Theoretical and experimental indicators of falls during pregnancy as assessed by postural perturbations

Highlights

• Pregnancy is often accompanied by decrements in postural control and falls.

• Postural responses of self-identified pregnant fallers and nonfallers were compared.

• Pregnant fallers had smaller peak COP–COG values in response to perturbations.

• A two-segmented model indicated that pregnant nonfallers had higher ankle stiffness.

• Increasing ankle stiffness could be a strategy to prevent falls in pregnant women.

Abstract

Throughout pregnancy, women experience physical, physiological, and hormonal alterations that are often accompanied by decreased postural control. According to one study, nearly 27% of pregnant women fell while pregnant. This study had two objectives: (1) to characterize the postural responses of pregnant fallers, nonfallers, and controls to surface perturbations, and (2) to develop a mathematical model to gain insights into the postural control strategies of each group. This retrospective analysis used experimental data obtained from 15 women with a fall history during pregnancy, 14 women without a fall history during pregnancy, and 40 nonpregnant controls. Small, medium, and large translational support surface perturbations in the anterior and posterior directions were performed during the pregnant participants’ second and third trimesters. A two-segmented mathematical model of bipedal stance was developed and parameterized, and optimization tools were used to identify ankle and hip stiffness, viscosity, and the feedback time delay by searching for the best fits to experimental COP data. The peak differences between the center of pressure and center of gravity (COP-COG) values were significantly smaller for the pregnant fallers compared with the pregnant nonfallers and controls (p < 0.01). Perturbation magnitude was a significant factor (p < 0.01), but perturbation direction was not (p = 0.24). Model fits were obtained with a mean goodness of fit value of R² = 0.92. Theoretical results indicated that pregnant nonfallers had higher ankle stiffness compared with the pregnant fallers and the controls, which suggests that ankle stiffness itself may be the dominant reason for the different dynamic response characteristics (e.g., peak COP–COG) observed. We conclude that increasing ankle stiffness could be an important strategy to prevent falling by pregnant women.

Introduction

As pregnancy advances, women undergo various physical, physiological, and hormonal alterations. For example, they typically gain 11–16 kg in weight [1]. These weight gains are primarily concentrated in the abdominal region and can increase lumbar lordosis [2]. Hormonal fluctuations can increase ligamentous laxity [3], [4], and changes in plantar foot pressures are observed [5]. Such alterations can lead to balance problems. According to one study, nearly 27% of pregnant women experienced an accidental fall [6], which is a rate comparable to the 30% rate of falls observed in individuals aged 65 yrs and older [7]. Falls that cause fractures and sprains can contribute to the fear of falling [8], while very serious falls can terminate maternal or fetal life [9], [10].

Several researchers have studied the changes in postural control during pregnancy. Butler et al. reported that the center of pressure (COP) excursion in a pregnant group increased in length compared with a control group during quiet stance, and that the amount of weight gained was not significantly associated with the postural sway measures investigated [11]. Nagai et al. showed an increased area of body sway and length of anterior–posterior (A/P) body sway in a pregnant group compared with nonpregnant controls during quiet stance, and that high anxiety correlated with instability [12]. Oliveira et al. reported that pregnant women exhibited larger elliptical fits to COP trajectories as pregnancy progressed and higher COP frequency content along the A/P direction in the absence of visual inputs [13]. However, none of these studies addressed changes of postural control in response to external perturbations.

McCrory et al. investigated pregnant women's responses to A/P support surface translations [14]. Their main finding was that pregnant fallers, who reported at least one fall during pregnancy, had a truncated COP displacement immediately in response to the perturbation compared with pregnant nonfallers and controls.

Both the COP and center of gravity (COG) variables have been used individually to quantify postural stability in biomechanics studies. In terms of postural control during quiet standing, the COG and COP can be interpreted as the controlled and controlling variables, respectively [15], [16], [17], where the COP is proportional to the ankle torque [18]. COP and COG can also be measured simultaneously and the scalar difference between COP and COG (COP–COG) can be computed as a metric to characterize postural control. The COP–COG has been characterized in both the time and frequency domain. The most common metrics associated with the COP–COG variable are amplitude [18], standard deviation [17], root mean square [19], peak magnitude of displacement [20], latencies of initial and peak displacement [20], and frequency spectra [21]. COP–COG metrics have been applied, for example, to elderly stroke patients [19] and healthy elderly and young subjects [17], but have not yet been used to characterize balance in pregnant women.

In terms of mathematical modeling, the inverted pendulum, a one-link representation capturing a single degree of freedom, is the simplest mathematical model for describing bipedal postural control [22]. Simple one and two degree of freedom models have been used to study the effects of biofeedback on individuals with vestibular loss [23] and the risk of falling due to obesity [24]. However, to the best of our to our knowledge, these models have not been applied to pregnancy.

The specific goals of this study are (1) to investigate whether COP–COG can differentiate pregnant fallers from nonfallers; and (2) to use mathematical models to gain insights into the differences in postural control strategies between pregnant fallers and nonfallers.