Measurement of wear in retrieved acetabular sockets

How to measure wear following total hip arthroplasty

Total hip arthroplasty has yielded excellent results in decreasing pain and enhancing function in patients with hip degenerative disease. However, the problems associated with prosthetic failure and the consequent need for revision surgery still represent a major clinical issue. The most common reasons for revision surgery include implant loosening, periprosthetic osteolysis, infection, malalignment, stiffness, implant failure or fracture, and wear. The need for eliminating or reducing wear plays a crucial role in refining prosthesis composition and design. In this regard, it is important to develop new techniques for more accurate and reproducible measurement of wear. This should allow earlier detection of increased wear and thus permit earlier identification of patients who are at risk, and also help to identify faulty implant designs.
This work is aimed at discussing the most common in vivo and in vitro methods used for evaluating the wear of hip prosthesis components.

Wear patterns of, and wear volume formulae for, cylindrically elongated acetabular cup liners

This study analyzed the wear patterns of, and wear volume formulae for, cylindrically elongated acetabular cup liners. The geometric patterns of the wear surface were first classified, then wear volume formulae were derived by integral calculus. SolidWorks((R)) software or published formulae were used to verify the accuracy of the proposed formulae. The analytical results showed that the wear shape of the liner can be categorized into seven wear patterns, including the special case of wear at 90 degrees , and the seven corresponding wear formulae were derived. In addition, wear of the cylindrical elongation might add considerably to the volume loss of the liner, depending on the height and shape of the elongation and the depth and direction of the linear penetration, being maximally 21% in the investigated model. The proposed wear formulae and patterns will be useful for more accurate performance evaluation of existing hip components implanted in patients and for the designing of new hip components.

Evaluating the accuracy of wear formulae for acetabular cup liners

This study proposes two methods for exploring the wear volume of a worn liner. The first method is a numerical method, in which SolidWorks software is used to create models of the worn out regions of liners at various wear directions and depths. The second method is an experimental one, in which a machining center is used to mill polyoxymethylene to manufacture worn and unworn liner models, then the volumes of the models are measured. The results show that the SolidWorks software is a good tool for presenting the wear pattern and volume of a worn liner. The formula provided by Ilchmann is the most suitable for computing liner volume loss, but is not accurate enough. This study suggests that a more accurate wear formula is required. This is crucial for accurate evaluation of the performance of hip components implanted in patients, as well as for designing new hip components.

Long-term radiographic assessment of cemented polyethylene acetabular cups

In vitro studies demonstrating excessive wear in polyethylene cups sterilized using gamma irradiation and stored in air led to the abandonment of this sterilization technique. We evaluated the clinical wear performance of a metal femoral component on a polyethylene cup in a hip prosthesis from a selected subset of implants in a group of patients followed for at least 20 years and assessed the time dependency of variation in penetration rates. We measured penetration in 33 polyethylene cups in 25 patients who had a Charnley low-friction arthroplasty between 1982 and 1984. All patients had Charnley Ogee cups implanted for more than 20 years and sterilized using the gamma irradiation in air technique. If degradation occurred over time in vivo, it was not reflected by an increased penetration rate with increasing time in vivo; even after 20 years of implantation, the degree of wear remained low. This suggests gamma irradiation affects wear on ultra-high-molecular-weight polyethylene by reducing wear secondary to the crosslinking, by increasing wear as shown through in vitro studies of heavily oxidized samples, or by oxidation resulting from prolonged shelf life. The effect of progressive oxidation in vivo does not appear to affect wear in vivo.

LEVEL OF EVIDENCE

Level IV, therapeutic study. See the Guidelines for Authors for a complete description of levels of evidence.

Precision of wear measurement using the shadowgraph technique

The shadowgraph technique is a relatively easy-to-use and inexpensive method of wear measurement from explanted acetabular cups. In this technique, from a mould of the internal surface of the cup, measurements of linear wear and wear angle can be obtained, from which volumetric wear can be calculated. In this study the measurement precision of this technique was assessed and the influence of different observers and multiple moulds was also determined. It was found that linear wear (coefficient of variation (CV) = 1.49 per cent) can be more precisely determined than wear angle (CV = 8.18 per cent) and that both the observer and the mould can significantly influence the results obtained, although the influence of the mould is considerably less than that of the observer.

Association between dislocation, impingement, and articular geometry in retrieved acetabular polyethylene cups

We used a novel analysis technique and retrieved acetabular polyethylene liners to investigate relationships between impingement damage, dislocation history, and polyethylene liner geometry. Forty-eight polyethylene liners with neutral rim elevation were visually assessed for peripheral rim damage consistent with impingement between the liner rim and femoral component. Liner articular geometry was measured using a digital stylus, and the geometric relationships between the liner rim and center of the femoral head were characterized by lip height, head center inset, and head penetration into the polyethylene liner. Thirteen (27%) retrieved liners had rim impingement damage and 12 (25%) had a history of dislocation. The proportion of liners with impingement was not significantly different among dislocated and stable liners. Lip height, head center inset, and head penetration were not significantly different between liners with and without impingement damage. Dislocated liners had a significantly shorter head center inset, with significantly less head penetration because of shorter functional duration. Designing polyethylene liners with a sufficiently deep articular surface that exceeds 0.95 mm may prove beneficial for decreasing the prevalence of early dislocation, independent of impingement damage.

Wear analysis of failed acetabular polyethylene: a comparison of analytical methods

There are many methods for analysing wear volume in failed polyethylene acetabular components. We compared a radiological technique with three recognised ex vivo methods of measurement. We tested 18 ultra-high-molecular-weight polyethylene acetabular components revised for wear and aseptic loosening, of which 13 had pre-revision radiographs, from which the wear volume was calculated based upon the linear wear. We used a shadowgraph technique on silicone casts of all of the retrievals and a coordinate measuring method on the components directly. For these techniques, the wear vector was calculated for each component and the wear volume extrapolated using mathematical equations. The volumetric wear was also measured directly using a fluid-displacement method. The results of each technique were compared. The series had high wear volumes (mean 1385 mm(3); 730 to 1850) and high wear rates (mean 205 mm(3)/year; 92 to 363). There were wide variations in the measurements of wear volume between the radiological and the other techniques. Radiograph-derived wear volume correlated poorly with that of the fluid-displacement method, co-ordinate measuring method and shadowgraph methods, becoming less accurate as the wear increased. The mean overestimation in radiological wear volume was 47.7% of the fluid-displacement method wear volume. Fluid-displacement method, coordinate measuring method and shadowgraph determinations of wear volume were all better than that of the radiograph-derived linear measurements since they took into account the direction of wear. However, only radiological techniques can be used in vivo and remain useful for monitoring linear wear in the clinical setting. Interpretation of radiological measurements of acetabular wear must be done judiciously in the clinical setting. In vitro laboratory techniques, in particular the fluid-displacement method, remain the most accurate and reliable methods of assessing the wear of acetabular polyethylene.

Estimation of the wear volume after total hip replacement. A simple access to geometrical concepts

Various formulas have been proposed to calculate the volume of prosthetic wear from the penetration depth of the head as assessed on plain radiographs, based on idealized, three-dimensional geometrical models of a prosthetic hip. However, for most published formulas no (or no simple) derivation is available and not all of them are correct. We describe a simple geometrical model that allows for transparent derivation of equations for various components of prosthetic wear volume and compare the calculated volumes with those obtained from published equations. These components are: (1) a right generalized cylinder resulting from a linear shift of a half spherical part of the prosthetic head into the hemispherical cup, (2) an additional wedge that is "cut" from the cup and (3) the wear from an optional additional cylindrical portion of the cup. We emphasize that calculation of a three-dimensional wear volume from linear penetration depth should be based on a geometrical concept that is transparent and simple enough for clinical research, such as the one presented. The incorrect formula of Kabo et al. should be completely abandoned.

Validation of a digital image processing software package for the in vivo measurement of wear in cemented Charnley total hip arthroplasties

Computer-generated images were used to assess image processing software employed in the radiographic evaluation of penetration in total hip replacement. The images were corrupted using Laplacian noise and smoothed to simulate different modulation transfer functions in a range associated with hospital digital radiographic systems. With no corruption, the penetration depth measurements were both precise and accurate. However, as the noise increased so did the inaccuracy and imprecision to levels that may make changes in the penetration observed clinically difficult to discern between follow-up assessments. Simulated rotation of the wire marker produced significant bias in the measured penetration depth. The use of these simulated radiographs allows the evaluation of the software used to process the digital images alone rather than the whole measurement system.

Validation of a micro-CT technique for measuring volumetric wear in retrieved acetabular liners

In this study, a novel micro-CT-based technique for evaluating wear in retrieved acetabular liners was introduced and validated. Six UHMWPE acetabular components ranging in implantation time from 2.7 to 14.4 years were collected and evaluated with the use of a high-resolution micro-CT scanner. The components were scanned with a uniform volumetric resolution of 74 microns (16-bit precision) with the use of a 1,024 x 1,024 in-plane image matrix. Manual rigid 3D image registration of the interior hemispherical portion of the acetabular cup with geometric primitives by trained observers allowed for isolation, visualization, and measurement of the wear volume. Results for these six components indicated an average wear rate of 65 mm(3)/year. Overall scanner error was quantified gravimetrically and associated with a maximum uncertainty of 0.6%. Intra-- and interobserver uncertainty analysis showed the method to be both accurate and repeatable.

Validation of acetabular cup wear volume based on direct and two-dimensional measurements: hip simulator analysis

The volumetric wear in retrieved cups can be assessed by mathematical conversion based on linear measurements and by a fluid-displacement method. We used a hip simulator model to produce wear in 22-, 28-, and 32-mm hip implants and then assessed the volumetric wear using a gravimetric wear method. We then compared the findings with those obtained with the linear and fluid-displacement methods. For the linear method, we translated the linear wear to the volumetric wear using the equations developed by Charnley et al., Kabo et al., and Hashimoto et al. The fluid-displacement method showed the strongest correlation with the gravimetric wear method, and it was found to overestimate the volume slightly (by 3%-9%). According to the linear wear conversion, however, the equation by Kabo widely underestimated the volume by 33%-40%. The equation used by Charnley tended to overestimate the volume (by 4%-17%), whereas Hashimoto's equation tended to slightly underestimate the volume (by 2%-12%). The fluid-displacement method demonstrated an average error of 0.34% +/- 13.40% when the wear exceeded 400 mm(3). The linear wear was thus converted to the volume wear most accurately using Hashimoto's equation, with the average error being -3.8% +/- 14.0%. Of the four measurement modalities, the fluid-displacement method showed the most accurate results. We therefore confirmed that the fluid-displacement method is the most accurate way to determine volumetric wear in retrieved cups.

The nature and dissemination of UHMWPE wear debris retrieved from periprosthetic tissue of THR

The role of wear debris in provoking joint replacement failure through bone resorption is now supported by much research. This study presents the analysis of 104 tissue samples using laser diffraction wear particle analysis in conjunction with standard histologic methods. The number and volume distributions were correlated to a range of joint and patient parameters. The median particle diameter by number was 0.69 microm. No particles smaller than 0.113 microm were resolved. No variation in terms of particle distribution was found among joint types. The ability of particles to migrate away from their point of origin was found to be inversely proportional to their size. The numbers of particles per gram of tissue found in various regions around the prosthesis varied little. Further, the numbers of particles in tissue samples shown to have a chronic foreign-body reaction was > 1 x 10(9) particles/gram.

The effect of socket design, materials and liner thickness on the wear of the porous coated anatomic total hip replacement

The wear of joint replacement prostheses represents the greatest challenge to their continued development. Parameters such as polyethylene quality, liner thickness and metal backing have all been implicated as potential detractors in the search for the lowest-wearing socket. This study examined the effect of these parameters through an extensive study of the two versions of the porous coated anatomic (PCA) hip prosthesis (one-piece socket and snaplock socket). For the whole cohort the wear rate was found to be 88 (SE 10) mm3/year and the clinical wear factor was 2.00 (SE 0.28) x 10(-6) mm3/N m. When the two socket types were investigated individually, the wear factors found were 2.39 (SE 0.44) x 10(-6)mm3/N m and 0.99 (SE 0.25) x 10(-6) mm3/N m for the one-piece and snaplock, respectively. This illustrates that the metal backing per se does not predispose these sockets to rapid wear. The good wear performance of the snaplock liner may be attributed to the high quality of the ultra-high molecular weight polyethylene (UHMWPE) used and the shorter implantation period compared to that for the one-piece design. No correlation was found between the thickness of the liner and the clinical wear factor. Within the range of thicknesses tested here, UHMWPE thickness is not an influential parameter for the hip prosthesis and this is confirmed

Variation in the wear rate during the life of a total hip arthroplasty: a simulator and retrieval study

The limitation of wear is fundamental to the optimization of total hip arthroplasty longevity. The maintenance of the supersmooth femoral head surface is considered to be paramount in maximizing prosthesis life expectancy. Ex vivo studies have failed to substantiate a relationship between roughness and the clinical wear factor, however. A hip simulator wear study was undertaken to investigate this contradiction. Three explanted femoral heads were articulated for 5 million cycles against new acetabular liners. The simulator wear rate was 5 times the ex vivo value. This difference can be explained only if the explant head roughness was not that which existed for most of the joint's life. The relationship between surface roughness and wear deduced for simulator testing is substantially different from that of unidirectional wear screening methods. The multiphasic nature of wear in cementless joints has been illustrated: a wear-in period, followed by a steady-state phase, until a head-roughening event causes a rapid wear period.

Prevention of loosening in total hip replacements using guided bone regeneration

The aim of this study was to prevent wear debris from reaching the interface of the acetabular cup and femoral component by using a partially occlusive expanded polytetrafluoroethylene membrane. This membrane initially acted as a physical seal, which became incorporated by bone and soft tissue, forming a secondary biologic seal, preventing loosening. An animal model was developed to test the hypothesis. The model replicated the mechanisms of loosening where the effects of wear debris were studied. Using femoral heads with the appropriate roughness, a goat model produced the radiologic and histologic presentation of loosening as observed in total hip replacements in humans. Loosening was assessed by measurement of the radiolucent lines, and was attributed to wear debris by histologic investigation. The expanded polytetrafluoroethylene membrane prevented acetabular implant loosening to a statistical significance of 0.02 in a blinded assessment when compared with the control groups. Loosening of the first 5 mm of the proximomedial aspect of the femur also was prevented. The authors of the current study prevented wear particle-induced osteolysis in the acetabular component by using an expanded polytetrafluoroethylene membrane to seal the bone-cement interface.

Evaluation of a hip joint simulator

To evaluate the functioning of the Durham hip joint wear simulator, the wear rates of ultra high molecular weight polyethylene (UHMWPE) and polytetrafluoroethylene (PTFE) acetabular cups articulating against 22 mm diameter cobalt-chromium-molybdenum (CoCrMo) femoral heads were studied. A wear test was conducted in a lubricant of distilled water at 37 degrees C for a duration of 4.8 million cycles. The average penetration rate for the CoCrMo femoral heads against UHMWPE acetabular cups was 0.03 mm/10(6) cycles, while penetration rate for PTFE cups was some twenty times greater. These results are of a similar order of magnitude to other simulator studies in distilled water and are in a similar ratio to clinical data.

A comparison between gravimetric and volumetric techniques of wear measurement of UHMWPE acetabular cups against zirconia and cobalt-chromium-molybdenum femoral heads in a hip simulator

Five cobalt-chromium-molybdenum (CoCrMo) and five zirconia femoral head components have been wear tested against 28 mm diameter ultra high molecular weight polyethylene (UHMWPE) acetabular cups for 5 million cycles in the Durham hip joint wear simulator using bovine serum as a lubricant. Wear measurements used gravimetric and volumetric techniques and no statistically significant difference was found between the measurement methods. The wear rates of the acetabular cups against both femoral heads are presented for both measurement methods. The UHMWPE acetabular cups showed a statistically significant higher linear wear rate for the first 2 million cycles than the lower linear wear rate from 2 million cycles to the end of the test, against both femoral head materials. Over the full duration of the wear test, the wear rates of acetabular cups articulating against zirconia femoral heads were lower than against CoCrMo femoral heads. The wear rates up to 2 million cycles and from 2 to 5 million cycles for both femoral head materials were consistent with other studies.

The influence of femoral head surface roughness on the wear of ultrahigh molecular weight polyethylene sockets in cementless total hip replacement

A theoretical relationship was recently proposed relating the wear behavior of polymetric bearing materials articulating against hard counterfaces.(1) This model attempts to predict the influence of surface roughness on wear. Laboratory-based studies have been used to establish the validity of these relationships, but their application to the clinical situation has not been investigated fully. Forty-two retrieved PCA hip joints have been assessed. The total wear volume was calculated from the penetration measured using the shadowgraph method, and roughness of the articulating surfaces was recorded using noncontacting profilometry. The roughness of the explanted femoral heads was observed to increase (median S(a) - 10. 35 nm worn region, 3.05 nm peripheral region), while that of the acetabular liner fell dramatically (median S(a) - 41 nm worn region, 212 nm unworn region). No evidence of a relationship between the topography of the worn regions of the femoral head and that of the acetabular liner could be found. Similarly, the strength of the association between the surface roughness and the clinical wear factor was considerably poorer than that achieved in laboratory experiments. A number of reasons for this observation are proposed. Most deleterious was considered to be the inability of the roughness parameters to describe the damaging features of the surface adequately. Uncertainty as to when the surface of the component degrades during its life serves to introduce further doubt as to the application of the wear models in the clinical environment. In conclusion, this study fails to provide clinical evidence to substantiate the relationship between surface finish and wear rate. The adoption of standardized measurement parameters and techniques would facilitate the direct comparison of joint types and the selection of the most advantageous materials.

The association between rates of wear in retrieved acetabular components and the radius of the femoral head

Current evidence suggests that loosening of the acetabular socket is related to the volume of ultra-high molecular weight polyethylene wear debris generated at the articulating surfaces, through a process of bone resorption. Therefore it is important that the rate of volumetric wear is minimized in an attempt to extend the useful life of the procedure. Laboratory evidence indicates that a reduction in sliding distance would be beneficial in achieving this target and may be attained by a reduction in femoral head radius. To investigate the relationship between femoral head size and the rates of both volumetric wear and penetration, 200 acetabular components were retrieved at the time of revision surgery. The joints had femoral heads ranging in size from 11.1 to 19.8 mm radius. For those sockets that were loose at revision surgery, a significant correlation was observed, between the rate of volumetric wear and the radius of the femoral head. For this cohort an increase in radius of 1 mm resulted in an increased rate of volumetric wear of 5.1 (SE 1.4) mm3/yr. However, the explained variance in the regression was low and exemplifies the multifactorial nature of the wear process. In particular, it is anticipated that the activity of the patient will have a significant effect on the rate at which the debris is produced. No significant correlation was observed between the rate of linear wear and femoral head radius. These results would indicate a benefit in using head sizes of a smaller radius, which generate debris at a reduced rate, and therefore require more time to accumulate large volumes of wear products.

The estimation of acetabular cup wear volume from two-dimensional measurements: a comprehensive analysis

This paper describes a mathematical investigation of the relationship between wear volume, wear depth and wear direction in acetabular components. The analysis takes into account the cylindrical and conical portions at the mouth of certain types of socket and also incorporates the effect of an initial radial discrepancy between the femoral head and socket. Published formulae for converting linear wear measurements to wear volumes are shown to be incorrect. Wear volume is shown to be highly dependent on the wear direction, increasing by more than 90 per cent over a 60 degrees range. Cylindrical and conical portions of a 22.225 mm cup may contribute up to about a third as much wear as the hemispherical socket. At low wear depths, the neglect of a radial discrepancy between the components can result in an overestimation of wear volume in excess of 100 per cent.

Prevalence of impingement in explanted Charnley acetabular components

: Impingement of the femoral neck against the rim of the socket bore has often been cited as one of the contributory factors in the acetabular loosening process. However, there has been little research on its clinical prevalence or on the effects of both linear wear and the diameter of the femoral neck. With this aim in mind, 74 Charnley prostheses were examined after being removed at revision surgery and the sockets interrogated for evidence of impingement. The probability of impingement was assessed using logistic regression analysis. A strong positive association was observed between penetration depth and impingement (chi2 = 12.8; P = 0.0004) regardless of differences in the femoral neck diameter. Further, the introduction of femoral components which comprised a reduced diameter neck had a positive effect, in that the 50% probability of impingement occurred at approximately 2 mm of penetration. For those components with standard necks, the 50% probability of impingement occurred at zero mm of penetration. If impingement is a problem then a reduced diameter neck would appear to be a solution in cutting rates of long-term loosening. However, whether or not this reduction in rim damage, and therefore impingement, is clinically significant in terms of loosening can only be fully assessed from long-term survival analysis and comparison with autopsy retrieved specimens.

Discrepancy between penetration depths derived from radiographic and direct measurement of acetabular components

The most common technique for assessing penetration due to wear in acetabular components is with the aid of the most recent serial radiograph. This approach, which is often termed the uni-radiographic method, has been shown to underestimate the more reliable value of the penetration depth deduced from direct measurement of explanted sockets. In this article the causes of the discrepancies between the two data sets are explored. Ninety-six sockets were available from revision surgery for which both the penetration depth and angle could be measured using the shadowgraphic technique in both the coronal and wear planes. Further, the penetration depth for each of the sockets was also assessed from pre-revision X-rays. A significant discrepancy was observed between the penetration depths measured in the wear plane of the replica delta Pw and that measured from the radiograph, delta PX-ray. The discrepancy was greatest for loose sockets as opposed to those that were still fixed at revision surgery. Using the corresponding data from the shadowgraph measurements, it was possible to deduce that the errors have arisen from the radiographic measurement of wear in the coronal plane and the formula used in calculating delta PX-ray. If these errors (which cannot be calculated from the X-ray data alone) were taken into consideration, then the systematic bias between radiographic and shadowgraphic measurement was greatly reduced. The largest portion of the discrepancy was accounted for by wear occurring out of the plane of the radiograph, and this, in general, coincides with the coronal plane. Overall, these results indicate that the accurate measurement of wear from serial radiographs is not possible and that improved performance in terms of accuracy can only be achieved when a three-dimensional system is used. An alternative method for deducing the radiographic penetration depth is proposed which, theoretically, negates the error arising from the inaccuracy of the formula.

The effect of surface topography of retrieved femoral heads on the wear of UHMWPE sockets

The study was undertaken to investigate the association, if any, between the surface roughness of 35 explanted femoral heads and the clinical wear factors of the corresponding polyethylene sockets. The wear of the socket was measured using a shadowgraph technique. The surface topography was investigated using a Rodenstock RM 600 non-contacting profilometer, and six parameters were used to characterise the roughness. Further, qualitative inspections of the femoral surface were undertaken using both a Joel JSM-IC-848 scanning electron microscope and a Zeiss Axiotech microscope with a differential interference contrast facility. Two parameters were found to correlate with the clinical wear factor, namely the skewness of the amplitude distribution function, Sk, and the arithmetic mean roughness, Ra. Simple parameters describing the peak heights of the asperities were found not to have a significant association with the clinical wear factor. The exponent in the power relationship between the arithmetic mean roughness and kclinical was found to be equal to 0.5 (SE: 0.2). This value is significantly smaller than that found in laboratory experiments and may be due to the non-uniform nature of the roughening of the femoral head, three-body wear and the effect of other clinical factors on the wear process. Further, extremely strong correlations were detected between the differing roughness parameters, which would suggest that any attempt to deduce which one is the most important in affecting the wear of the polymeric counterface is fraught with difficulty. However, further investigation of those parameters, such as the reduced peak height or the material filled profile peak area, which may better describe the effect of the counterface on the wear of the UHMWPE surface would appear to be prudent.

Wear in retrieved Charnley acetabular sockets

One hundred and twenty-nine Charnley acetabular components were acquired at the time of revision surgery and a tribological investigation undertaken. The relative occurrence of pitting in the unworn and worn regions of the sockets suggest that most of the cement ingress occurs during the early part of the service life. The penetration depth of the explanted sockets was determined using the shadowgraph technique. Observation of the profiles in the wear planes suggest that, in general, the creep component was not a significant proportion of the overall change in the inner bore of the socket. Using weighted ordinary least squares regression, in which the intercept was not assumed to be zero, mean penetration and wear volume rates of 0.02 (SE = 0.02) mm/year and 55 (SE = 5) mm3/year, respectively, were recorded and are in agreement with other retrieval studies. In neither case was the intercept found to be significantly different from zero. A mean clinical wear factor, Kclinical, equal to 2.1 (SE = 0.2) x 10(-6) mm3/N m was calculated which is considerably larger than that found in laboratory experiments which purport to reflect in vivo conditions. In this analysis, a significant positive intercept was observed [96 (SE = 36) mm3] and may be evidence of the small initial penetration due to creep reported in simulator experiments. A strong positive association between kclinical and the arithmetical mean roughness, Ra, of the femoral head was also demonstrated although the rate of change was not as great as that cited for laboratory experiments.