Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Placenta. Author manuscript; available in PMC 2011 January 1.
Published in final edited form as:
PMCID: PMC2808440

Decreased Adherence and Spontaneous Separation of Fetal Membrane Layers--Amnion and Choriodecidua—a Possible Part of the Normal Weakening Process


The fetal membrane (FM) layers, amnion and choriodecidua, are frequently noted to have varying degrees of separation following delivery. FM layers normally separate prior to rupture during in vitro biomechanical testing. We hypothesized that the adherence between amnion and choriodecidua decreases prior to delivery resulting in separation of the FM layers and facilitating FM rupture.


FM from 232 consecutively delivered patients were examined to determine the extent of spontaneous separation of the FM layers at delivery. Percent separation was determined by the weight of separated FM tissue divided by the total FM weight. Separately, the adherence between intact FM layers was determined. FM adherence was tested following term vaginal delivery (13), term unlabored cesarean section (10), and preterm delivery (6).


Subjects enrolled in the two studies had similar demographic and clinical characteristics. FM separation was present in 92.1% of membranes. Only 4.3% of FM delivered following spontaneous rupture of the fetal membranes (SROM) had no detectable separation. 64.7% of FM had greater than 10% separation. FM from term vaginal deliveries had significantly more separation and were less adherent than FM of term unlabored, elective cesarean section (39.0 ± 34.4% vs 22.5 ± 30.9%, p=.046 and 0.041 ± 0.018 N/cm vs 0.048 ± 0.019 N/cm, p< .005). Preterm FM had less separation and were more adherent than term FM (9.95 ± 17.7% vs 37.5 ± 34.4% and 0.070 ± 0.040 N/cm vs 0.044 ± 0.020 N/cm; both p< .001).


Separation of the amnion from choriodecidua at delivery is almost universal. Increased separation is associated with decreased adherence as measured in vitro. Increased separation and decreased adherence are seen both with increasing gestation and with labor suggesting both biochemical and mechanical etiologies. The data are consistent with the hypothesis that FM layer separation is part of the FM weakening process during normal parturition.


Rupture of the fetal membranes (FM) typically follows the onset of uterine contractions; however, in up to 10% of term births, premature rupture of FM (PROM) precedes the onset of contractions. In preterm births, pPROM is the sentinel event in approximately one third of cases and is thus the proximate cause of significant neonatal mortality and morbidity.

The mechanisms by which term or preterm FM weaken and rupture are not completely understood. Previously, we have shown that term FM weaken prior to the onset of labor [1] and that weakening is not solely due to the acute mechanical forces of labor contractions, but rather the result of a biochemical process characterized by apoptosis and extracellular matrix remodeling [1,2]. Furthermore, weakening of term FM is not homogenously distributed across the entire surface, but is more localized to the area that overlies the uterine cervix [1, 3]. This physiologic “weak zone” is present prior to the onset of labor in the third trimester, as demonstrated from studies of term FM from elective, scheduled, unlabored cesarean sections as well as from spontaneous vaginal delivery (SVD) specimens [1,3]. The acute stretch forces of contractions during the normal labor process may secondarily lead to the rupture of the biochemically weakened FM.

We have observed that layers of the FM (amnion and choriodecidua) are frequently partially separated at delivery. Significant separation has frequently been a reason for elimination of specific FM from research studies, particularly biomechanical studies, by our group and others. Often this separation is in the region of the weak zone. Additionally, during in vitro experiments designed to explore the FM rupture process, we video-documented consistent separation of the amnion from the choriodecidua layers occurring prior to rupture [4]. In the same report we demonstrated that the work [energy] required to rupture FM pieces with fused amnion and choriodecidua is significantly greater than the sum of the work required to rupture adjacent, manually-separated pieces of individual amnion and choriodecidua [4]. Thus part of the work necessary to rupture normal, adherent FM is expended during the separation of FM layers and this separation of the FM reduces the amount of force necessary to rupture the membranes.

Meinert et al have also reported FM separation and presented specific evidence that changes in the proteoglycans, decorin and biglycan, and especially increases in the glycosaminoglycan, hyaluronan, at the interface of the amnion and choriodecidua contribute to this process [5, 6]. They suggested that the changes in decorin and biglycan may result in collagen fiber disorganization and that hyaluronan may absorb water, with resultant increases in tissue pressure, causing FM separation and weakness.

Based upon these observations, we hypothesized that FM separation is a part of the FM weakening process facilitating rupture of membranes. We further hypothesized that if separation of the layers of FM were part of the normal FM weakening process, a mechanism for gradual weakening of the tissue bonds fusing the amnion and choriodecidua, like that proposed by Meinert et al [6], must exist which would decrease the adherence between the amnion and choriodecidua at the end of gestation. Physical principles dictate that separation of the FM layers should occur when mechanical shear forces exceed the adherence, as determined by these tissue bonds. Thus separation would become more likely as adherence decreases. The aims of this study were to determine the degree of separation of FM layers in normal deliveries and, secondly, to relate the degree of separation to the strength of adherence between the FM layers. To facilitate the latter, we have developed and reported an in vitro testing methodology to measure the adherence between intact FM layers [7]. If our hypothesis that FM separation is a normal physiological process is correct, FM from clinically defined groups (i.e. Cesarean section delivery without labor, AROM, SROM, preterm and term) which show greater separation should show parallel decreases in adherence.


Tissue Collection

In Vivo Study- Determination of the normal degree of Separation of Amnion and Choriodecidua

This study was approved by the Institutional Review Board of the MetroHealth Medical Center. MetroHealth Medical Center is a referral center with approximately 3600 annual births, a significant proportion of which are high risk and/or premature births. Following delivery, all placentae with attached FM are routinely stored in a cold storage facility for one week. Placentae are generally placed in cold storage within 30 to 60 minutes of delivery. During two study periods of eight weeks duration, during each of the summers of 2006 and 2007, FM from all deliveries were considered for study. Two procedures were followed: 1) For deliveries occurring between 8 am to 5 pm, FM were collected shortly after delivery and examined immediately; 2) For deliveries occurring between 5 pm and 8 am, FM were collected and processed the following morning. FM delivered more than 24 hours earlier were excluded. As our objective was to study normal tissue, placentae from labored cesarean sections, with clinical chorioamnionitis, pre-eclampsia, IUGR, FM of multiples, those earmarked for pathological examination for clinical management, and those with gross placental anomalies were excluded. It is not the practice of our pathology service to routinely examine all placentas, thus histological information is not available for FM in the in vivo study.

In Vitro Study-Adherence Testing

To determine the force of adherence between FM amnion and choriodecidua, we utilized our previously published equipment and protocol to perform standard T-peel testing [7]. FM (n=29) were collected for these studies as follows: term vaginal delivery (n=13: SROM, n=7; AROM, n=6), term unlabored cesarean section (n=10), preterm vaginal delivery (n=5: SROM, n=3; AROM n=2) and indicated preterm unlabored cesarean section (n=1). The same exclusion criteria were used as for the separation study. In addition, FM with histological chorioamnionitis were also excluded. The FM were processed, as described below, within 20 minutes following delivery.

Quantification of Spontaneous FM Separation

FM were cut from the placental disc, washed with Hanks Balanced Salt Solution, removed of blood clots, and then laid flat with amnion side up. Membranes were visually examined for areas of obvious spontaneous separation by teams of two investigators (DK, PS, JS, KB, JN) working together. Separation was usually evident by differences in appearance and translucency between separated versus adherent areas of FM. If an obvious plane of separation was determined and confirmed by both investigators, the amnion was gently manually lifted until the amnion could no longer be lifted without the choriodecidua also being pulled up (Figure 1A). All areas where the amnion was separated from the choriodecidua were then excised. Hanks solution was gently blotted off the FM prior to weighing. For each placenta, the separated areas (amnion plus choriodecidua) were weighed together. The remaining adherent membranes were then weighed. The proportions of separated FM were reported by dividing the weight of the separated FM by the total weight of the combined (separated plus adherent) membranes.

Figure 1
A. Determination of Percent Spontaneous FM Separation; B. FM in T-peel adherence testing apparatus

This methodology assumes that equal surface areas of FM fragments of the same patient have equal weight. FM are known to be heterogeneous, however. Specifically, FM in the area of the weak zone are known to have decreased decidua which may decrease fragment weight [1, 8]. The weak zone may also have increased hylauronan with absorbed water which may increase fragment weight [6]. We have compared equal size FM fragments cut from the separated and non-separated regions of the same placenta and found differences in weight/surface area of less than 5% (data not shown).

Effect of Refrigeration Storage Time upon FM Separation

As not all placentae were examined immediately after birth, we determined the possible effects of delay in examination and refrigeration upon spontaneous FM separation. Six FM with varying degrees of spontaneous FM separation were obtained immediately following delivery. Gentian violet was used as a dye to mark along the line of separation between amnion and choriodecicua. The marked FM were then placed in a refrigerator (−4°C). Placentae were re-examined by two investigators every 8 hours to assess for progression of the spontaneous FM peel front. Subsequent observation for up to 24 hours did not show progression of the line of separation between the amnion and choriodecidua. Separately, the percentage of spontaneous FM separation was evaluated in 20 fresh FM less than one hour after delivery and compared with the extent of spontaneous FM separation in 200 refrigerated FM, one to less than 24 hours old. No difference in degree of spontaneous FM separation [fresh (36.1 ± 32.7%) versus refrigerated (37.2 ± 34.1%); p=0.88] was found.

Determination of Adherence between Amnion and Choriodecidua

FM used for the adherence study were obtained fresh from the delivery room and immediately taken to the laboratory for study. The force of FM adherence was determined using our published T-peel testing protocol [7]. Briefly, the entire FM was cut from the placental disc. Membranes were gently washed with refrigerated Minimum Essential Medium with Earles Salts (MEM: Mediatech, Manassas, Va) and all blood clots were removed. Membrane pieces were kept moist with MEM throughout the peel testing process. Once removed from the disc, FM were laid flat in MEM. A series of cuts were made starting from the placental disc edge and extending to the incision/tear line. This allowed the FM to lay completely flat. Serial secondary cuts were made along a line extending from the disc to the incision/tear line (cesarean section/vaginal delivery) to obtain tissue test pieces (4 × 6 cm). Additional FM sections, parallel and adjacent to this group, were cut maintaining similar orientation. Tracing paper was used to document the orientation of FM to the placental disc edge as well as the incision/tear (cesarean section/vaginal delivery) line. A 4 × 6 cm cork sheet template placed gently over the FM guided the placement of a 4 cm microtome blade to obtain uniform sized FM pieces. This allowed 9 to 15 test samples per FM. Location of samples was mapped on the tracing paper to maintain sample orientation and samples were stored in MEM and refrigerated until testing. These FM test strips were used to determine individual and average force of FM layer adherence for each placenta. Commercial industrial tensile testing equipment (701 Universal Test System, Com-Ten Industries, St. Petersburg, Fl) was adapted to perform a standard engineering T-peel test (ASTM D1876) to measure FM adhesive force with a sensitivity of 0.05 mg. Each 4 × 6 cm piece was loaded in a customized clamp system on the T-peel tester where the lower clamp is held stationary while the upper clamp, attached to the load cell, is driven vertically (peel speed 25.4 cm/min) by the peel tester, peeling the amnion from the choriodecidua (Figure 1B, 1C). Average peel force per individual sample tested was determined from the force (N) versus displacement (cm) curve. Force of adherence (N/cm) is reported as peel force per sample divided by the undeformed sample width of the tested FM fragments (4 cm). Average adherence force for each FM was determined by dividing the sum of the average forces of all samples by the number of all samples for that FM (average of 12 pieces per FM).

Clinical Data Collection and Analysis

Maternal records and the obstetric summary were reviewed for each patient. For the purpose of this study, the time of initial patient assessment by labor and delivery nursing staff in the triage room was taken as the admission time. Comparisons among groups were analyzed using ANOVA and the Student’s t-test on Statview (Version 5.0.1). Differences were considered statistically significant if p<0.05 (two-tail).


Patient characteristics

Our objective was to study consecutive, normal FM. From a total of 770 deliveries during the two clinical study periods, 232 FM (30%) were examined. Of the 538 unexamined FM, 207 were sent to pathology for clinical management, 302 FM were not used because they were more than 24 hours post-delivery before they could be examined, and 29 were discarded due to exclusion criteria. Additional FM (29) were collected for in vitro analysis of FM layer adherence. The demographic and labor details for these FM were similar to those used for the in vivo studies of spontaneous FM separation (Table 1).

Table 1
Patient characteristics.

Frequency of spontaneous FM separation

The frequency distribution of spontaneous FM separation revealed a wide range of spontaneous separation among 232 FM (Figure 2). Some degree of FM separation was seen in 92.1% of membranes. Of these, 64.7% exhibited greater than 10% separation of the amnion from the choriodecidua (Figure 2). No separation was detectable in 7.9% of the examined FM, and in only 4.3% of FM delivered following SROM. When separation was detected, it always involved some part of the rupture tear line. There was no association between percent of spontaneous FM separation or FM layer adherence with fetal gender, maternal race, maternal tobacco use, antibiotic use during labor, documented infection(s) during pregnancy or labor, or oxytocin use (Table 2).

Figure 2
Distribution of Percent Spontaneous Separation in 232 FM
Table 2
Clinical factors associated with percentage spontaneous FM separation and FM adherence.

Labor characteristics and FM layer separation and adherence

FM from term vaginal delivery following all modes of rupture (SROM and AROM) had significantly more spontaneous separation than FM from term unlabored, elective cesarean section (p=0.046; Figure 3A, Table 2). In vitro analysis of adherence was consistent with this finding. That is, FM from term vaginal delivery following all modes of rupture had significantly lower adherence compared to FM from term unlabored, elective cesarean section (p=0.005; Figure 3B, Table 2). This suggests that labor decreases adherence facilitating FM layer separation.

Figure 3
Effect of Labor upon Spontaneous FM Separation and Adherence

FM from term SVD following SROM had significantly more separation than FM from term SVD following AROM (p=0.046) or from term unlabored, elective cesarean section (p=0.008; Figure 4A, Table 2). Consistently, FM from term SVD following SROM had significantly lower adherence compared to FM from term unlabored, elective cesarean section (p=0.015; Figure 4B, Table 2). The association between FM separation, and admission (to the hospital) to delivery time, was also assessed. Patients with SVD following SROM, had a shorter admission to delivery time than those from SVD following AROM, (496 ± 303 versus 608 ± 375 minutes; p=0.029.) Thus, deliveries requiring the least intervention showed less adherence and more separation of the FM layers at birth. They also proceeded more rapidly.

Figure 4
Effect of Mode of Delivery on Spontaneous FM Separation and Adherence

Effect of gestational length on FM separation and adherence

Preterm FM had significantly less separation than term FM for all modes of delivery (p<0.001; Figure 5A, Table 2). Additionally, in vitro testing demonstrated that preterm FM had significantly greater adherence compared to term FM for all modes of delivery (p<0.001; Figure 5B, Table 2). The differences in percentage of separation and adherence between preterm and term fetal membranes are much greater than the differences due to labor suggesting that changes in gestational length have a larger effect (Table 2).

Figure 5
Effect of Gestational Length upon FM Separation and Adherence


The key finding of this study is that some degree of spontaneous separation of FM layers (amnion and choriodecidua) at the time of delivery is nearly universal. Further, increased separation, coupled with decreased adherence, is associated with both labor and increased gestational length, suggesting that the process has both mechanical and maturational (presumably biochemical) etiologies. Finally, greater spontaneous FM separation in association with less adherence is seen more frequently with normal birthing processes that do not require intervention (i.e., spontaneous vaginal delivery, spontaneous rupture of membranes, and term delivery). Together, these studies support the concept that FM layer separation is part of the FM weakening process during normal parturition.

There is limited medical literature focusing upon prenatal FM layer separation as documented by ultrasound. The amnion and chorion are normally separate prior to 14 weeks gestation and fuse between 14–16 weeks gestation. Separation of amnion from choriodecidua after 16 weeks gestation is associated with adverse perinatal outcomes: including PROM, chorioamnionitis, preterm birth [9], fetal chromosomal abnormalities [10, 11], amniotic bands, umbilical cord strangulation and fetal death [12,13]. Prenatal FM separation is common after invasive procedures, occurring in up to 25% of amniocentesis procedures and 12–47% of fetal surgeries [9, 12, 13]. Total (100%) FM separation is associated with extremely adverse perinatal outcomes [10, 14]. This association between FM separation and adverse perinatal outcomes emphasizes the importance of understanding the mechanism underlying spontaneous FM separation in both normal and complicated pregnancies.

While our previous work, along with that of others, has suggested that FM layer separation may be a normal part of the FM rupture process, the steps leading to FM layer separation are not clearly defined [46, 15, 16]. Current understanding of the FM weakening and rupture process suggests that FM weaken as a result of a programmed biochemical process associated with cellular apoptosis, extra-cellular matrix remodeling and degradation [3, 8,16,17]. Additional findings demonstrate that the FM weakening process initiates in the late third trimester and uterine contractions during labor likely facilitate further weakening and membrane rupture [1]. In the present study, the observation that some spontaneous separation of FM layers occurs even in term unlabored, elective Cesarean deliveries suggests that spontaneous FM separation is due, at least in part, to biochemical remodeling of the amnion-choriodecidua interface near the end of gestation. Comparison of separation and adherence of preterm FM versus term FM further supports this concept as spontaneous FM separation was significantly more frequent and extensive in term FM (≥37 weeks) compared to preterm FM (<37 weeks), independent of duration of labor. Furthermore, FM layer adherence decreased with increasing gestation.

Data from other groups also suggest biochemical changes at the amnion-choriodecidua interface promote FM separation and weakness. Meinert et al [5, 6] and Helmig et al [18] have reported changes in the proteoglycans decorin and biglycan, in conjunction with large increases in hyaluronan at this interface during late gestation. Decorin links fibrillar collagen and organizes it into coordinated bundles. Biglycan has the opposite effect by interfering with decorin organization [19, 20]. Meinert et al reported increases in biglycan in the para-cervical weak zone of FM at end gestation. Furthermore, Meinert et al demonstrated a threefold increase in hyaluronan over the last week of gestation, localized to the interface between the amnion and choriodecidua in the para-cervical weak zone [6]. Hyaluronan is known to absorb water thereby increasing tissue pressure in the extra-cellular matrix [21]. They suggest that the increased tissue pressure coupled with collagen fiber disorganization results in FM separation.

Although we have previously reported that the FM weakening process is primarily mediated by biochemical remodeling that occurs prior to labor rather than during labor itself [13], it is clear from the data above that mechanical forces of labor also affect fetal membrane adhesiveness and membrane separation. Term FM from vaginal deliveries show less adherence and more separation than term FM obtained after cesarean section without labor. Although the differences in both separation and adherence due to labor are much less than those due to gestation, to the extent that mechanical forces of labor cause separation of the FM layers, they must be considered part of the weakening process. Furthermore, FM separation is consistently seen along the tear line, in the region of the normal weak zone. It presumably exacerbates the weakening due to biochemical remodeling in this region.

In conclusion, we have demonstrated that some degree of spontaneous separation of FM layers is nearly universal following delivery, and that the propensity for separation can be quantified by determining the force of adherence using a novel methodology of FM T-Peel testing. Furthermore, we now suggest that the programmed biochemically-mediated process of FM weakening and rupture includes both weakening of the FM layers [1, 3] and, in addition, decreased adherence resulting in the separation of the FM layers. The contractions of labor augment the decreases in adherence, as well as FM separation, and thereby also contribute to the FM weakening. We speculate that FM layer separation is a part of the normal parturition process and facilitates rupture of the FM.


Support: This work was supported by NIH HD48476 grant to JJM.


Rupture of Membranes
Premature Rupture of Membranes
Spontaneous Rupture of Membranes
Artificial Rupture of Membranes
Spontaneous Vaginal Delivery


Location where study was performed: MetroHealth Medical Center, Cleveland, Ohio

Presentation: Annual Meeting of the SGI, San Diego, Ca, March 29, 2008

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. El Khwad M, Stetzer B, Moore RM, Kumar D, Mercer B, Arikat S, Redline RW, Mansour JM, Moore JJ. Term human fetal membranes have a weak zone overlying the lower uterine pole and cervix before onset of labor. Biol Reprod. 2005;72:720–6. [PubMed]
2. Pandey V, Jaremko K, Moore RM, Mercer BM, Stetzer B, Kumar D, Fox JM, Mansour JM, Moore JJ. The force required to rupture fetal membranes paradoxically increases with acute in vitro repeated stretching. Am J Obstet Gynecol. 2007;196:165.e1–7. [PubMed]
3. El Khwad M, Pandey V, Stetzer B, Mercer BM, Kumar D, Fox J, Redline RW, Mansour JM, Moore JJ. Fetal membranes from term vaginal deliveries have a zone of weakness exhibiting characteristics of apoptosis and remodeling. J Soc Gynecol Investig. 2006;13:191–5. [PubMed]
4. Arikat S, Novince RW, Mercer BM, Kumar D, Fox JM, Mansour JM, Moore JJ. Separation of amnion from choriodecidua is an integral event to the rupture of normal term fetal membranes and constitutes a significant component of the work required. Am J Obstet Gynecol. 2006;194:211–7. [PubMed]
5. Meinert M, Eriksen GV, Petersen AC, Helmig RB, Laurent C, Uldbjerg N, Malmström A. Proteoglycans and hyaluronan in human fetal membranes. Am J Obstet Gynecol. 2001;184:697–85. [PubMed]
6. Meinert M, Malmström A, Tufvesson E, Westergren-Thorsson G, Petersen AC, Laurent C, Uldbjerg N, Eriksen GV. Labour induces increased concentrations of biglycan and hyaluronan in human fetal membranes. Placenta. 2007;28:482–6. [PubMed]
7. Kumar D, Novince R, Strohl A, Mercer BM, Mansour JM, Moore RM, Moore JJ. A New Methodology to Measure Strength of Adherence of the Fetal Membrane Components, Amnion and the Choriodecidua. Placenta. 2009 doi: 10.1016/j.placenta.2009.03.014. [PMC free article] [PubMed] [Cross Ref]
8. McLaren J, Malak TM, Bell SC. Structural characteristics of term human fetal membranes prior to labor: identification of an area of altered morphology the cervix. Hum Reprod. 1999;14:237–41. [PubMed]
9. Sydorak RM, Hirose S, Sandberg PL, Filly RA, Harrison MR, Farmer DL, Albanese CT. Chorioamniotic membrane separation following fetal surgery. J Perinatol. 2002;22:407–10. [PubMed]
10. Bromley B, Shipp TD, Benacerraf BR. Amnion-chorion separation after 17 weeks’ gestation. Obstet Gynecol. 1999;94:1024–6. [PubMed]
11. Ulm B, Ulm MR, Bernaschek G. Unfused amnion and chorion after 14 weeks of gestation: associated fetal structural and chromosomal abnormalities. Ultrasound Obstet Gynecol. 1999;13:392–5. [PubMed]
12. Graf JL, Bealer JF, Gibbs DL, Adzick NS, Harrison MR. Chorioamniotic membrane separation: a potentially lethal finding. Fetal Diagn Ther. 1997;12:81–4. [PubMed]
13. Levine D, Callen PW, Pender SG, McArdle CR, Messina L, Shekhar A, Wong GP. Chorioamniotic separation after second-trimester genetic amniocentesis: importance and frequency. Radiology. 1998;209:175–81. [PubMed]
14. Lewi L, Hanssens M, Spitz B, Deprest J. Complete chorioamniotic membrane separation. Case report and review of the literature. Fetal Diagn Ther. 2004;19:78–82. [PubMed]
15. Moore RM, Mansour JM, Redline RW, Mercer BM, Moore JJ. The physiology of fetal membrane rupture: insight gained from the determination of physical properties. Placenta. 2006;27:1037–51. [PubMed]
16. Maymon E, Romero R, Pacora P, Gervasi MT, Bianco K, Ghezzi F, Yoon BH. Evidence for the participation of interstitial collagenase (matrix metalloproteinase 1) in preterm premature rupture of membranes. Am J Obstet Gynecol. 2000;183:914–20. [PubMed]
17. Oxlund H, Helmig R, Halaburt JT, Uldbjerg N. Biomechanical analysis of human chorioamniotic membranes. Eur J Obstet Gynecol Reprod Biol. 1990;34:247–55. [PubMed]
18. Helmig R, Oxlund H, Petersen LK, Uldbjerg N. Different biomechanical properties of human fetal membranes obtained before and after delivery. Eur J Obstet Gynecol Reprod Biol. 1993;48:183–9. [PubMed]
19. Uldbjerg N, Danielsen CC. A study of the interaction in vitro between type I collagen and a small dermatan sulphate proteoglycan. Biochem J. 1988;251:643–8. [PubMed]
20. Danielson KG, Baribault H, Holmes DF, Graham H, Kadler KE, Iozzo RV. Targeted disruption of decorin leads to abnormal collagen fibril morphology and skin fragility. J Cell Biol. 1997;136:729–43. [PMC free article] [PubMed]
21. Roughley PJ, Lee ER. Cartilage proteoglycans: structure and potential functions. Microsc Res Tech. 1994;28:385–97. [PubMed]