral cornea of selected species. Invest Ophthalmol Vis Sci 1981;21:759–65.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Linked articles 161059

    • Funding This work was supported by grants from the Key Clinic Medicine Research Program, the Ministry of Health, China (2007-2009); the Science and Technology Development Fund, Science and Technology Commission of Shanghai Pudong New Area Municipality, PR China (No PKJ2007-Y13); Joint Key Project of New Frontier Technology in Shanghai Municipal Hospitals, Shanghai, PR China (No SHDCI2007104); Shanghai Excellent Leading Scholars Program, Science and Technology Commission of Shanghai Municipality, Shanghai (10XD1401100); Outstanding Doctoral Research Program by Fudan University (2009-2011).

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Ethics approval was provided by the Ethics Committee of Eye & Ear, Nose Throat Hospital of Fudan University.

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

    Linked Articles