The effect of washing and wearing on sports bra function

ABSTRACT

Sports bras provide support by restricting breast motion during exercise, which may prevent damage to breast structures. Laundering affects the mechanical properties of some sports bra materials. Bra function on the wearer after washing is unreported, meaning sports bra durability is unknown. This intervention study compared subjective and objective performance of sports bras that were washed, and worn/washed, to a control. Twenty-two females were assigned three identical sports bras; control, washed and worn/washed (washed after 60-min wear). Pre-intervention: breast position was recorded while standing and running in each bra. Comfort, fit and aesthetics were rated. Intervention: participants undertook their normal exercise in their worn/washed bra. Post-intervention: after 25 washes (n = 19), pre-intervention testing was repeated. Breast volume and control bra motion were consistent pre/post-intervention; however, post-intervention breast motion increased (20% mediolaterally, 16% superioinferiorly) in washed bras and (32% mediolaterally, 25% superioinferiorly) worn/washed bras. Post-intervention washed, and worn/washed bras were perceived as less supportive and washed bras less comfortable than worn/washed bras. Sports bra support reduced after 25 washes; this was compounded by wear. Participants detected reduced support, but comfort was sustained, suggesting replacement may not be considered. Guidelines on sports bra durability are recommended for breast health.

KEYWORDS:

• Breast
• laundering
• kinematics
• skin strain

Introduction

Breast tissue has limited intrinsic support, and vigorous movement may cause damage to anatomical breast structures (Norris, Mills et al., 2020). Sports bras provide support by restricting three-dimensional (3D) breast motion during exercise (Steele et al., 2018; Zhou et al., 2011), which reduces breast pain, discomfort and embarrassment (Chen et al., 2011; Risius et al., 2017). Sports bras are restrictive to control breast movement, but contain elastic to allow respiration (Bowles et al., 2012; Page & Steele, 1999), shape retention and fit (McCann & Bryson, 2014; Yu & Zhou, 2016). Elasticity, absorbency and durability are important considerations for sports bra materials (Starr, 2002). Elastane, polyester and Lycra are common materials in sports bras (Page & Steele, 1999) as they are lightweight, durable and washable (Yu & Zhou, 2016). Despite potential susceptibility to breakdown, bonding and lamination is common in sports bras as it can offer support, shape retention and stretch (Yu et al., 2006).

As sports bras are not for single use, assessment of laundering effects on their functional properties is essential (Senthilkumar & Anbumani, 2012). The elastane, polyamide or polyester in sports bras provide diverse mechanical properties and are easy to wash and dry quickly (McCann & Bryson, 2014; Yu & Zhou, 2016). However, research into elastic fabrics for tight-fitting sportswear identified shrinkage and reduced recovery up to 20 washes (Easter & Ankenman, 2006; Senthilkumar & Anbumani, 2012). Sports bras may also incorporate knitted fabrics as they allow large extensibility and high recovery rates (Yu & Zhou, 2016); however, they may also shrink during washing, due to the heat, moisture and mechanical action experienced (Mikučioniené & Čepukoné, 2017). In general, research suggests that repeated laundering negatively impacts physical and mechanical properties of various materials commonly used in sports bras.

Industry practice sees some manufacturers and brands investigating washing effects on the ability of particular sports bras and bra components (such as straps or padding in the cups) to withstand certain tests, including mechanical tension tests (anecdotally after ~25 washes). However, data on bra function on the wearer after a certain number of washes have not been identified. Anecdotally, it has been suggested that sports bras should be replaced with footwear, every 3–6 months, or after 25 washes (McGhee et al., 2008), but without an evidence base to support these claims the bra industry is unable to provide consumers with advice on sports bra durability or replacement.

Literature suggests that sports bra function is determined by many factors, including support, fit and comfort. Measurement of breast (nipple) range of motion using 3D motion capture (Scurr et al., 2010) or more recently electromagnetic motion sensors (Norris, Jones et al., 2020) is typically used as a surrogate measure of bra support (Scurr et al., 2010), during activities such as running, which has low stride variance (~3%; Jordan et al., 2006), enabling comparisons across conditions and time and comparison to previous literature.

Changes in mechanical properties of a bra over time may affect bra fit due to stretching or shrinkage, which could have negative health implications (McGhee & Steele, 2010; White & Scurr, 2012; Wilson & Sellwood, 1976). Assessing bra fit lacks definitive guidelines; however, research has used subjective assessments of key fit parameters by trained individuals (McGhee & Steele, 2010). Additionally, changes in the objective position of key breast and bra landmarks over time may also help us understand changes in bra fit. Changes in sports bra fit and function may alter wearer’s perceptions of comfort and support.

Therefore, this study investigates the effect of washing a sports bra compared to wearing and washing a sports bra on support, fit and satisfaction of the garment. From 0 to 25 washes, it is hypothesised that there will be significant changes in breast motion during running in a sports bra that is washed and in a sports bra that is worn/washed, compared to a similar control bra. It is also hypothesised that subjective and objective bra fit will change for the washed bra, and the worn/washed bra from 0 to 25 wash cycles and compared to the control bra. Finally, it is hypothesised that bra discomfort and garment dissatisfaction will increase in the washed bra, and in the worn/washed bra from 0 to 25 washes and compared to the control bra.

Methods

The University of Portsmouth Research Ethics Committee has approved the study (SFEC 2018-130A).

Participants

Twenty-two females volunteered to participate (mean (standard deviation) age 21.6 (2.0) years, mass 62.8 (9.8) kg, overbust circumference 88.6 (6.4) cm, underbust circumference 75.7 (5.3) cm, total breast volume 1481.7 (434.9) cc, underband size range 32–36 (mode 32) and cup size range A–F (mode D)). Participants had a washing machine, had not given birth, had no surgical breast procedures and were undertaking ≥60 minutes of exercise each week. Following an explanation of the procedures, informed consent was obtained, and bra size established by a trained bra fitter in the chosen sports bra.

The bra

The chosen sports bra was Lululemon Enlite Sports Bra (Figure 1); this bra was chosen as it was rated highly by Runner's World at the time (Runners World, 2020). Participants were assigned three of these sports bras in the appropriate size: the control bra, the washed bra and the worn/washed bra.

Figure 1. The Lululemon Enlite sports bra. A combination style, made from 56% nylon and 44% lycra elastane; cup lining made from 80% polyester and 20% elastane. The sports bra is a bonded, laminated sports bra, which is marketed as high support. This sports bra is designed for running and includes lycra and built-in cups, a bonded underband and a stitch-free hook-and-eye closure (Zhou et al., 2011). Marker and electromagnetic sensor positions (grey; Starr, 2002) and definitions of breast projection, breast drop and breast separation.

Procedure: static tests

Participant’s body mass was recorded. Bra fit was subjectively assessed in the three bras by a trained bra fitter using the professional bra fitting criteria (McGhee & Steele, 2010). To understand the stability of participants’ breast size over time, bare-breast volume was measured using three-dimensional (3D) scanning (Artec Eva 3D, Luxembourg). During scanning, markers were attached to the suprasternal notch (STN), xiphoid process (XP), seventh cervical vertebrae (C7) and eighth thoracic vertebrae (T8; Mills et al., 2016) to account for changes in posture (torso orientation; Figure 1). Participants’ breast boundary was marked on the skin using the lift and fold method (Lee et al., 2004). To measure changes in breast projection, breast drop and breast separation (Figure 1), participants were scanned in each sports bra.

Procedure: dynamic tests

Participants completed a 10 km/h (Bowles et al., 2012) treadmill run (Scurr et al., 2010) for 30 s in each sports bra (random order). Breast and torso movements were measured using five microelectromagnetic sensors (240 Hz, Micro Sensor 1.8TM, diameter = 1.8 mm, mass = <1.0 g, LIBERTY, Polhemus, Canada; Norris, Jones et al., 2020) positioned on the same landmarks as Figure 1, plus a left nipple sensor.

Following each run, participants rated the comfort and support for each bra using a numerical rating scale from 0 (no support, very uncomfortable) to 10 (maximum support, very comfortable). Bra clasp settings were standardised across bras and testing sessions. To assess protocol reliability, participants undertook one running trial twice (random order).

The intervention

Participants were given their washed bra, worn/washed bra and manufacturer washing guidelines (which included machine washing at 30°C). The control bra remained in the laboratory and was not worn or washed. Participants were advised to undertake their normal exercise activities whilst wearing a worn/washed bra. After each ~60 minutes of wear, participants washed both the washed bra, and the worn/washed bra, following similar washing and drying routines with similar detergent and washing machine.

Post-intervention testing

Following 25 washes, 19 participants returned with the washed bra, and the worn/washed bra for testing session 2 (three participants dropped out). The tests undertaken in session 1 were repeated in the control, washed and worn/washed bras. Participants were then verbally encouraged to continue the study, wearing and/or washing the same bras for another 25 washes. Two participants (34A and 34C) continued and returned after 50 washes for session 3, where the same tests were repeated. Weekly, throughout the intervention, participants were prompted to record their washing and wearing and washing activities and bra feedback using an online form. The research team used this form to monitor adherence.

Data analysis

Data from the 3D scans were converted to models using Artec Studio 12 and 13 (Artec Group, San Jose). Breast volume was calculated from a bare-breast scan (Geomagic V12; 3D Systems, USA), by dissecting the breast from the torso using a manually marked breast boundary. Dissected breasts were then individually filled to obtain breast volume. For static 3D scans, an LCS with the origin at the sternal notch (MATLAB, Version 2017b, MathWorks, UK) was used to calculate sagittal plane breast projection, frontal plane breast drop and transverse plane breast separation (Figure 1). For dynamic tests, breast and torso movement data were exported to Visual 3D (Version 4.96.4, C-Motion, USA) and low pass Butterworth filtered (cut-off at 13 Hz) (Mills et al., 2014). A different torso segment model created an LCS with the origin midway between STN and C7. Midway between XP and T8 defined the distal end of the torso segment (Mills et al., 2016). Nipple and torso movements were assessed anterioposteriorly (AP), mediolaterally (ML) and superioinferiorly (SI) and then relative breast motion calculated over 10 gait cycles.

Statistical analysis

Breast volume data were non-parametric and compared between testing sessions 1 and 2 using a Wilcoxon Signed Ranks Test. One-way repeated measures ANOVA with one factor (bra condition four levels; control, washed, worn/washed and reliability retest) compared test and retest trials in session 1. Data was normally distributed and sphericity assumed (Mauchly's Test of Sphericity). Breast motion and breast position (projection, drop and separation) in the control, washed and worn/washed bras across testing sessions 1 and 2 were compared using two-way repeated measures ANOVAs with two factors (testing session, two levels; bras, three levels). All data were normally distributed, Sphericity assumed (Mauchly's test of Sphericity) or Greenhouse Geisser correction factor used. Perceived bra support and bra comfort were compared across bra conditions and testing sessions using a non-parametric Wilcoxon Signed Ranks Test. The alpha level was 0.05, unless multiple post-hoc tests were performed and a Bonferroni adjustment used. Post-hoc effect sizes were calculated as small (d = 0.2), medium (d = 0.5) and large (d = 0.8; Cohen, 1988) using G*Power (3.1.7, Germany). As only two participants progressed to session 3 (50 washes), statistical comparisons were not appropriate. Both participants were smaller breasted (34A and 34C) and the results did not reveal any trends.

Results

During the intervention, participants undertook a variety of activities in the worn/washed bra: gym (41%), running (13%), home workout (12%), trampolining (10%), football (<5%), exercise class (<5%), high-intensity interval training (<5%), strength training (<5%), netball (<5%), walking (<5%), tennis (<5%), circuits (<5%), and cycling (<5%). Nineteen participants completed 25 wash/wear cycles in an average of 85 days. Two participants completed 50 wash/wear cycles in an average of 224 days.

Mean total breast volume was 1,481 (434) cc in session 1 and remained similar in session 2 – 1,470 (441) cc (W = 75.0₍₁₅₎, p = 0.39). Test and retest running trials in session 1 showed no significant differences in anterioposterior (F = 0.71_(1,4), p = 0.55), mediolateral (F = 0.87_(1,4), p = 0.46) and superioinferior (F = 0.38_(1,4), p = 0.77) breast motion (Figure 2).

Figure 2. Mean (standard deviation error bars) breast range of motion during running at 10 kph and static breast position from 3D scans in each bra condition in session 1 (0 washes, n = 22) and in session 2 (25 washes, n = 19). For breast range of motion, brackets show significant differences, where p < 0.02. Brackets show significant differences for breast drop, where T = 2.58₍₁₇₎, p = 0.01, d = 0.42, and for breast separation, where T = −3.24₍₁₇₎, p < 0.01, d = 0.76.

During session 1, there was no significant difference in breast motion across the bra conditions or across testing sessions 1 and 2 for the control bra. There was, however, a 20% (1 mm) increase in mediolateral and 16% (2.7 mm) increase in superioinferior breast motion from session 1 to 2 when running wearing the washed bra (mediolateral, T = −2.45₍₁₇₎, p = 0.01, d = 0.44; superioinferior, T = −2.71₍₁₇₎, p ≤ 0.01, d = 0.60). There was a 32% (1.6 mm) increase in mediolateral and 25% (4.2 mm) increase in superioinferior breast motion from session 1 to 2 when running wearing the worn/washed bra (mediolateral, T = −4.21₍₁₇₎, p < 0.01, d = 0.94; superioinferior, T = −5.31₍₁₇₎, p < 0.01, d = 1.22). The only interaction effect was a significant increase of 7% (1.4 mm) in superioinferior breast motion in the worn/washed bra compared to the washed bra in session 2 (T = −1.80₍₁₇₎, p < 0.05, d = 0.45; Figure 2).

In session 1, the subjective assessment of bra fit failed on a similar number of participants across all bras (one participant failed one fit parameter in each bra). However, bra fit appeared to deteriorate in sessions 2 and 3, in particular, for the worn/washed bra with at least one fit parameter failing in 7 out of 19 participants in session 2 compared to 1 participant out of 22 in session 1. The objective assessment of static breast position in each bra showed no significant difference in position between the control, washed and worn/washed bras in session 1 and no difference in static breast position for the control bra between sessions 1 and 2. Overall, there was limited change in breast position for any of the bras across the two sessions (Figure 2).

For the washed bra, there was a significant decrease in perceived support and bra comfort from sessions 1 to 2. Participants perceived less support in the worn/washed bra from sessions 1 to 2, but comfort remained the same (Figure 3). More participants perceived their breasts to move in the worn/washed bra in session 2 and for the bra itself to move rather than stay in place when compared to the same bra in session 1 and compared to the washed and control bras. The experience of rubbing and chafing remains similar across the three bras and the two testing sessions.

Figure 3. Median (a) perceived sports bra support and (b) bra comfort following the running trial; c) percentage of participants who experienced issues in each bra during each session (session 1, 0 washes (n = 22), session 2, 25 washes (n = 19) and session 3, 50 washes (n = 2)).

Discussion and implications

The sports bra industry offers no evidence-based consumer advice on the effect of washing and wearing on sports bra support, fit and satisfaction. Whilst it is recognised that different bra components, materials, construction and finishes may change the outcomes of this study for each bra, being able to advise the consumer that washing deteriorates the support, fit or satisfaction of a sports bra may encourage more appropriate consumer replacement strategies.

During this intervention study, participants undertook a variety of activities while wearing the worn/washed bra, taking an average of nearly 3 months to complete 25 hours of wear (25 washes). Over this time period, participant’s breast volume remained consistent. During running trials in testing session 1, the control, washed, and worn/washed bras all demonstrated similar magnitudes of breast motion and the control bra demonstrated consistent breast motion from testing session 1 to session 2, suggesting comparable conditions.

During the running trials in testing session 2, there was an increase in breast motion mediolaterally and superioinferiorly in the washed bra condition (20% and 16%) and the worn/washed bra condition (32% and 25%). These results suggest that washing alone had an effect on the support provided by this sports bra, but this effect was further compounded by wear, accepting hypothesis 1. With the control bra showing no change in breast motion across sessions and the test/retest trials showing no difference, this suggests that a difference did occur in the intervention bras. However, although the difference in intervention bra breast motion between sessions was significant, the absolute change was small (<4.2 mm), which may question the meaningfulness of this outcome. No studies have been located that reported breast motion data in three-dimensions across high-support bra conditions during running; however, Norris et al. reported differences in breast motion between a low-support bra and a high-support bra of ~4 mm (ml) and 12 mm (SI).

An increase in breast motion during running has been linked to increases in breast pain (Scurr et al., 2010), increases in the risk of breast tissue strain (Norris, Mills et al., 2020), embarrassment (Burnett et al., 2015) and changes in mechanical performance, e.g., gait parameters (White et al., 2011). Interestingly, the anterioposterior breast motion did not change from sessions 1 to 2, suggesting that the bra maintained its support in this direction. Disappointingly, only two participants continued the study to reach 50 washes and their breast motion showed no obvious trends across testing sessions.

Previous research into elastic fabrics in tight-fitting sportswear identified shrinkage following initial washing (Easter & Ankenman, 2006; Senthilkumar & Anbumani, 2012). Therefore, it may have been hypothesised that sports bra washing would lead to shrinkage causing a tightening of the sports bra, and consequently a reduction in breast motion. However, the results showed an increase in breast motion in the sports bras that were washed. Visually, the bras displayed no obvious signs of degradation, other than discolouration and one broken clasp. However, the washed, and the worn/washed bras were much more malleable (less stiff) than the control bra and did not retain their shape as well, although no mechanical testing was undertaken on these bras.

Across all the bras in session 1, subjective fit parameters only failed for one participant, compared to seven in session 2. The subjective fit assessment did not highlight one particular fit parameter that degraded more. There were no obvious trends in the subjective fit outcomes that help us understand the deterioration in support in the washed, and the worn/washed bras. However, hypothesis two is accepted as subjective fit deteriorated from sessions 1 to 2 in the washed, and the worn/washed bras.

During running, it was assumed that changes in anterioposterior breast motion would be accompanied by changes in static breast projection, as the bra compressed the breasts more or less. Similarly, changes in superioinferior breast motion were assumed to be accompanied by changes in breast drop as the bra’s ability to lift the breast changes and finally, mediolateral motion changes would be accompanied by changes in breast separation as the bra's ability to hold the breasts in the horizontal plane changes. However, despite changes in mediolateral and superioinferior breast motion, breast position showed minimal change from sessions 1 to 2. These results suggest that after 25 washes, the washed and the worn/washed bras were still able to reposition the breasts statically, but were not able to hold the breasts in this position during running. As static breast position remained relatively unchanged from sessions 1 to 2, hypothesis two is rejected.

In session 2, participants were not blinded to the bra condition and they perceived greater reductions in support in the worn/washed bra, compared to the washed bra, but the worn/washed bra was perceived as more comfortable than the washed bra. For the worn/washed bra, these outcomes may point to an increase in stretch in session 2, which reduced support (both actual and perceived) but was considered more comfortable.

Conclusions

Washing reduced sports bra support (~20%), which was compounded further by wear (~32%). It was difficult to detect these changes statically through subjective fit or breast position in each bra. However, in session 2, the washed and worn/washed bras were both perceived as less supportive, but the washed bra was also perceived as less comfortable than the bra that had been worn/washed. This suggests that participants could detect reduced support, but still found it comfortable in the worn/washed bra. For the worn/washed bra, had participants perceived both reduced support and comfort, they may have been more willing to replace this bra, but with only a reduction in perceived support and no change in comfort, they may wear the bra for longer despite the loss of support. Reduced breast support has negative health implications, and so this study suggests that objective guidelines on sports bra replacement could be beneficial to consumers.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported that there is no funding associated with the work featured in this article.