Skip to main content

Is Vertical Mucosal Thickness a Cause of Early Crestal Bone Loss or a Marker of Junction Position and Abutment Handling? A Critical Review of the Peri-Implant Soft-Tissue Seal and a Factorial Randomized Trial Protocol With Causal Mediation Analysis

Paper type
Critical review with an embedded factorial randomized controlled trial protocol
Status
Original Work · Working Paper
Keywords:
peri-implant mucosa · supracrestal tissue height · vertical soft-tissue thickness · marginal bone loss · platform switching · abutment disconnection · subcrestal placement · mediation analysis
Contents
  1. 1Introduction and Clinical Significance
  2. 2Critical Review of the Established Literature
  3. 3Identification of the Specific Research Gap
  4. 4Scientific Objective and Hypothesis
  5. 5Theoretical Foundation: An Exclusion-Zone Model of Early Bone Level
  6. 6Methodology
  7. 7Evidence Synthesis and Explicitly Labeled Theoretical Predictions
  8. 8Discussion: Mechanisms and Competing Explanations
  9. 9Implications for Implant Surgery and Restorative Dentistry
  10. 10Technical and Clinical Limitations
  11. 11Proposed Experimental and Clinical Validation
  12. 12Future Research Directions
  13. 13Evidence-Calibrated Conclusion

Abstract

Background: After an implant is exposed to the oral cavity, a mucosal seal of roughly constant vertical dimension forms around it, and crestal bone commonly remodels during the first year. A series of controlled clinical studies reports substantially greater early bone loss where vertical mucosal thickness is 2 mm or less, even with platform-switched connections, and less loss when thin tissue is thickened. Most of this evidence comes from one research group; other studies attribute early bone loss to abutment height, junction position, or repeated abutment disconnection.

Objective: To appraise the evidence that mucosal thickness determines early marginal bone stability, to formulate the competing mechanisms explicitly, and to design a trial that distinguishes them.

Methods: Critical narrative review of animal histology, human histology, controlled clinical studies, randomized trials, and meta-analyses. A geometric model was formulated in which bone level is set by the more apical of two exclusion zones, one for supracrestal tissue height and one for the inflammatory infiltrate at the implant-abutment junction. A 2 × 2 factorial randomized trial protocol with prespecified causal mediation analysis was written in accordance with SPIRIT and CONSORT.

Findings: Histological evidence that a minimum supracrestal tissue dimension is established at the expense of bone when mucosa is thin is consistent across animal studies. Clinical evidence for a thickness effect is consistent in direction and large in the originating group's studies, and contradicted or modified in others. Junction position and abutment handling have independent support. We hypothesize that more than half of the effect of thin mucosa on one-year bone level, measured from the implant shoulder, is removed when junction depth and abutment disconnection are controlled. Retention of more than half of the effect would refute the hypothesis.

Conclusions: Soft-tissue thickness matters to early bone level; how much of its effect is causal and how much reflects correlated surgical and prosthetic variables is unresolved. The model and trial proposed here are untested.

Keywords: peri-implant mucosa; supracrestal tissue height; vertical soft-tissue thickness; marginal bone loss; platform switching; abutment disconnection; subcrestal placement; mediation analysis

1. Introduction and Clinical Significance

Marginal bone level around an implant is the radiographic outcome clinicians watch most closely. Classical success criteria allowed for bone loss during the first year of function and required stability thereafter [1], an allowance that reflected the behavior of external-hexagon implants with matching abutments placed at the crest. Early bone loss is not merely cosmetic. Implants that lose more bone in the first months are more likely to reach levels that compromise their prognosis [2], and exposed microrough threads are harder to keep free of biofilm. Understanding what sets the initial bone level is therefore a matter of long-term peri-implant health and of esthetics in the anterior zone.

Explanations for early remodeling have accumulated in layers. The first located the cause at the implant-abutment junction: a microgap colonized by bacteria, with bone retreating to a standard distance from it. The second located it in the soft tissue: the mucosa requires a minimum vertical dimension to form an epithelial and connective-tissue seal, and where that dimension is lacking it is obtained by bone resorption. The third located it in clinical handling: every removal of an abutment tears the connective-tissue attachment, which re-forms more apically. Platform switching, subcrestal placement, tall abutments, and one-time abutment protocols are each a response to one of these explanations. Soft-tissue thickening is the most recent, advanced by studies reporting that vertical mucosal thickness at placement is a strong determinant of bone loss and that thickening thin tissue prevents it.

The clinical appeal of that proposition is obvious, and it has been widely adopted. Its evidential basis deserves scrutiny for three reasons. The largest effects come from a single group. Thickness is correlated with other variables, biotype, ridge anatomy, the depth at which the surgeon seats the implant, the abutment height the restorative dentist can then use, that have their own claims on bone level. And the interventions that modify thickness also modify those variables. This paper asks how the thickness effect can be separated from its correlates.

2. Critical Review of the Established Literature

The soft-tissue seal: histology

Comparative histology in the dog established that the peri-implant mucosa, like gingiva, has a junctional epithelium about 2 mm long and a zone of connective tissue about 1 to 1.5 mm high between the apical termination of the epithelium and the bone crest [3]. The connective tissue adjacent to the implant is poor in cells and vessels and rich in collagen fibers that run parallel to the implant surface; a thin inner zone rich in fibroblasts was described in contact with the titanium [4, 5]. In human volunteers, the mucosal attachment to experimental abutments matured over several weeks to dimensions similar to those in animals [6]. A dimension of this order, usually quoted as 3 to 4 mm, was found around different implant systems, one- and two-piece [7–9], and the consensus description of peri-implant health adopts it [10].

The experiment most often cited for a minimum dimension is that of Berglundh and Lindhe, who reduced the vertical thickness of the mucosa before abutment connection in dogs and found that wound healing consistently included bone resorption that restored a mucosal attachment of approximately the usual dimension [11]. This is the histological origin of the thickness hypothesis.

The junction: microgap and infiltrate

A parallel line of animal work examined the implant-abutment interface. In the canine mandible, crestal bone around two-piece implants remodeled to a position roughly 2 mm apical of the interface regardless of whether the implant was submerged, whereas one-piece implants without an interface at the crest showed minimal change [12]; the size of the microgap had little influence, and movement between components was implicated [13]. An inflammatory cell infiltrate was described in the connective tissue at the level of the junction [14], and peak inflammatory cell density was shown to be centered on the interface, with bone loss greater the deeper the interface lay [15]. On this account bone level is set by a zone of inflammation around the junction.

Platform switching, connecting a narrower abutment to the implant so that the junction lies inward from the implant shoulder, was introduced on the reasoning that the infiltrate would be shifted away from bone [16]. Meta-analyses found less marginal bone loss with platform switching than with matching connections, with the effect increasing with the size of the horizontal offset [17, 18].

Abutment disconnection

In the dog, repeated disconnection and reconnection of abutments resulted in a more apically positioned connective-tissue attachment and additional marginal bone resorption relative to undisturbed abutments [19]. Clinical evidence is mixed in size but consistent in direction. A randomized trial found no significant short-term effect of healing-abutment disconnection and reconnection on tissues [20]; a meta-analysis by the same first author found significantly greater bone loss with multiple disconnections, the weighted mean difference being about 0.2 mm [21]; a systematic review and meta-analysis of the one-abutment one-time protocol reported a small but significant reduction in bone loss [22]; and a randomized trial found interproximal bone levels favored one-time abutment placement [23].

Abutment height and junction depth

The height of the prosthetic abutment, which sets the vertical distance from the junction to the crown margin and thus the space available for the mucosal seal on a stable surface, was associated with marginal bone loss in a clinical study, shorter abutments showing more loss [24]; a randomized trial found less early interproximal bone loss with taller abutments [25]; and a cluster-randomized trial reported that establishment of biological width was influenced by abutment height irrespective of vertical mucosal thickness [26]. That last finding is in direct tension with a primary role for thickness.

Subcrestal placement moves the junction and the implant shoulder below the crest. A systematic review and meta-analysis of platform-switched implants found that subcrestal placement was not associated with greater marginal bone loss than equicrestal placement and may reduce implant-surface exposure [27]. A prospective intra-subject comparison showed that adapting the vertical position of conical-connection implants to the soft-tissue thickness, placing them deeper where tissue was thin, prevented early exposure of implant surface [28], following an earlier report from the same group that initial soft-tissue thickness influenced peri-implant bone remodeling [29].

Vertical mucosal thickness: clinical studies

In a one-year prospective controlled study, implants placed where the crestal mucosa was 2 mm thick or less underwent significantly more crestal bone loss, approaching 1.5 mm, than implants in thicker tissue [30]. Because the test implants in that study were positioned supracrestally, later work addressed bone-level implants: with platform-switched connections, one-year bone loss was 1.17 mm in thin tissue against 0.21 mm in thick tissue [31]; with horizontally matching connections, thin tissue lost about 1.7 mm, thick tissue about 0.3 mm, and thin tissue thickened with an allogeneic membrane about 0.45 mm [32]; and around platform-switched bone-level implants, thickening reduced early bone loss to the level seen with naturally thick tissue [33]. A subsequent study examined bone stability around implants restored on titanium bases in relation to vertical thickness [34]. Two trials bear directly on the role of implant depth. In thin tissue, implants placed 1.5 mm subcrestally showed 0.18 mm of bone loss relative to the implant at two years, against 0.51 mm for epicrestal implants whose tissue had been thickened by tenting over a healing abutment, although the crest above the subcrestal implants had remodeled by about 1.2 mm [35]. An independent randomized trial of platform-switching implants found more early bone change with supracrestal tissue height of 2 mm or less at crestally placed implants, no difference between crestal and subcrestal placement where tissue height exceeded 2 mm, and a benefit of subcrestal placement where it did not [36].

Meta-analyses pooling these and other studies found less marginal bone loss around implants with thick than with thin mucosa, with a weighted mean difference of about 0.8 mm in one [37, 38]. The pooled estimates are dominated by the originating group's data. A prospective clinical and histological cohort from another group found no significant difference in bone loss between thin and thick mucosa around platform-switching bone-level implants at one or three years [39]. Systematic review of soft-tissue augmentation concluded that grafting to increase mucosal thickness resulted in less marginal bone loss over time, on limited evidence, and that augmentation of keratinized tissue improved bleeding and plaque indices [40]; the accompanying consensus report endorsed these conclusions with reservations about the evidence base [41]. Width of keratinized mucosa is a distinct variable with its own literature [42], and the concept of a peri-implant phenotype now groups mucosal thickness, keratinized width, supracrestal tissue height, and bone thickness as related but separable dimensions [43].

Appraisal

The direction of the thickness association is consistent. Its magnitude and independence are uncertain, for identifiable reasons. Thickness is measured once, at flap elevation, with a probe, to the nearest half millimeter, and dichotomized at 2 mm; the cut-point is conventional. Allocation to thin and thick groups is by nature, so those groups may differ in bone anatomy and vascularity. Thickening is achieved by adding a graft or by leaving a healing abutment submerged, both of which alter healing in other ways, and in some designs by placing the implant deeper, which changes junction position. Radiographic bone level is measured from the implant shoulder; when implants are placed at different depths, the same biological event yields different measured loss. Radiographic measurement itself is subject to projection and reader error [44]. And replication outside the originating center is limited, with at least one null result and one trial pointing to abutment height instead.

3. Identification of the Specific Research Gap

Established: a supracrestal soft-tissue dimension of about 3 to 4 mm forms around implants and, in animals, is obtained by bone resorption when mucosa is thinned; an inflammatory zone is centered on the implant-abutment junction; platform switching reduces early bone loss; repeated disconnection is at least mildly detrimental; thin mucosa is associated with more early bone loss in several controlled clinical studies.

Uncertain: whether the clinical thickness effect is a direct causal effect of tissue dimension or is transmitted through, or confounded with, junction position, abutment height, and abutment handling; and whether thickening is beneficial when those are optimized.

Existing designs cannot resolve this. One-factor comparisons of thin versus thick tissue cannot separate thickness from its natural correlates. Trials of thickening change thickness together with other aspects of healing. Studies of abutment height or disconnection seldom measure thickness. No trial has crossed a thickness intervention with a junction or handling intervention, and no study has measured the realized supracrestal tissue height as a mediator and asked how much of an intervention's effect passes through it. Outcome definitions that reference the implant shoulder conflate bone remodeling with implant positioning.

The specific gap is a factorial randomized comparison in thin-mucosa sites, with standardized implant depth, a measured mediator, and bone level expressed both relative to the implant shoulder and relative to the original crest.

4. Scientific Objective and Hypothesis

The objective is to estimate the causal effect of vertical mucosal thickness on one-year marginal bone level and to determine how much of that effect is independent of junction position and abutment handling.

Hypothesis H11: in sites with vertical mucosal thickness of 2 mm or less receiving platform-switched, conical-connection implants, more than 50 percent of the effect of soft-tissue thickening on one-year bone level relative to the implant shoulder is eliminated when implants are placed subcrestally by a standardized depth and restored with a one-time definitive abutment; equivalently, the effect of thickening under optimized junction conditions is less than half its effect under conventional conditions.

H11 is falsified if the effect of thickening in the optimized-junction arm is more than half of its effect in the conventional arm, with a confidence interval for the ratio that excludes 0.5 from below. A second, mediation-based statement, H11b, is that the proportion of the thickening effect mediated by measured supracrestal tissue height at abutment connection exceeds 50 percent; H11b is falsified if the proportion mediated has a confidence interval lying below 0.5, which would indicate that thickening acts chiefly by a route other than tissue height, for example by a barrier or vascular effect of the graft.

The two statements address different questions. H11 asks whether thickening is needed once the junction is managed. H11b asks whether thickening works through the dimension it is supposed to change.

5. Theoretical Foundation: An Exclusion-Zone Model of Early Bone Level

Two requirements

The histological literature supports two spatial requirements. First, the mucosal seal occupies a vertical dimension H*, measured from the mucosal margin to the most coronal bone-to-implant contact, of roughly 3 to 4 mm. Second, the junction between implant and abutment is surrounded by a zone of inflamed connective tissue of radius r, of the order of 1 to 1.5 mm in animal studies, within which bone is not maintained. Bone settles at the most coronal level that satisfies both.

Supracrestal tissue requirement

Let T be the vertical thickness of mucosa over the crest at placement and d the depth of the implant shoulder below the original crest (positive when subcrestal, zero when equicrestal, negative when supracrestal). If the available tissue height is less than H*, the deficit is made up by resorption of crestal bone. The resulting position of first bone-to-implant contact apical to the implant shoulder is

B_S = max(0, H^* - T - d) (1)

in millimeters. Loss measured from the original crest is max(0, H* − T), which does not depend on d: subcrestal placement does not prevent the crest from remodeling, it prevents the remodeling from exposing implant surface. The distinction explains how studies using different reference points can reach different conclusions from the same biology.

Junction requirement

Let s be the horizontal offset of the junction from the implant shoulder (zero for matching connections, positive for platform switching) and v the vertical distance from the junction to the level of the implant shoulder (zero for flat-to-flat connections at the shoulder). The distance from the junction to the nearest bone at the shoulder is the hypotenuse of s and v, and bone is excluded within radius r of the junction. Considering bone along the outer implant surface, the exclusion extends apical to the shoulder by

B_J = max(0, √(r^2 - s^2) - v) for s < r (2)

and zero for s greater than or equal to r. A horizontal offset approaching r removes the junctional contribution, the rationale of platform switching [16]; the dependence on offset size agrees qualitatively with meta-analytic findings [17]. The radius r is not a constant. It is expected to increase with bacterial leakage and micromovement at the connection and with each disruption of the attachment by abutment removal, which we represent as

r = r_0 + n ρ (3)

where r0 is the radius for an undisturbed stable connection, n the number of disconnections, and ρ the increment per disconnection. The linear form is a convenience; the animal data show an effect of repeated disconnection without establishing its shape [19].

Combined prediction

The predicted bone level apical to the implant shoulder after initial remodeling is

B = max(B_S, B_J) (4)

The maximum, and not the sum, is appropriate because each requirement specifies a boundary and bone conforms to the more apical one. Three regimes follow. Where tissue is thick and the junction well managed, both terms are zero and no implant surface is exposed. Where tissue is thin and the implant is at the crest, BS dominates and bone loss is proportional to the tissue deficit: this is the regime of the original thickness studies. Where tissue is adequate but the junction is at the shoulder with no offset and abutments are repeatedly removed, BJ dominates and thickness is irrelevant: this is the regime in which abutment height and handling appear decisive and thickness does not.

The model accounts for conflicting clinical findings by supposing that studies sampled different regimes. It also makes a prediction that distinguishes it from a pure thickness account: in thin tissue, subcrestal placement by a depth equal to the deficit should abolish implant-surface exposure without any thickening, provided the junction term is controlled. And it identifies a limitation of thickening by graft: a graft raises T only if it survives as vital tissue of stable dimension, and the gain must be verified by measurement.

6. Methodology

Review methods

The review is narrative and critical. Animal and human histology, controlled clinical studies, randomized trials, and meta-analyses were considered, with attention to the origin of data pooled in meta-analyses. No formal risk-of-bias assessment was performed.

Trial design

The proposed study is a multicenter, 2 × 2 factorial, randomized controlled trial, with a protocol conforming to SPIRIT [45] and reporting to CONSORT [46]. Factor A is soft-tissue thickening (connective-tissue graft or substitute placed over the implant at surgery versus no thickening). Factor B is junction management (optimized: subcrestal placement by a standardized depth with a definitive abutment placed at surgery or at uncovering and never removed; conventional: equicrestal placement with healing abutment and the customary sequence of disconnections for impression and delivery). The four cells are shown in Table 1. Centers outside the group that originated the thickness hypothesis are essential, since independent replication is one of the purposes.

Table 1. Cells of the proposed 2 × 2 factorial trial in thin-mucosa sites.
CellSoft tissueJunction management
1No thickeningConventional: equicrestal; healing abutment; standard disconnections
2ThickeningConventional
3No thickeningOptimized: subcrestal by standardized depth; one-time definitive abutment
4ThickeningOptimized

Participants

Adults requiring a single implant in a healed posterior site, with adequate bone width and height for a standard implant without bone augmentation, are screened. After flap elevation, vertical mucosal thickness is measured at the crest with a calibrated probe by a trained examiner; sites with thickness of 2 mm or less are eligible and randomized. Sites with thicker mucosa are not randomized but are followed as an observational reference cohort receiving conventional junction management without thickening, to anchor the comparison with naturally thick tissue. Smokers of more than ten cigarettes per day, uncontrolled diabetes, and untreated periodontitis are exclusions, to limit variance from other causes of bone loss.

Interventions

All sites receive the same implant system with a conical, platform-switched connection and a microrough surface to the shoulder. In optimized cells the shoulder is placed 1.5 mm below the buccal and lingual crest, verified with a depth gauge; in conventional cells, at the crest. Thickening uses a single specified material and technique at all centers, with the choice between autogenous connective tissue and a substitute made at protocol finalization and fixed thereafter. In optimized cells a definitive abutment of standardized height is connected and impressions are made at abutment level; in conventional cells, healing abutments are removed for implant-level impression, try-in, and delivery, a minimum of three disconnections. Restorations are screw-retained crowns with standardized emergence.

Randomization and blinding

Randomization is stratified by center and by jaw, using permuted blocks of varying size with concealed central allocation revealed after thickness measurement. Surgeons cannot be blinded. Radiographic examiners are blinded to thickening and, as far as radiographs permit, to junction management; statistical analysis is performed on coded groups.

Outcomes

The primary outcome is the position of the first radiographic bone-to-implant contact relative to the implant shoulder at 12 months after loading, the mean of mesial and distal sites, on standardized periapical radiographs taken with individualized holders. A co-primary descriptive outcome is change in bone level relative to the original crest, obtained by referencing the baseline radiograph taken at placement; reporting both allows the two reference frames in Equation 1 to be compared. The mediator is supracrestal tissue height at the time of definitive abutment connection or, in conventional cells, at final impression, measured as the distance from the mucosal margin to the bone crest by transmucosal probing under anesthesia and, in a subsample, by ultrasonography. Secondary outcomes are mucosal thickness at 12 months, probing depth, bleeding on probing, mucosal recession, keratinized mucosa width, patient-reported morbidity of grafting, and operative time. Radiographic measurement error is estimated from duplicate radiographs in a subsample.

Statistical analysis

The primary analysis is a linear model for the primary outcome with terms for factor A, factor B, and their interaction, adjusted for stratification variables, with patient as the unit. H11 concerns the interaction: the simple effect of thickening within conventional junction management, δconv, and within optimized junction management, δopt, are estimated, and the ratio δopt/δconv is reported with a bootstrap confidence interval.

For H11b, causal mediation analysis decomposes the total effect of thickening on bone level into a natural indirect effect through supracrestal tissue height and a natural direct effect by other paths [47]. With TE the total effect, NDE the natural direct effect, and NIE the natural indirect effect,

TE = NDE + NIE, PM = NIE/(NDE + NIE) (5)

where PM is the proportion mediated. Randomization of thickening removes confounding of the exposure-outcome and exposure-mediator relations; the mediator-outcome relation is not randomized, and its estimation assumes no unmeasured common causes of tissue height and bone level. Measured baseline covariates (bone crest width, keratinized mucosa width, site, age) are included, and a sensitivity analysis reports how strong an unmeasured confounder would have to be to change the conclusion. The analysis is conducted within each level of factor B and pooled if no interaction with the mediator is found.

Sample size

A factorial trial powered for an interaction needs approximately four times the sample required for a main effect of the same size. For planning, suppose the effect of thickening under conventional management is 0.8 mm and under optimized management 0.3 mm, a difference in effects of 0.5 mm, with a standard deviation of 0.6 mm for the outcome; these are assumptions chosen within the range of published thin-versus-thick differences and must be revisited after a pilot. The standard error of the interaction contrast with n per cell is 2σ divided by the square root of n. Detecting 0.5 mm with two-sided α of 0.05 and power 0.80 requires n of approximately 45 per cell, or 180 randomized patients, and 210 with allowance for 15 percent attrition. The trial would have high power for the main effects and moderate precision for the ratio in H11. The mediation analysis has lower power and its result should be interpreted with its confidence interval and not as a significance test.

7. Evidence Synthesis and Explicitly Labeled Theoretical Predictions

Synthesis

Table 2 arranges the principal clinical evidence by the mechanism each study bears on and notes what was and was not controlled.

Table 2. Clinical evidence on determinants of early marginal bone level, by mechanism.
MechanismSupporting evidenceNot controlled or uncertain
Vertical mucosal thickness [30, 31, 33, 37]Controlled studies and meta-analyses: more bone loss with thin tissue; thickening protectiveSingle-group dominance; junction depth and abutment handling; reference point of measurement
Thickness not determinant [26, 39]Independent cohort without association; cluster trial attributing effect to abutment heightDifferent implant designs and protocols; smaller samples
Junction offset [17, 18]Meta-analyses of platform switchingHeterogeneous designs; thickness seldom recorded
Abutment disconnection [21–23]Meta-analyses and a randomized trial: small benefit of one-time abutmentShort follow-up; small effect sizes
Abutment height [24, 25]Cohort and randomized trial: less loss with taller abutmentsConfounding with tissue thickness in observational data
Junction depth [27, 28]Meta-analysis and intra-subject study: subcrestal placement limits implant exposureOutcome reference frame varies

Theoretical predictions

The following are predictions of the exclusion-zone model and have not been tested.

Prediction 1. In cell 1 (thin tissue, conventional junction), first bone contact at 12 months will lie apical to the implant shoulder by an amount close to the tissue deficit, of the order of 1 to 1.5 mm.

Prediction 2. In cell 3 (thin tissue, no thickening, optimized junction), bone measured from the original crest will remodel as in cell 1, but first bone contact will lie at or coronal to the implant shoulder, because the 1.5 mm subcrestal depth absorbs the deficit. Two published trials report a pattern of this kind [35, 36]; the prediction is therefore a consistency check on the model more than a new test.

Prediction 3. Thickening will reduce remodeling measured from the original crest in both junction conditions if, and only to the extent that, measured supracrestal tissue height increases.

Prediction 4. The effect of thickening on the shoulder-referenced outcome will be large in conventional cells and small in optimized cells, giving a ratio below 0.5 (H11).

Prediction 5. In the observational thick-tissue cohort under conventional management, bone loss will reflect the junction term alone and will be modest with a platform-switched conical connection.

Prediction 6. Thickening will carry measurable costs in operative time and patient-reported morbidity, relevant to whether a small residual benefit under optimized conditions justifies it.

8. Discussion: Mechanisms and Competing Explanations

The exclusion-zone model is a way of organizing three mechanisms so that their predictions can be told apart. It has obvious limitations as biology. The supracrestal dimension H* is treated as a fixed requirement, whereas it varies between individuals, sites, and abutment materials and geometries; the human data show dimensions similar to, and not identical with, those in animals [6]. The junctional radius is taken from animal studies of matching connections with documented micromovement [13, 15]; conical connections with minimal micromovement may have a much smaller radius, in which case the junction term is negligible in the proposed trial and the comparison reduces to thickness versus depth. The model ignores the horizontal dimension of bone: a thin buccal plate may resorb for vascular reasons unrelated to either zone. It treats early remodeling as a one-time adjustment and says nothing about later, inflammation-driven loss.

The strongest competing account is the direct one advanced by proponents of thickening: thin tissue is poorly vascularized, heals with more inflammation, seals less effectively, and so permits bone loss regardless of where the junction lies. On this view thickening should help even with subcrestal placement and a one-time abutment, the ratio in H11 will be near 1, and the hypothesis is refuted. The trial is designed so that this outcome is as readable as the opposite one. The large benefit of thickening reported with horizontally matching connections placed at the crest [32] is what the direct account predicts and is also what the model predicts for that regime, so it does not discriminate. The two trials that varied depth do: in thin tissue, subcrestal placement without thickening gave less implant-referenced bone loss than epicrestal placement with tissue tenting, while the crest above the implant remodeled by roughly the amount the model would attribute to the tissue deficit [35], and subcrestal placement was beneficial only where tissue height was insufficient [36]. These findings are in line with Prediction 2. What neither trial tested is whether thickening adds anything once depth and abutment handling are optimized, which is the comparison H11 requires.

A second competing account gives primacy to abutment height: the seal needs a sufficient length of undisturbed abutment surface, and short abutments force it onto the junction and implant shoulder [24, 26]. Abutment height and subcrestal depth are partly interchangeable in their effect on the distance from crown margin to bone, and partly distinct, since a tall abutment in thin tissue is exposed or forces the crown margin supragingivally. The trial fixes abutment height within cells to keep this variable from drifting, which means it cannot estimate the effect of height; a three-factor design would be needed and would be prohibitively large.

A third issue is the meaning of the mediator. Supracrestal tissue height at abutment connection is a composite of original thickness, any gain from grafting, and any gain from bone remodeling that has already occurred by that time. If bone has resorbed before the mediator is measured, tissue height will have increased because of bone loss, and a naive analysis would find that greater tissue height predicts more bone loss. The timing of mediator measurement is therefore critical; measuring it early, before substantial remodeling, and referencing it to the original crest on the baseline radiograph reduces this reverse dependence without eliminating it. The assumption of no unmeasured mediator-outcome confounding is also strong: local vascularity or bone density could influence both. These are reasons to regard H11b as exploratory and H11, which rests on randomization alone, as primary.

Measurement deserves equal attention. A probe reading at the crest after flap elevation is sensitive to flap design, tissue compression, and the precise site. Dichotomizing at 2 mm discards information and invites misclassification of sites near the threshold; analyzing thickness as a continuous variable in the observational cohort, and reporting the dose-response form predicted by Equation 1, a linear decline to zero at H*, would be a sharper test than a two-group comparison. Radiographic bone level measured mesially and distally says nothing about the buccal aspect, where thin tissue and thin bone coincide and where recession is clinically visible.

Finally, the dependence of pooled estimates on a single group is a statement about the evidence and not about the quality of that group's work, which introduced the question and pursued it systematically. Independent replication is the ordinary requirement of any clinical claim, and the one null cohort [39] and the abutment-height trial [26] show that it is not a formality here.

9. Implications for Implant Surgery and Restorative Dentistry

On present evidence, vertical mucosal thickness should be measured at surgery, because it identifies sites at risk of early implant-surface exposure and the measurement costs nothing. Where tissue is thin, several responses have support: placing a platform-switched, conical-connection implant subcrestally by an amount related to the deficit [27, 28]; using an abutment of adequate height [25]; minimizing disconnections [22]; or thickening the tissue [33, 40]. These are not mutually exclusive, and no trial shows which combination is sufficient. Thickening adds surgical time, cost, and, with autogenous tissue, a donor site; the model suggests it may add little where depth and abutment handling are already optimized, but that is a hypothesis.

Restorative clinicians influence two of the three mechanisms. Abutment-level workflows that avoid repeated removal, and selection of abutment height with the tissue dimension in mind, are within their control. Communication of the measured thickness from surgeon to restorative dentist should be routine.

For interpretation of radiographs, bone remodeling to the implant shoulder after subcrestal placement is expected and is not pathological loss; reporting bone level relative to the shoulder and relative to the original crest avoids confusion.

10. Technical and Clinical Limitations

The review is narrative. The model is a geometric caricature with parameters taken from animal studies of older implant designs. The trial is restricted to single posterior implants with one implant system, which limits generalization to anterior sites, where esthetics and thin buccal bone dominate, and to other connections. Surgeons are unblinded, and performance bias in handling of thin tissue is possible. The conventional arm involves a fixed number of disconnections that may exceed or fall short of local practice. Twelve months captures early remodeling and not long-term stability or peri-implantitis. The mediation analysis depends on untestable assumptions. The sample-size inputs are assumptions, and the interaction test has limited power if the true difference in effects is smaller than assumed. Exclusion of heavy smokers and uncontrolled diabetes improves internal validity at the cost of applicability.

11. Proposed Experimental and Clinical Validation

The trial is the principal validation. Two preliminary steps are needed. First, a measurement study establishes the repeatability of crestal mucosal thickness measurement and of supracrestal tissue height measurement by probing and by ultrasonography; if repeatability is poor, the mediation analysis is not feasible. Second, a pilot of 40 patients estimates the outcome standard deviation and confirms that the subcrestal depth and abutment protocol can be delivered consistently across centers. The trial then proceeds with an independent data monitoring committee and a prespecified statistical analysis plan registered before enrollment.

The model itself can be examined on existing data without new patients: individual-patient data from published trials that recorded mucosal thickness, implant depth, and abutment protocol could be pooled to test the piecewise-linear dose-response of Equation 1 and the absence of a thickness effect above the required height. Such a re-analysis would need the cooperation of the original investigators. Histological validation of the junctional radius for modern conical connections, in an animal model with and without repeated disconnection, would establish whether the junction term remains relevant. Refutation criteria are stated in Section 4.

12. Future Research Directions

Several questions follow. Non-invasive measurement of mucosal thickness before surgery, by ultrasonography or by registered optical and tomographic scans, would allow planning of implant depth and abutment height in advance and would make thickness a continuous, reproducible variable. The buccal aspect, invisible on periapical radiographs, needs three-dimensional outcome assessment with controlled artifact. The long-term consequence of early remodeling pattern for peri-implantitis risk is unknown and is the question that ultimately determines how much early bone level matters. Abutment material and surface, which influence the quality of the connective-tissue interface, are a further dimension the present model omits. And cost-effectiveness analysis of thickening versus depth adjustment would inform practice once the effects are known.

13. Evidence-Calibrated Conclusion

Animal histology supports a minimum supracrestal tissue dimension obtained at the expense of bone and an inflammatory zone at the implant-abutment junction. Controlled clinical studies, most from one group, associate thin vertical mucosa with greater early bone loss and report benefit from thickening; meta-analyses agree in direction; independent evidence is limited and partly discordant. Platform switching, taller abutments, subcrestal placement, and avoidance of repeated disconnection each have clinical support of low to moderate certainty. Whether thickness acts directly or mainly through junction position and handling has not been tested. The exclusion-zone model proposed here reconciles the conflicting findings as different regimes and predicts that thickening adds little when junction depth and abutment handling are optimized. That prediction may be wrong, and the factorial trial described is designed to show it if so. No patient data are reported.

Cite this paper

Jain S, Gupta A. Is Vertical Mucosal Thickness a Cause of Early Crestal Bone Loss or a Marker of Junction Position and Abutment Handling? A Critical Review of the Peri-Implant Soft-Tissue Seal and a Factorial Randomized Trial Protocol With Causal Mediation Analysis. Working paper. Center for Implant Dentistry; 2026. Available from: https://bayareaimplantdentistry.com/clinician-research-and-publications/soft-tissue-thickness-and-marginal-bone-cause-or-proxy

References

  1. Albrektsson T, Zarb G, Worthington P, Eriksson AR. The long-term efficacy of currently used dental implants: a review and proposed criteria of success. Int J Oral Maxillofac Implants. 1986;1(1):11-25. PMID: 3527955 (opens in a new tab).
  2. Galindo-Moreno P, León-Cano A, Ortega-Oller I, Monje A, O Valle F, Catena A. Marginal bone loss as success criterion in implant dentistry: beyond 2 mm. Clin Oral Implants Res. 2015;26(4):e28-e34. doi:10.1111/clr.12324 (opens in a new tab) PMID: 24383987 (opens in a new tab).
  3. Berglundh T, Lindhe J, Ericsson I, Marinello CP, Liljenberg B, Thomsen P. The soft tissue barrier at implants and teeth. Clin Oral Implants Res. 1991;2(2):81-90. doi:10.1034/j.1600-0501.1991.020206.x (opens in a new tab) PMID: 1809403 (opens in a new tab).
  4. Moon IS, Berglundh T, Abrahamsson I, Linder E, Lindhe J. The barrier between the keratinized mucosa and the dental implant. An experimental study in the dog. J Clin Periodontol. 1999;26(10):658-63. doi:10.1034/j.1600-051x.1999.261005.x (opens in a new tab) PMID: 10522777 (opens in a new tab).
  5. Buser D, Weber HP, Donath K, Fiorellini JP, Paquette DW, Williams RC. Soft tissue reactions to non-submerged unloaded titanium implants in beagle dogs. J Periodontol. 1992;63(3):225-35. doi:10.1902/jop.1992.63.3.225 (opens in a new tab) PMID: 1593416 (opens in a new tab).
  6. Tomasi C, Tessarolo F, Caola I, Wennström J, Nollo G, Berglundh T. Morphogenesis of peri-implant mucosa revisited: an experimental study in humans. Clin Oral Implants Res. 2014;25(9):997-1003. doi:10.1111/clr.12223 (opens in a new tab) PMID: 23799997 (opens in a new tab).
  7. Abrahamsson I, Berglundh T, Wennström J, Lindhe J. The peri-implant hard and soft tissues at different implant systems. A comparative study in the dog. Clin Oral Implants Res. 1996;7(3):212-9. doi:10.1034/j.1600-0501.1996.070303.x (opens in a new tab) PMID: 9151585 (opens in a new tab).
  8. Cochran DL, Hermann JS, Schenk RK, Higginbottom FL, Buser D. Biologic width around titanium implants. A histometric analysis of the implanto-gingival junction around unloaded and loaded nonsubmerged implants in the canine mandible. J Periodontol. 1997;68(2):186-98. doi:10.1902/jop.1997.68.2.186 (opens in a new tab) PMID: 9058338 (opens in a new tab).
  9. Hermann JS, Buser D, Schenk RK, Schoolfield JD, Cochran DL. Biologic Width around one- and two-piece titanium implants. Clin Oral Implants Res. 2001;12(6):559-71. doi:10.1034/j.1600-0501.2001.120603.x (opens in a new tab) PMID: 11737099 (opens in a new tab).
  10. Araujo MG, Lindhe J. Peri-implant health. J Periodontol. 2018;89 Suppl 1:S249-S256. doi:10.1002/JPER.16-0424 (opens in a new tab) PMID: 29926949 (opens in a new tab).
  11. Berglundh T, Lindhe J. Dimension of the periimplant mucosa. Biological width revisited. J Clin Periodontol. 1996;23(10):971-3. doi:10.1111/j.1600-051x.1996.tb00520.x (opens in a new tab) PMID: 8915028 (opens in a new tab).
  12. Hermann JS, Buser D, Schenk RK, Cochran DL. Crestal bone changes around titanium implants. A histometric evaluation of unloaded non-submerged and submerged implants in the canine mandible. J Periodontol. 2000;71(9):1412-24. doi:10.1902/jop.2000.71.9.1412 (opens in a new tab) PMID: 11022770 (opens in a new tab).
  13. Hermann JS, Schoolfield JD, Schenk RK, Buser D, Cochran DL. Influence of the size of the microgap on crestal bone changes around titanium implants. A histometric evaluation of unloaded non-submerged implants in the canine mandible. J Periodontol. 2001;72(10):1372-83. doi:10.1902/jop.2001.72.10.1372 (opens in a new tab) PMID: 11699479 (opens in a new tab).
  14. Ericsson I, Persson LG, Berglundh T, Marinello CP, Lindhe J, Klinge B. Different types of inflammatory reactions in peri-implant soft tissues. J Clin Periodontol. 1995;22(3):255-61. doi:10.1111/j.1600-051x.1995.tb00143.x (opens in a new tab) PMID: 7790533 (opens in a new tab).
  15. Broggini N, McManus LM, Hermann JS, Medina R, Schenk RK, Buser D, et al. Peri-implant inflammation defined by the implant-abutment interface. J Dent Res. 2006;85(5):473-8. doi:10.1177/154405910608500515 (opens in a new tab) PMID: 16632764 (opens in a new tab).
  16. Lazzara RJ, Porter SS. Platform switching: a new concept in implant dentistry for controlling postrestorative crestal bone levels. Int J Periodontics Restorative Dent. 2006;26(1):9-17. PMID: 16515092 (opens in a new tab).
  17. Atieh MA, Ibrahim HM, Atieh AH. Platform switching for marginal bone preservation around dental implants: a systematic review and meta-analysis. J Periodontol. 2010;81(10):1350-66. doi:10.1902/jop.2010.100232 (opens in a new tab) PMID: 20575657 (opens in a new tab).
  18. Strietzel FP, Neumann K, Hertel M. Impact of platform switching on marginal peri-implant bone-level changes. A systematic review and meta-analysis. Clin Oral Implants Res. 2015;26(3):342-58. doi:10.1111/clr.12339 (opens in a new tab) PMID: 24438506 (opens in a new tab).
  19. Abrahamsson I, Berglundh T, Lindhe J. The mucosal barrier following abutment dis/reconnection. An experimental study in dogs. J Clin Periodontol. 1997;24(8):568-72. doi:10.1111/j.1600-051x.1997.tb00230.x (opens in a new tab) PMID: 9266344 (opens in a new tab).
  20. Koutouzis T, Koutouzis G, Gadalla H, Neiva R. The effect of healing abutment reconnection and disconnection on soft and hard peri-implant tissues: a short-term randomized controlled clinical trial. Int J Oral Maxillofac Implants. 2013;28(3):807-14. doi:10.11607/jomi.3022 (opens in a new tab) PMID: 23748313 (opens in a new tab).
  21. Koutouzis T, Gholami F, Reynolds J, Lundgren T, Kotsakis GA. Abutment Disconnection/Reconnection Affects Peri-implant Marginal Bone Levels: A Meta-Analysis. Int J Oral Maxillofac Implants. 2017;32(3):575–581. doi:10.11607/jomi.5367 (opens in a new tab) PMID: 28334059 (opens in a new tab).
  22. Atieh MA, Tawse-Smith A, Alsabeeha NHM, Ma S, Duncan WJ. The One Abutment-One Time Protocol: A Systematic Review and Meta-Analysis. J Periodontol. 2017;88(11):1173-1185. doi:10.1902/jop.2017.170238 (opens in a new tab) PMID: 28625076 (opens in a new tab).
  23. Molina A, Sanz-Sánchez I, Martín C, Blanco J, Sanz M. The effect of one-time abutment placement on interproximal bone levels and peri-implant soft tissues: a prospective randomized clinical trial. Clin Oral Implants Res. 2017;28(4):443-452. doi:10.1111/clr.12818 (opens in a new tab) PMID: 27016157 (opens in a new tab).
  24. Galindo-Moreno P, León-Cano A, Ortega-Oller I, Monje A, Suárez F, ÓValle F, et al. Prosthetic Abutment Height is a Key Factor in Peri-implant Marginal Bone Loss. J Dent Res. 2014;93(7 Suppl):80S-85S. doi:10.1177/0022034513519800 (opens in a new tab) PMID: 24621670 (opens in a new tab).
  25. Blanco J, Pico A, Caneiro L, Nóvoa L, Batalla P, Martín-Lancharro P. Effect of abutment height on interproximal implant bone level in the early healing: A randomized clinical trial. Clin Oral Implants Res. 2018;29(1):108-117. doi:10.1111/clr.13108 (opens in a new tab) PMID: 29222809 (opens in a new tab).
  26. Spinato S, Stacchi C, Lombardi T, Bernardello F, Messina M, Zaffe D. Biological width establishment around dental implants is influenced by abutment height irrespective of vertical mucosal thickness: A cluster randomized controlled trial. Clin Oral Implants Res. 2019;30(7):649-659. doi:10.1111/clr.13450 (opens in a new tab) PMID: 31033035 (opens in a new tab).
  27. Valles C, Rodríguez-Ciurana X, Clementini M, Baglivo M, Paniagua B, Nart J. Influence of subcrestal implant placement compared with equicrestal position on the peri-implant hard and soft tissues around platform-switched implants: a systematic review and meta-analysis. Clin Oral Investig. 2018;22(2):555-570. doi:10.1007/s00784-017-2301-1 (opens in a new tab) PMID: 29313133 (opens in a new tab).
  28. Vervaeke S, Matthys C, Nassar R, Christiaens V, Cosyn J, De Bruyn H. Adapting the vertical position of implants with a conical connection in relation to soft tissue thickness prevents early implant surface exposure: A 2-year prospective intra-subject comparison. J Clin Periodontol. 2018;45(5):605-612. doi:10.1111/jcpe.12871 (opens in a new tab) PMID: 29359339 (opens in a new tab).
  29. Vervaeke S, Dierens M, Besseler J, De Bruyn H. The influence of initial soft tissue thickness on peri-implant bone remodeling. Clin Implant Dent Relat Res. 2014;16(2):238-47. doi:10.1111/j.1708-8208.2012.00474.x (opens in a new tab) PMID: 22758656 (opens in a new tab).
  30. Linkevicius T, Apse P, Grybauskas S, Puisys A. The influence of soft tissue thickness on crestal bone changes around implants: a 1-year prospective controlled clinical trial. Int J Oral Maxillofac Implants. 2009;24(4):712-9. PMID: 19885413 (opens in a new tab).
  31. Linkevicius T, Puisys A, Steigmann M, Vindasiute E, Linkeviciene L. Influence of Vertical Soft Tissue Thickness on Crestal Bone Changes Around Implants with Platform Switching: A Comparative Clinical Study. Clin Implant Dent Relat Res. 2015;17(6):1228-36. doi:10.1111/cid.12222 (opens in a new tab) PMID: 24673875 (opens in a new tab).
  32. Linkevicius T, Puisys A, Linkeviciene L, Peciuliene V, Schlee M. Crestal Bone Stability around Implants with Horizontally Matching Connection after Soft Tissue Thickening: A Prospective Clinical Trial. Clin Implant Dent Relat Res. 2015;17(3):497-508. doi:10.1111/cid.12155 (opens in a new tab) PMID: 24103157 (opens in a new tab).
  33. Puisys A, Linkevicius T. The influence of mucosal tissue thickening on crestal bone stability around bone-level implants. A prospective controlled clinical trial. Clin Oral Implants Res. 2015;26(2):123-9. doi:10.1111/clr.12301 (opens in a new tab) PMID: 24313250 (opens in a new tab).
  34. Linkevicius T, Linkevicius R, Alkimavicius J, Linkeviciene L, Andrijauskas P, Puisys A. Influence of titanium base, lithium disilicate restoration and vertical soft tissue thickness on bone stability around triangular-shaped implants: A prospective clinical trial. Clin Oral Implants Res. 2018;29(7):716-724. doi:10.1111/clr.13263 (opens in a new tab) PMID: 29855100 (opens in a new tab).
  35. Linkevicius T, Puisys A, Linkevicius R, Alkimavicius J, Gineviciute E, Linkeviciene L. The influence of submerged healing abutment or subcrestal implant placement on soft tissue thickness and crestal bone stability. A 2-year randomized clinical trial. Clin Implant Dent Relat Res. 2020;22(4):497-506. doi:10.1111/cid.12903 (opens in a new tab) PMID: 32250061 (opens in a new tab).
  36. Terzioglu B, Ayyildiz BG. Effect of supracrestal tissue height on marginal bone level changes at platform-switching dental implants placed crestally and subcrestally: A randomized clinical-trial. J Dent. 2024;148:105219. doi:10.1016/j.jdent.2024.105219 (opens in a new tab) PMID: 38960001 (opens in a new tab).
  37. Suárez-López Del Amo F, Lin GH, Monje A, Galindo-Moreno P, Wang HL. Influence of Soft Tissue Thickness on Peri-Implant Marginal Bone Loss: A Systematic Review and Meta-Analysis. J Periodontol. 2016;87(6):690-9. doi:10.1902/jop.2016.150571 (opens in a new tab) PMID: 26777766 (opens in a new tab).
  38. Díaz-Sánchez M, Soto-Peñaloza D, Peñarrocha-Oltra D, Peñarrocha-Diago M. Influence of supracrestal tissue attachment thickness on radiographic bone level around dental implants: A systematic review and meta-analysis. J Periodontal Res. 2019;54(6):573-588. doi:10.1111/jre.12663 (opens in a new tab) PMID: 31231802 (opens in a new tab).
  39. Canullo L, Camacho-Alonso F, Tallarico M, Meloni SM, Xhanari E, Penarrocha-Oltra D. Mucosa Thickness and Peri-implant Crestal Bone Stability: A Clinical and Histologic Prospective Cohort Trial. Int J Oral Maxillofac Implants. 2017;32(5):675–681. doi:10.11607/jomi.5349 (opens in a new tab) PMID: 28296981 (opens in a new tab).
  40. Thoma DS, Naenni N, Figuero E, Hämmerle CHF, Schwarz F, Jung RE, et al. Effects of soft tissue augmentation procedures on peri-implant health or disease: A systematic review and meta-analysis. Clin Oral Implants Res. 2018;29 Suppl 15:32-49. doi:10.1111/clr.13114 (opens in a new tab) PMID: 29498129 (opens in a new tab).
  41. Giannobile WV, Jung RE, Schwarz F, Groups of the 2nd Osteology Foundation Consensus Meeting. Evidence-based knowledge on the aesthetics and maintenance of peri-implant soft tissues: Osteology Foundation Consensus Report Part 1-Effects of soft tissue augmentation procedures on the maintenance of peri-implant soft tissue health. Clin Oral Implants Res. 2018;29 Suppl 15:7-10. doi:10.1111/clr.13110 (opens in a new tab) PMID: 29498127 (opens in a new tab).
  42. Lin GH, Chan HL, Wang HL. The significance of keratinized mucosa on implant health: a systematic review. J Periodontol. 2013;84(12):1755-67. doi:10.1902/jop.2013.120688 (opens in a new tab) PMID: 23451989 (opens in a new tab).
  43. Avila-Ortiz G, Gonzalez-Martin O, Couso-Queiruga E, Wang HL. The peri-implant phenotype. J Periodontol. 2020;91(3):283-288. doi:10.1002/JPER.19-0566 (opens in a new tab) PMID: 32027021 (opens in a new tab).
  44. De Bruyn H, Vandeweghe S, Ruyffelaert C, Cosyn J, Sennerby L. Radiographic evaluation of modern oral implants with emphasis on crestal bone level and relevance to peri-implant health. Periodontol 2000. 2013;62(1):256-70. doi:10.1111/prd.12004 (opens in a new tab) PMID: 23574471 (opens in a new tab).
  45. Chan AW, Tetzlaff JM, Altman DG, Laupacis A, Gøtzsche PC, Krleža-Jerić K, et al. SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med. 2013;158(3):200-7. doi:10.7326/0003-4819-158-3-201302050-00583 (opens in a new tab) PMID: 23295957 (opens in a new tab).
  46. Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c332. doi:10.1136/bmj.c332 (opens in a new tab) PMID: 20332509 (opens in a new tab).
  47. VanderWeele TJ. Mediation Analysis: A Practitioner's Guide. Annu Rev Public Health. 2016;37:17-32. doi:10.1146/annurev-publhealth-032315-021402 (opens in a new tab) PMID: 26653405 (opens in a new tab).
  48. Reference verification note. 47 of 47 references are PubMed-indexed; their bibliographic records (authors, title, journal, year, volume, pages, DOI, PMID) were retrieved programmatically from the PubMed database on November 17, 2026. Database retrieval confirms that each record exists as cited; it does not replace full-text reading by the responsible authors of every statement attributed to a source.