Secondary Logo

Journal Logo


The Relationship of Prosthetic Sock Ply Thickness to Percentage of Transtibial Limb Volume Outside of the Socket

Jasken, Jenna CPO; Hall, Michelle MS, CPO, FAAOP(D)

Author Information
Journal of Prosthetics and Orthotics: July 2018 - Volume 30 - Issue 3 - p 140-144
doi: 10.1097/JPO.0000000000000193
  • Free


Prosthetists aim to deliver a comfortable and intimately fitting socket for each patient. Throughout the past decade, total surface bearing (TSB) sockets for individuals with transtibial amputation have gained popularity. The American Board for Certification in Orthotics, Prosthetics & Pedorthics (ABC) 2015 Practice Analysis of Certified Practitioners in the Disciplines of Orthotics and Prosthetics indicated that 39.2% of individuals with transtibial amputations use TSB sockets and 17.8% use patella-tendon bearing (PTB) sockets.1 The TSB socket design uses a liner made of urethane, silicone, or other gel-like materials. Compared with the traditional PTB socket that loads soft tissue and relieves bony prominences, the TSB socket loads the entire limb more equally across its surface.2 The use of the liner helps with this force distribution in TSB sockets.2 Depending on the manufacturer and patient's limb rigidity, different amounts of positive model reduction are necessary for a properly fitting socket. For example, the Össur Iceross Original liner has a volume reduction of 3% to 5% proximally and gradually decreases to a 0% distally based on circumferential measurements.3 Ohio WillowWood also recommends a volume reduction when using its Alpha Basic, Classic, and Hybrid liners. For a 3-mm liner, the global reduction is 3% to 5%; for a 6-mm liner, 5% to 7%; and for a 9-mm liner, 7% to 9%.4 These reductions may be achieved by using plaster or digital (computer-aided design and manufacturing [CAD/CAM]) methods.

Similarly, throughout the fitting process, prosthetists use sock ply, or thickness, to assess the fit of the socket and then increase or decrease the overall volume accordingly to achieve the best fit. Ply is defined as the number of single yarns twisted together to make a fiber-like material; this material is subsequently woven into a sock fabric.5 With new materials such as cotton and polyester, the thickness between one ply of the new materials is not the same thickness as one ply of wool.5 There is limited research on the thickness of each ply and how it relates to volume.

When reducing the socket volume, prosthetists using plaster models may place a sock within the check socket and fill it with plaster or fill the socket without a sock and reduce the percentage volume by removing plaster by hand. Those using CAD/CAM may digitally decrease the model by a percentage of the original volume. To increase the size of the socket, prosthetists using plaster may fill the check socket, place the appropriate sock ply over the positive model, cast over the model and sock, and then fill the new cast, whereas those using CAD/CAM may increase the socket volume digitally. These methods attempt to alter the volume of the socket to mimic the effect of sock ply on the patient's limb. They rely heavily on clinical experience.

Previous studies have attempted to quantify various aspects related to transtibial limb volume, socket fit, and sock ply thickness. However, none have provided guidelines for approximating volumetric change due to the use of prosthetic socks.6–8 This information may be helpful to clinicians in determining socket volume changes necessary to optimize the fit for each iteration.

Bolt et al.6 analyzed variation in transtibial residual limb model volumes by randomly measuring nine transtibial models that were fabricated by OIM Orthopedie (Orthopedische Instrument Makerij, Assen, the Netherlands) using prosthetic casts of patients who had undergone amputation. Each model had a total of 40 measurements using five different measuring techniques: water immersion, the Design TT system, the Omega Tracer system, circumferential measurements, and anthropometric measurements.

The Design TT system was developed for manufacturing a prosthetic socket for persons with transtibial amputation by using two digital photographs (one frontal and one lateral) of the residual limb against a black background.6 A calibration device was placed on the model to calculate the right distances and volumes from the reference points.6 The photographs were loaded into the software, and from there, a three-dimensional socket model was calculated.6 The anthropometric measurements taken were length from the most proximal location (anterior dot) to the distal end; length from the proximal location to the widest or smallest diameter; anteroposterior distance at the most proximal, at the widest/smallest, and at the most distal location; and lateromedial distance at the most proximal, at the widest/smallest, and at the most distal location.6

Variables of the study were the sock ply applied to each model and how much stretch was applied to each sock when it was donned. The results showed that the Omega Tracer had a low repeatability coefficient (45 mL), whereas the anthropometric measurements had a higher repeatability coefficient (155 mL).6 A repeatability coefficient of 45 mL indicates that there is a 95% chance that a next measurement result with this method will fall within 45 mL of the initial measurement, independent of observer, session, or occasion.6 The repeatability coefficients show that the CAD/CAM system is a very reliable instrument and the anthropometric measurements are rather imprecise.6 Based on this, future studies interested in volumetric measures should consider use of a scanning system.

Zachariah et al.7 attempted to characterize the magnitude and distribution of volume changes in persons with transtibial amputation upon socket doffing. Residual limbs were measured using an optical scanner twice: upon removal of the prosthesis and at a 2-week interval. The largest source of error they found was the tremor-like movements of the residual limb.7 Once the limb had stabilized, the proper sock ply was donned and the residual limb was measured. They found that the volume change was highest immediately upon on socket doffing and then stabilized after 35 minutes of socket removal.7 Along with limb stabilization, they also found that there were not consistent proximal-to-distal trends in volume increase, which suggests that prosthetic socks may be adequate for within the day volume control.7

Cagle et al.8 conducted a study testing to see if the thickness of one multiply sock of ply P had the same thickness of a stack of reduced-ply socks of total ply P. In addition, they examined if the thickness of single sock stacked on top of the other equals the sum of the single multiply sock thickness.8 The socks were tested using a custom-designed instrument. Six different sock models from Knit Rite ranging from 1 to 6 plies were tested.8 The results showed that a single multiply sock tended to be thinner than a stack of socks of that added up to the same ply number. Three 1-ply socks were 20% greater in thickness than one 3-ply sock.8 A limitation to this study was that the socks used had never been worn in a prosthesis. However, they had been tested in a previous study.8 This implies that these socks had undergone mechanical stress, which could result in an underestimation of the thickness compared with a sock that had not undergone mechanical stress. Those authors did not examine these effects on overall volume.

This study attempts to quantify and compare the percentage of volumetric changes caused by the application of prosthetic socks to foam prosthetic limb models. This inquiry was pursued in the hope to provide initial guidance to prosthetists in determining the amount to enlarge/reduce a socket based on sock ply fit.


Three sample foam models of the residual limbs (A, B, and C) were randomly scanned with each combination of sock ply. The resultant volumetric change from baseline was calculated. The three sample foam models differed in length and circumference. The sample foam positive models were fabricated for this study based on a convenience sample of positive plaster models previously used to fabricate transtibial check sockets for prosthetic patients in our clinic. Sample foam models were used to eliminate the confounding effects of individual patients whose volume may fluctuate during the testing period.

At the time of the study, our clinic was not using CAD/CAM technology for prostheses, but was for orthoses. Therefore, all scanning was completed by the one experienced orthotist, who was familiar with the CAD/CAM system (Rodin4D Global CAD/CAM solution for CPO, France). The Rodin4D CAD/CAM system was used for all trials. The absolute accuracy of this scanning system is 0.75 mm and the practical accuracy is 0.13 mm.9 One experienced prosthetist applied socks to each model and was blinded to the sock ply number. The prosthetist was instructed to apply the socks as consistently as possible in a manner similar to that which he would when working with a patient during a fitting. Neither the orthotist nor prosthetist was an investigator in this study, so both were blinded to the purpose of this study.

For consistency of donning the socks, similar to that on an actual patient, each sample foam model had an appropriately sized liner (Össur Americas, 3 mm Iceross Comfort Cushion, Foothill Ranch, CA) applied to it. This type of liner may be used in clinical practice with a TSB socket. The same liner remained on each model throughout the duration of the study to avoid any differences in the way the liner was donned. A volumetric scan of the liner (0-ply), while applied to the model, was taken six times as part of the random order of the trials. The average of these liner trials was used as each model's baseline.

For each trial, appropriately sized prosthetic sock(s) (Royal Knit, Summit, MO, USA) were randomly applied to each model over the liner by the experienced prosthetist, then were scanned by the experienced orthotist and removed by the research investigator before the next trial. Ten different individual or combined sock trials (zero, one, two, three, five, one + one, one + two, two + three, two + one, and three + two ply) were completed on each model. Each sock combination was repeated six times and the average was used for comparisons. The 180 trials (60 per model) were completed in a random order, as determined through a random list generator,10 over 4 days owing to time constraints of the participating clinicians.

For a consistent volume measurement for each trial, the distal patella was marked with a plastic rivet and taped to the liner with double-sided tape. A metal pipe was affixed to each model to hold the model stationary and horizontal for all trials. Plastic stands were fabricated to hold the models for this study and the metal pipes were covered with additional polyvinyl chloride (PVC) pipe. Plastic was used in these instances to avoid scan distortion from the proximity of the metal.

Each volume measurement was normalized by the baseline measurement to compare the differently sized models. A volumetric change from baseline was calculated for each individual or combined sock test. It was then normalized by dividing by the baseline measurement. These were calculated using Microsoft Excel (Microsoft Office 2010).

Custom software was developed using MATLAB (R2015b; MathWorks, Natick, MA, USA) for statistical analyses. One-way analysis of variance (ANOVA) tests were used to compare the volumetric differences in sock ply or baseline with each other, as well as to compare sock ply of a single sock with the same ply from multiple socks. A significance level of 0.05 was used. Tukey-Kramer post hoc analyses were performed in instances where significant differences were found.

Although all of the socks were applied by a single trained prosthetist who was instructed to don the sock consistently, it is unknown if the socks were consistently stretched during donning. This potential variance due to human error was not measured. Any such errors would have likely served to stretch the sock and thereby reduce the resultant volumetric change that occurred through application. Although the use of the mean of multiple randomized trials and models should have accounted for this difference, errors may still have occurred. Future studies may wish to measure the resultant sock stretch and standardize sock application, perhaps through use of a machine that could measure the tension of each sock donned.


The normalized average mean volume data are found in Table 1. The volumetric percentage change was calculated as the difference between the sock ply volume and baseline volume, divided by baseline volume and then multiplied by 100. This is also listed in Table 1.

Table 1
Table 1:
Percentage of normalized volumetric mean from baseline

Comparisons of single socks of differing plies did not necessarily correspond to a greater volumetric change (Figure 1). The percentage volumetric change of a one-ply sock was statistically smaller than that of a three-ply (p = 0.0004) or a five-ply (p = 0.0206) sock. However, no differences were found when comparing a one-ply with a two-ply (p = 0.3818) sock, a two-ply with a three-ply (p = 0.0461) sock, and a two-ply with a five-ply (p = 0.5234) sock.

Figure 1
Figure 1:
Comparison of the effect of sock ply thickness on proportional difference in volumetric change (ccm) from the baseline. *Significant difference between one ply versus three ply and one ply versus five ply.

The order of sock application for the same ply did not result in a significant difference in volumetric change from baseline (one + two ply vs. two + one ply, p = 0.9039; two + three ply vs. three + two ply, p = 0.1004). The use of multiple socks versus a single sock for the same ply varied in the volumetric results. The combination of two + three–ply and three + two–ply socks resulted in a significantly greater volume change than using a single five-ply sock (p < 0.0001; Figure 2). Conversely, no statistically significant differences were observed when comparing the volumetric change for the application of a single three-ply sock versus a one + two–ply or two + one–ply sock (p =0.1117; Figure 3). There was also no significant difference when comparing the volumetric change for application of a single two-ply versus a one + one–ply sock (p = 1.182).

Figure 2
Figure 2:
Comparisons of proportional differences in volumetric change (ccm) of a single five-ply sock with two + three– and three + two–ply sock combinations. *Significant difference between a two + three–ply or three + two–ply sock compared with a single five-ply sock.
Figure 3
Figure 3:
Comparison of proportional difference in volumetric change (ccm) of a single three-ply sock with one + two– and two + one–ply sock combinations. No significant difference was found between these comparisons.


This study determined that there are significant differences in the resultant percentage of volumetric changes in models of transtibial limbs outside of the socket based on sock ply thickness. Depending on the various sock ply and combination of multiple socks, significant differences were found. The order of application of the socks was found not to have significant differences.

Our results showed that multiple sock ply combinations, specifically two + three and three + two ply, are significantly greater in resultant volumetric change than use of a single five-ply sock. However, the order of application did not show a significant difference. Conversely, there were no significant differences in volumetric change when comparing a single three-ply to a one + two–ply or a two + one–ply sock. The amount of stretch of each sock was not measured in this study. This could be a reason as to why there were no significant differences comparing a three-ply with a one + two–ply or a two + one–ply sock, but there were significant differences when comparing a five-ply with a two + three– and three + two–ply sock. This is similar to the results found by Cagle et al.,8 who found that three 1-ply socks were 20% greater in thickness than one 3-ply sock; one 3-ply sock + two 1-ply socks averaged 30% greater thickness than one 5-ply sock. This finding is important for prosthetists when considering a percentage to increase/decrease the volume of a socket, as the use of multiple socks will not have the same outcome as the use of a single sock. Similarly, this result may have clinical implications in educating patients to wear the least number of individual socks for the correct ply as this will directly affect the socket fit and interpretation by the prosthetist and the fit of the socket.

Conventional wisdom among many prosthetists holds that the greater the sock ply, the greater the increase of volume. Although this study found that the volumetric change for five-ply is significantly greater than for one-ply socks, it did not find a difference between a two-ply or three-ply sock. Sanders et al.5 measured the thickness of prosthetic socks of different materials and plies to investigate the relationships between sock ply and thickness. Their results showed that new, unused socks of the same ply were of comparable thickness with two exceptions of wool, which tended to be thicker than average, and cotton, which tended to be thinner than average.5 They used a custom electronic height gauge to measure initial sock thickness. The mean thickness for all the socks tested was as follows: 1.6 mm for one ply, 2.3 mm for three ply, and 2.6 mm for five ply.5 The study of Sanders et al.5 used new socks that had never been tested or worn. This was done to ensure that no socks were put under mechanical stress.5 The results from these studies confirm the clinical intuition that thickness/volume increases with the use of increased sock ply. However, there is less of a difference between smaller ply changes than would be expected. It should be noted that both this and Sanders et al.'s studies were conducted in a laboratory, not on patients, so further research is needed to provide greater insights into the effect of these ply changes within the socket on a live patient.

This study may also be limited in its direct applicability to human subjects, as these results do not account for the compression of the sock within the prosthetic socket. The volume measurement was taken at the distal patella, which potentially is proximal to the posterior wall and thereby may be outside of the socket trim lines. This methodological simplification on foam models outside of the socket may have resulted in slightly greater volumetric changes than what may occur within the socket. Therefore, prosthetists should use these results with caution. The reproducibility of these results on human subjects within a socket may be difficult because of the location of the proximal height for the volume measurement. These factors should be considered and addressed in future studies.


This study found a statistically significant increase in overall limb volume, compared with baseline, with the application of a five-ply versus one-ply sock. The use of a single prosthetic sock generally resulted in mean volumetric increase of 5% to 10% over the baseline. This should be used as a baseline guide for prosthetists, as it may be a slight overestimation compared with what would be found within the socket under compression by the human limb. The use of multiple socks may result in a greater volumetric increase of limb volume than using a single sock of the same ply, so prosthetists should educate their patients to use the least number of socks for the correct ply fit. Future studies are needed to determine the in-socket effects of sock ply and resultant volumetric changes in order to more accurately guide prosthetists in their socket modifications.


The authors extend their appreciation to Gillette Children's Specialty Healthcare for supporting this research. Staff members spent time in addition to their clinical practice to help with the study.


1. American Board for Certification in Orthotics, Prosthetics & Pedorthics, Inc. Practice analysis of certified practitioners in the disciplines of orthotics and prosthetics. 2015. Accessed February 1, 2016.
2. Smith D, Michael J, Bowker J. Atlas of Amputations and Limb Deficiencies Surgical, Prosthetic and Rehabilitation Principles. 3rd Ed. Rosemount, IL: American Academy of Orthopaedic Surgeons; 2004.
3. Ossur life without limitations. 2003.
4. WillowWood. 2011. Accessed September 1, 2016.
5. Sanders JE, Cagle JC, Harrison DS, Karchin A. Amputee socks: how does sock ply relate to sock thickness? Prosthet Orthot Int 2012;36(1):77–86.
6. Bolt A, de Boer-Wilzing VG, Geertzen JH, et al. Variation in measurements of transtibial stump model volume: a comparison of five methods. Am J Phys Med Rehabil 2010;89(5):376–384.
7. Zachariah S, Saxena R, Fergason J, Sanders J. Shape and volume change in the transtibial residuum over the short term: preliminary investigation of six subjects. J Rehabil Res Dev 2004;41(5):683–694.
8. Cagle JC, Yu AJ, Coil MA, Sanders JE. Amputee socks: thickness of multiple socks. Prosthet Orthot Int 2014;38(5):405–412.
9. Rodin4D Global CAD/CAM solution for CPO. 2003.
10. True Random Number Service. 1998.

residual limb; volume; sock; artificial limb; prosthetic limb; sock ply; prosthesis; limb volume

Copyright © 2018 American Academy of Orthotists and Prosthetists