Since the 1980s, there has been an increasing number of research studies that have shown that a well-designed and well-fitted sports bra can reduce excessive breast motion and/or the associated exercise-induced breast pain that women often incur during physical activity (10,14,17,21). In fact, clinicians and researchers have suggested that a supportive and well-fitted sports bra should be considered an essential piece of sporting equipment, not just a piece of underwear (19). Previous research has revealed, however, that sports bras are not the most common breast support choice during exercise in young Australian women, whereby only 41% of female participants who were surveyed reported wearing a sports bra when they participated in physical activity (5). When the deterrents to sports bra usage were further investigated, it was determined that these same women reported that they extremely disliked the shoulder straps of commercially available sports bras, mainly due to the shoulder straps cutting into and/or slipping off their shoulders (6). More importantly, other reports have highlighted the substantial negative health implications associated with inappropriate bra shoulder strap fit and/or design. These include the presence of keloid scarring on the skin directly under the shoulder strap (15) and soft tissue damage to the shoulder (9), further described in detail in the following paragraphs, as a result of excessive downward pressure of the bra strap. Despite being the most disliked feature in current sport bra designs (5,6) and the negative health consequences associated with badly designed or fitted bra straps, no research was found that systematically investigated the effects of modifying current shoulder strap designs on the comfort and support provided by a sports bra.
The purpose of shoulder straps is to hold a bra in place, not to provide support to the breasts (8). However, it has been reported that some women, especially those with large breasts, tighten their bra straps in an attempt to hold their breast tissue off the anterior chest wall (13). This causes much of the weight of the breasts to be borne by the bra shoulder straps and, in turn, causes the bra straps to exert high pressure on the shoulders of the wearer. This resultant pressure on the shoulders not only can lead to neck and shoulder pain but also can result in the development of deep bra strap furrows that are often seen in the shoulders of women with large breasts and are usually a direct result of narrow bra straps cutting into the soft tissues in the shoulders (9). In addition, the downward pressure exerted by the bra strap on the clavicle can narrow the costoclavicular passage and affect the neurovascular bundle (9), resulting in puffy blue hands and possible nerve dysfunction (22), including paresthesia of the fifth digit of the hand (9). It has been recommended that women who suffer any of these negative health consequences associated with high bra strap pressures should remove the shoulder straps from their bras, widen the straps, or insert a bra strap cushion to decrease the pressure exerted on their shoulders and therefore reduce the related symptoms including neck and shoulder pain (9). No published literature could be found, however, quantifying the pressures that bra shoulder straps exert on the shoulders of bra wearers, or the effect of inserting a bra strap cushion on the bra shoulder strap pressures or shoulder comfort of women.
Traditionally, bra straps were designed to traverse the shoulders of the wearer and attach to the girdle of the bra in a direct line with each nipple (7) (see Fig. 1A). To prevent straps from slipping off the shoulder, however, this traditional strap configuration has been modified to include racer-back, crossed-back, and T-bar designs (see Fig. 1). Anecdotally, it is thought that the traditional shoulder strap orientation is more effective in limiting vertical breast motion compared with other strap configurations, particularly crossed-back strap designs, as a vertically positioned strap is directly aligned with the force vector generated by breast motion during activities such as running, which incorporate vertical trunk motion (20). However, biomechanical testing completed at the Australian Institute of Sport in 2006 suggested that a crossed-back strap design was more effective in limiting breast motion than the traditional shoulder strap orientation, although no details of this investigation are available in the published scientific literature (1). Although one publication has investigated the effects of lift and gather on breast tissue as a result of differing strap conditions (24), despite the finding that the bra strap slipping off the shoulder of the wearer was one of the most disliked features in current sports bra designs (6), no research could be found that systematically investigated the effect of modifying shoulder strap orientation on vertical breast motion, resultant exercise-induced breast pain, and/or shoulder comfort.
Given the lack of research about the design of sports bra straps, as well as the negative health consequences associated with inappropriate bra strap design and fit, the purpose of this study was (i) to quantify the pressures that bra shoulder straps exert on the shoulder of the wearer and to investigate how bra strap cushions moderate this pressure and resultant shoulder comfort and (ii) to investigate the effects of variations in bra shoulder strap configuration (traditional vs crossed-back orientation) on vertical breast motion, exercise-induced breast discomfort, and shoulder pressure and comfort. On the basis of previous literature and biomechanical principles, it was hypothesized that (i) inserting bra strap cushions would decrease the pressure that bra shoulder straps exert on the shoulder of the wearer, in turn, increasing shoulder comfort; and (ii) a traditional shoulder strap orientation would be associated with less vertical breast motion and shoulder pressure and, in turn, less exercise-induced breast discomfort than a crossed-back strap alignment.
Fourteen healthy active women (mean age = 24.8 ± 3.8 yr, height = 166.4 ± 5.3 cm, mass = 66.6 ± 9.1 kg), who were professionally measured as a C+ bra cup, were recruited for this study from the staff and students at the University of Wollongong. Before inclusion, the chief Investigator (K-AB) fitted all potential participants in the test bra (Berlei Ultrasport Elite) in accordance with the sports bra manufacturer’s sizing manual (2). The most common bra size among the participants was 14C (6 of 14 participants) with the bandwidth size ranging from 10 to 14 (Australian sizing) and the cup measuring as either C or D cup size. Participant inclusion was limited by age (20–35 yr), bra size (C cup size or larger), any current or previous pregnancies, any breast surgery or interventions, or any musculoskeletal disorder that would affect running or walking. To avoid breast pain due to menstruation, the participants were tested within the week before the onset of their menses. The University of Wollongong Human Research Ethics Committee approved all recruiting and testing procedures (HE02/071), and all subjects gave written informed consent to participate in the study.
Five experimental conditions were tested during the study, with participants completing one running trial for each condition. A no strap condition was used as a base measurement. A traditional vertical strap alignment and a crossed-back strap alignment, with both strap alignments tested with and without the inclusion of a bra strap cushion formed the other four experimental conditions. The test bra used during all trials was the Ultrasport Elite sports bra (Berlei, Pacific Brands, New South Wales, Australia). This sports bra was selected as the test bra because the strap configuration could be easily modified while keeping the rest of the bra structure and materials consistent throughout every trial (see Fig. 2). The test bra was composed of white nylon, cotton, polyester, and Lycra; contained underwire; and was encapsulating in design. All participants wore a new bra for their test session to ensure that there were no adverse effects of either wear or washing on the bras, and the conditions were randomly allocated to ensure that bra order did not affect any results.
Commercially available, oval-shaped shoulder strap cushions (Amoena no. 192 shoulder cushions, Mount Waverley, Australia) were used for the strap cushion conditions. The cushions (130 × 50 mm) consisted of a foam rubber laminated with nylon tricot (see Fig. 3), with Velcro attached to the anterior aspect to hold the strap of the bra in position on the cushion. The cushion was positioned on the crest of each participant’s shoulders (see Fig. 2A), with all participants using the same white pair of cushions for each cushion trial.
All testing was conducted in the Biomechanics Research Laboratory at the University of Wollongong, and the participants wore their own running shorts and shoes during all testing for each bra strap condition. Although all participants were experienced treadmill runners, they were first provided with a familiarization trial on the laboratory treadmill (Powerjog GX 100; Expert Fitness UK, Wales, UK). They were then asked to run at a comfortable self-selected speed that they could maintain for 10 min without experiencing fatigue (mean speed = 7.6 km·h−1, range = 6.3–8.5 km·h−1). A self-selected speed was chosen as pilot testing, and the literature suggests that requiring participants to run at a standardized speed that differs from their self-selected speed can affect their gait pattern (12), which may in turn affect their breast motion. All data were collected on the same day for each participant, with participants required to complete five running trials of 5 min and 15 s in duration. Participants were given a 10-min rest period between running trials, during which the next bra shoulder strap setup was prepared.
Vertical breast displacement
The vertical displacement of each participant’s breasts relative to their torso were monitored during each running trial by tracking light-emitting diodes (2-mm diameter) placed directly onto the participant’s skin overlying the sternal notch, using double sided tape (3M), and over micropore tape (3M) placed in the center of each participant’s nipples. These marker placements were deemed the most appropriate as previous research has shown nipple motion to provide a good representation of vertical breast motion (17). As previous research has shown a strong association with vertical breast motion and resultant breast pain (10,14,17,21), vertical breast displacement was deemed suitable to represent breast motion in this comparative study.
The three-dimensional position of each light-emitting diode was tracked using an OptoTRAK (3020) motion capture system (Northern Digital, Incorporated, Ontario, Canada; duty cycle, 75%; voltage, 9 V; and marker frequency, 200 Hz). In each condition, the treadmill speed was gradually increased to the participant’s preselected running speed with data collected 5 min into the trial for a 15-s period during which the participant was running with a consistent, steady-state stride pattern. The treadmill speed was then decreased to a stop.
Vertical breast displacement for each trial was calculated (in centimeters) from the displacement data by removing torso motion (characterized by the sternal notch marker) from the nipple motion in the vertical plane. The mean values of the 10 most representative vertical breast displacement values, derived from cycles in which no marker data were occluded due to participant arm swing, were calculated during the 15-s sampling period per condition and used for data analysis.
Breast pain and shoulder comfort
To quantify the exercise-induced breast pain and shoulder comfort perceived by the participants during each trial, we asked the participants to indicate their pain and comfort level immediately after each trial using a 100-mm-long visual analog scale, which has been reported to be valid and reliable in previous research (3). Breast pain was rated from “0,” indicating no breast pain, to “100,” indicating the worst breast pain possible; and shoulder comfort was rated from “0,” indicating the most comfort possible, to “100,” indicating the least comfort possible. After completing all trials, participants answered qualitative questions regarding their overall bra comfort and their preferred shoulder strap configuration.
Bra shoulder strap pressure
The force, contact area, and pressure exerted by the bra shoulder strap onto the shoulder of each participant were collected using two custom-designed pliance pressure strip sensors (novel GmbH, Munich, Germany) as described previously in Bowles et al. (4) (see Fig. 2). These strips consisted of ten 1-cm2 sensors in parallel and were attached to the inner aspect of the bra shoulder strap or the shoulder cushion, always in direct contact with the participant’s shoulder. The sensor was positioned to lie on the crest of the shoulder, with even amounts of the sensor strip located on the anterior and posterior aspects of the shoulder. The sensor strips were “zeroed” once attached to the bra shoulder strap, before the bra was placed on the participant. Data from the pressure strips were collected at 50 Hz using a pliance mobile multi-interface box attached to a collection box (novel GmbH) and interfaced with a personal computer using the pliance Expert 8.2 online software (novel GmbH) for data collection. The pressure measurements (N·cm−2) were recorded for 15 s coinciding with the vertical breast displacement measures, 5 min into the running trial. Although measures of pressure were recorded simultaneously with vertical breast displacement, similar to previous research published (4), the data collection equipment for both measures could not be time locked, and therefore no assumptions of the exact temporal relationship between these measures can be made. Sensor data collection zones or “masks” were established before data collection using the novel-win software (novel GmbH) to group the ten sensors in each strip. As a result of preestablishing the masks, the following variables were derived through the pliance Expert 8.2 online software (novel GmbH): contact area (the sum of all of the loaded sensors, in both strips, during the time of the trial), maximum force (the total force measured over both strips), and mean and maximum pressure and pressure–time integral (measured and calculated for each strip with software displaying the highest value).
The mean and SD values were calculated first for each of the five experimental conditions (no strap, traditional, and crossed-back orientation, with and without strap cushions) for the average vertical breast displacement, the VAS scores charactering breast pain and shoulder comfort, and the bra shoulder strap pressure data. To quantify the effects of inserting the bra cushions under the bra straps and how any effects were moderated by strap orientation, we analyzed shoulder comfort, mean and maximal pressure, force, contact area, and pressure–time integral data using an ANOVA design with two within factors (strap cushion: with and without; and strap orientation: traditional and crossed-back).
To determine the effect of bra shoulder strap orientation on vertical breast displacement, mean pressure, pain and comfort scores, we used a one-way repeated-measures ANOVA design with the shoulder strap orientation (no strap, traditional, and crossed-back) as the independent variable. Where a significant difference was found, Bonferroni pairwise comparison tests were conducted to identify where the difference lay, with the alpha level adjusted for the number of comparisons in the test by the statistic software. All statistical analyzes were deemed significant to a level of P ≤ 0.05 (Statistical Package for the Social Sciences, version 19; SPSS Inc., Chicago, IL).
Effects of bra shoulder strap cushion
During the running trials maximal pressures ranging from 0.83 to 2.67 N·cm−2 and mean pressures ranging from 0.52 to 1.06 N·cm−2 were exerted by the bra straps on each shoulder of the wearer. Unexpectedly, there was no significant main effect of shoulder strap cushion on any of the contact area, force, pressure, or comfort results. However, there was a significant strap cushion–strap orientation interaction (P = 0.038). That is, although there was no change in pressure with and without the cushion under the traditionally orientated straps, the maximal pressure generated at the bra strap–shoulder interface under the crossed-back straps was significantly lower when the cushion was inserted compared to when there was no strap cushion (see Table 1).
Effects of bra shoulder strap orientation
There was a significant main effect of modifying the bra shoulder strap orientation on vertical breast displacement during running (P = 0.012). Bonferroni pairwise comparisons revealed that vertical breast displacement was significantly less in the crossed-back strap orientation compared with the no strap condition (P = 0.015; see Table 2). However, there was no significant difference in vertical breast displacement between the traditional strap orientation and the no strap condition (P = 0.076), although this value approached significance. There was also no significant difference between the traditional strap orientation and the crossed-back strap orientation (P = 0.834) in reducing vertical breast motion during running.
A significant main effect of shoulder strap orientation was also noted on breast pain during the running trials (P = 0.001; see Table 2). Post hoc analysis revealed a significant reduction in breast pain when participants used both the traditional and crossed-back strap orientations (both P = 0.002) compared with the no strap condition. No significant difference was found, however, between the two strap orientations (P = 0.259) in resultant exercise-induced breast pain. In addition, no significant between-condition difference was found in the shoulder comfort measures regardless of shoulder strap orientations (P = 0.453), although the crossed-back orientation resulted in significant increases in mean pressure values when compared with the traditional strap orientation (P = 0.011). When questioned on overall bra comfort, 11 of 14 participants reported that the crossed-back shoulder orientation was more comfortable for them than the traditional shoulder strap orientation.
This unique research is the first study published in the scientific literature to quantify the pressures that bra shoulder straps exert on the shoulder of bra wearers and to determine whether the insertion of a bra shoulder strap cushion could moderate these pressures. The effects of bra strap cushions on bra shoulder pressures as well as the effect of changes in bra shoulder strap orientations on breast motion and exercise-induced breast discomfort are discussed in the next section.
Effects of bra shoulder strap cushions
The pressures exerted by a single bra strap onto the participants’ shoulders (mean range = 0.52–1.06 N·cm−2, maximal pressure range = 0.83–2.67 N·cm−2) might initially seem low and therefore inconsequential. However, Jones and Hooper (11) cited previous work suggesting that continuous pressure values of up to 14 kPa (1.4 N·cm−2) should be avoided to prevent tissue damage while carrying a loaded backpack. Furthermore, measured values from the current study are in line with the mean values (70–110 mm Hg; 0.933–1.47 N·cm−2) that have been reported previously in children for shoulder–surface contact pressures applied by a backpack during walking, loaded with 10% of the child’s body mass (16). Of most concern is that all these measures are greater than 30 mm Hg (0.40 N·cm−2), which has been reported as the pressure threshold at which skin blood flow can be occluded (16). Given that women typically wear a bra on a daily basis during most of the hours they are awake, these pressures exerted by the bra strap onto the wearer’s shoulder have to be sustained for numerous hours every day, explaining the development of deep furrows in the soft tissue structures of the shoulder in some women over time (9). Furthermore, pressure values in the present study were measured for the shoulder straps of a sports bra that are traditionally wider than the straps of a fashion bra. As thinner bra straps would have less area to dissipate the same force, these current pressure measures are likely to underestimate the pressures exerted on the shoulders of women during activities of daily living.
Contrary to our hypothesis, inserting the bra shoulder strap cushion did not significantly decrease the mean shoulder pressure, contact area, or force measures exerted on the shoulder of the wearer. In fact, the strap cushion only significantly reduced the maximal pressure exerted on the shoulder while it was inserted under the crossed-back bra shoulder strap compared with the traditional shoulder strap orientation. We anticipated that the shoulder strap cushions would increase the area over which the shoulder strap forces could be distributed and, in turn, reduce the resultant pressures. However, unexpectedly, there was only a negligible increase in the contact area between the bra strap and the shoulder when the bra strap cushion was inserted, and this was accompanied by a similarly small increase in force, such that there was no substantial change in the mean pressures exerted on the shoulders during the running trials when the shoulder strap cushion was inserted. Visual inspection of the strap cushions revealed that the lateral portions of the cushions tended to lift off the wearers’ shoulders, particularly in the traditional shoulder strap orientation condition, and therefore did not increase the actual bra strap–shoulder contact area. Interestingly, in the crossed-back shoulder strap orientation, the bra strap transverses over the trapezius muscle on the posterior aspect of the shoulder rather than the bony spine of the scapula in the traditional strap orientation. Traversing the relatively flat muscle belly permitted the shoulder strap cushion to be placed in a more level position compared with when traversing the bony prominence in the traditional strap orientation. This, in turn, appears to allow the cushion to be more effective in the cross-back shoulder orientation, as evidenced by the significant reduction in the maximal shoulder pressure in this condition compared with the traditional strap orientation. On the basis of these results, we recommended that strap cushions should be redesigned to enable them to remain flat against the shoulder so that they can effectively increase the area over which the forces from the shoulder straps are exerted.
Effects of bra shoulder strap orientation
Contrary to our hypothesis and the anecdotal beliefs about strap orientation (20), the traditional shoulder strap orientation was not more successful in decreasing vertical breast motion or decreasing breast pain during the running trials compared with the crossed-back strap orientation. Although there were no significant differences in these variables when comparing the two experimental shoulder strap orientations, only the crossed-back shoulder strap orientation significantly reduced vertical breast motion relative to the no strap condition. With no other scientific literature found investigating bra shoulder strap orientations, we can only speculate as to why the results of this study were contrary to our hypothesis, yet consistent with the results found in the Australian Institute of Sport investigation (1). As stated previously, the shoulder straps of a bra should be used to anchor the bra rather than support the breasts (8), with breast support provided by the bra band. We speculate that if the bra band is well structured and provides the necessary breast support, orientation of the shoulder straps might not be as important in providing breast support as other sections of the bra. For example, McGhee and Steele (18) reported that increasing the elevation and compression of the breasts by innovative cup design in a sports bra decreased breast motion and related breast pain and improved sports bra comfort. Furthermore, Starr et al. (23) also stressed the importance of an encapsulating cup design for a sports bra to limit breast motion rather than a compression style sports bra. The material structure of the shoulder straps in this study was consistent across the two experimental strap orientations, with the modification in strap orientation resulting in no significant difference in the ability of the shoulder strap to reduce vertical breast displacement and related pain. Therefore, it is suggested that the material composition of the shoulder straps may be more important in holding the bra in place compared with the strap orientation, although this notion warrants further investigation.
Consistent with the vertical breast displacement data and the previous findings of a strong association between breast motion and exercise-induced breast pain (10,14,17,21), there was a significant decrease in breast pain in both the traditional and crossed-back strap orientations when compared with the no strap condition. Because exercise-induced breast pain can be severe enough to cause women to cease participating in physical activity, designing sport bras that can limit this breast pain is vital to improve women’s participation rates in physical activity and resultant overall health. Because there was no significant difference between the strap orientations in terms of decreasing exercise-induced breast pain, we suggest that either strap orientation has the potential to reduce exercise-induced breast pain for active women, relative to a strapless bra.
Interestingly, the crossed-back shoulder strap orientation imparted significantly more mean pressure and force on the shoulder of the wearer when compared with the traditional shoulder strap orientation, although these increases were not reflected in the shoulder comfort measures. We speculate that the increase in mean pressure in the crossed-back strap orientation was not functionally high enough to affect overall shoulder comfort. In fact, 11 of the 14 research participants reported after the running trials that they preferred the crossed-back orientation over the traditional strap orientation.
Two of the participants reported they perceived that the shoulder straps in the traditional orientation were not as “tight” over the shoulder as they would usually wear, possibly affecting the resultant pressure and comfort reading for this condition. As the posterior elastic section of the bra shoulder strap (visible in Fig. 2) had to be long enough to transverse the width of the back to attach to the opposite side in the crossed-back orientation, it was too long to be tightened to the participant’s preference in the traditional strap orientation trial for these two participants. Although this is a limitation of the current study, this problem was only apparent for two participants, and the measured vertical breast displacement during the running trial for one of these participants was actually greater with the crossed-back orientation when compared with the traditional strap orientation.
In conclusion, the insertion of a commercially available bra shoulder strap cushions was not effective in decreasing the pressure that the bra shoulder straps exerted on the wearer. However, further research is recommended to determine whether redesigning the strap cushions could increase the effective bra strap–shoulder contact area and enable the cushions to alleviate bra shoulder pressures and prevent bra straps digging into the shoulders of sports bra wearers. In addition, assuming that the band of a bra provides sufficient support, modifying strap–shoulder orientation from a traditional to a crossed-back configuration could alleviate the common problem of bra shoulder straps slipping off the shoulders of the wearer, without decreasing the overall efficacy of the sports bra.
This research was funded by an Australian Research Council Strategic Partnership with Industry Grant with Berlei (Australia) as the Industry Partner. Berlei (Australia), however, was not involved in any data collection, data analysis, or production of this article. Funding was also provided by the New South Wales Sporting Injuries Committee, and the custom novel pressure measuring strips were kindly donated by novel GmbH for this experiment.
No conflict of interest exists for either author, Kelly-Ann Bowles or Julie Steele.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.