Women’s shoes and knee osteoarthritis

D Casey Kerrigan, Jennifer L Lelas, Mark E Karvosky

We assessed whether wearing wide-heeled shoes has a similar effect on knee torque to narrow-heeled shoes by measuring the joint torques of 20 healthy women during walking. Wearing wideheeled shoes had a 30% greater effect on peak external knee flexor torque than walking barefoot. Walking with wide-heeled and narrow-heeled shoes increased peak knee varus torque by 26% and 22%, respectively. Our findings imply that wide-heeled shoes cause abnormal forces across the patellofemoral and medial compartments of the knee, which are the typical anatomical sites for degenerative joint changes.

We have previously reported that traditional narrow-base (stiletto) high-heeled shoes exaggerate the knee varus and knee sagittal torques during walking—these torques are thought to be relevant to development of knee osteoarthritis.¹  Women and men have similar knee varus and knee sagittal torques during barefoot walking, which suggests that intrinsic biomechanical differences do not explain the fact that knee osteoarthritis is twice as common in women than men.²  Although high-heeled shoes are traditionally thought of as stilettos with a narrowbase sole and heel, women’s dress shoes with wide-base soles and moderately high heels are routinely worn by women of all ages, in work and other settings. These shoes, typically associated with a wide toe-box, are generally perceived to be more comfortable and stable than narrow-base heeled shoes.³  Although these shoes could reduce the risk for falls, ankle injuries, and foot deformities compared with narrow-base heeled shoes, we hypothesised that women’s wide-base heeled shoes impose the same, if not greater, alterations in knee joint torques during walking as narrow-base heeled shoes.

We studied 20 healthy women, who were comfortable wearing wide-heeled and narrow-heeled dress shoes, with a mean age of 34·9 years (SD 7·1; range 23-44 years), mean height of 1·62 m (0·05), and mean weight of 59·1 kg (9·7). ur study protocol was approved by the Institutional Review Board and we obtained written informed consent from each participant. For each woman, one pair of her wide-base shoes and one pair of her narrow-base shoes of similar heel height (defined as the maximum height of the sole at the rear of the heel) were selected for comparison. We ensured that there were no substantial differences in average height between the wide-heeled and narrow-heeled shoes; mean heights were 7·0 cm (SD 1·0) for both the wide-heeled shoes and narrowheeled shoes. Narrow-base shoes were an average 1·2 cm in width (0·4) whereas wide-base shoes were 4·5 cm in width (1·2). We measured bilateral joint torques, as the women walked barefoot, in each pair of shoes, at a comfortable speed across a 10 m gait laboratory walkway, with a six camera videobased motion analysis system (VICON 512 system; Oxford Metrics Ltd, Oxford, UK) and two staggered force platforms (Advanced Mechanical Technology, Newton, MA, USA) embedded in the walkway. We calculated joint torques with a commercialised full-inverse dynamic model (VICON Clinical Manager; Oxford Metrics Ltd, Oxford, UK) adjusted for healthy body weight and height, and reported in Newton metres per kg metres (Nm/kgm). We calculated the average peak joint-torque value for each woman over three walks (average of both lower extremities) for each type of shoe and walking barefoot. We compared peak torque-values with a repeated measures ANOVA with post-hoc t test assessment (Stata Version 6·0).

Figure 1 shows that the general pattern of sagittal knee torque was similar between wide-base and narrow-base heeled shoes, yet was different for barefoot walking. The knee flexor torque normally present during early stance, was sustained into the midstance phase and the peak knee extensor torque during late stance was significantly reduced compared with barefoot walking (mean 0·23 Nm/kgm [SD 0·08]) for both the widebase and narrow-base shoes (0·16 Nm/kgm [0·08], p<0·0001, and 0·16 Nm/kgm [0·09], p<0·0001, respectively). In addition, the peak knee flexor torque for the wide-base heeled shoe (0·35 Nm/kgm [0·14]) was 30% greater (p<0·0001) than for women walking barefoot (0·27 Nm/kgm [0·13]).

Figure 2 shows that the coronal varus torque at the knee was increased throughout the stance period for both the wide-base and narrow-base heeled shoes compared with barefoot. The first peak was significantly greater than the barefoot peak value (0·33 Nm/kgm [0·07]) for both the wide-base (0·40 Nm/kgm [0·06], p<0·0001) and narrow-base shoe (0·39 Nm/kgm [0·06], p<0·0001). The second peak was also significantly greater than the barefoot peak value (0·27 Nm/kgm [0·06]) for both the wide-base (0·35 Nm/kgm [0·05], p<0·0001) and narrow-base shoe (0·34 Nm/kgm [0·05], p<0·0001).

Alhough sagittal knee torque was sustained in women who wore both wide-base and narrow-base heeled shoes, there was a 30% greater reduction in peak sagittal knee torque in women who wore wide-base heeled shoes compared with those who were barefoot. This increased knee torque increases the work of the quadriceps muscles, increases the strain through the patella tendon, and increases the pressure across the patellofemoral joint.⁴  The increased patellofemoral pressures occurring with each repetitive step might be important with respect to the development of degenerative joint changes within the patellofemoral compartment. Both types of shoes exaggerated the normal varus torque at the knee by 26% and 22% for the wide-heeled and narrow-heeled shoes, respectively. Increased varus torque imposes greater compressive forces through the medial aspect of the knee, a site typically prone to degenerative joint changes.⁵ 

We have shown that wide-heeled, women’s dress shoes cause the same, if not greater, alterations in knee torques, as narrow-heeled shoes. These findings may have particular importance with respect to the development of knee osteoarthritis, insofar as women tend to wear these wideheeled dress shoes routinely and for longer periods of time.

Department of Physical Medicine and Rehabilitation, Harvard Medical School (D C Kerrigan MD) and Center for Rehabilitation Science, Spaulding Rehabilitation Hospital (J L Lelas MS, M E Karvosky MA), Boston, MA 02114, USA

Correspondence to: Dr D Casey Kerrigan
(e-mail: ckerrigan@partners.org)