The Effect of Added Weight on Foot Anthropometry in Pregnant Women and Nulligravida Women

Foot anthropometry is altered by pregnancy. It is unknown if these changes are due to increased bodyweight and/or hormonal concentrations. The purpose of our study was to examine the effect of added weight on foot anthropometry in pregnancy. Fifteen primigravid women and 13 nulliparous controls participated. Controls were matched to pregnant women based on self-reported pre-pregnancy weight. After informed consent, data were collected on pregnant participants in each trimester and post-partum. Foot length, width, arch index, arch height index, arch rigidity index, and arch drop were assessed. Subsequently, pregnant participants in their first two trimesters donned a weighted pack such that total weight difference from pre-pregnancy weight was 124N. Foot measurements were repeated. Third trimester participants were only measured without a pack as they were at full-pregnancy weight. In post-partum, bodyweight plus pack-weight equaled third trimester weight. For control participants, bodyweight plus pack-weight equaled third trimester weight of the pregnant participant to whom they were matched. A MANOVA was performed with the independent variables of trimester, weight condition, and leg. Tukey post-hoc analyses were performed when appropriate (α=0.05).

pregnancy was related to further alterations in anthropometry. Increased pregnancy hormone concentrations likely affect foot anthropometry in primigravid women. Background A number of physiological and anthropometric changes occur during pregnancy and persist post-partum. Pregnant women experience substantial weight gain (1). Due to the increasing abdominal volume associated with the growing fetus, an anterior shift in the center of mass (2), increasing spinal lordosis (3), increasing thoracic extension (4), and decreased strength of the abdominal musculature (5) are noted during pregnancy. An increased serum concentration of relaxin increases the laxity of the sacro-pelvic ligaments to broaden the pelvic inlet to allow for childbirth (6). The effects of relaxin are also seen in peripheral joints, as increased joint laxity in the hand, foot, and knee are noted during pregnancy (7,8).
Anthropometric alterations specific to the foot have been widely reported. Anecdotally, women have described increased shoe size following pregnancy (9)(10)(11). Objective studies have reported increased foot length, width, and volume (7,12). Nyska et al. (1) noted a greater area of the foot in contact with the ground during pregnancy, while others have indicated a decreased arch height (13,14).
However, these alterations are also seen in obese individuals. Increased body weight and/or situational load bearing, such as wearing a rucksack, are related to increased plantar contact area, increased foot length and width, and decreased arch height (15)(16)(17)(18).
It is not known if the changes seen during pregnancy are due to the increased ligamentous laxity related to serum relaxin or to the increased body weight associated with pregnancy.
Once stretched beyond the elastic region, ligaments do not return to their original length (19). Thus, these changes to foot anthropometry are likely to persist post-partum (14).
The altered foot shape may be related to impaired foot function (14) and increased reports of foot pain in post-partum women (20). If this pain does not resolve, this could become a significant life-long concern. A better understanding of changes in foot structure during pregnancy, and their relationship to pain, may allow for improved methods of prevention and treatment.
We have previously reported the changes in foot anthropometry over the course of a first pregnancy (21). Specifically, we examined foot length and width, arch index, arch height index, arch drop, and arch rigidity index (21). Foot length increased significantly such that by the third trimester, the feet of pregnant participants were 3% longer than controls, and this increased length persisted post-partum (Cohen's D = 0.63 with a medium effect size) (21). Arch index also tended to be greater in the third trimester, indicating a lower arch (Cohen's D = 0.40 with a small effect size), although this was not noted post-partum (21).
Thus, this transient change may be due to added weight, increased laxity secondary to pregnancy hormones, or edema of the feet, all of which occur during pregnancy.
The purpose of this study was to examine the effect of added weight on foot anthropometry in primigravida women and nulligravida controls. We hypothesized that added weight, regardless of pregnancy status, will result in increased foot length, width, arch index, and arch drop and decreased arch height index and arch rigidity index.
Furthermore, we hypothesized that these changes will be greater in the primigravida women, particularly in the third trimester, due to the combined effects of added weight and the pregnancy-related hormones.

Participants
For this preliminary study, detection of only very large effect size was sought. To detect a very large effect size when α = 0.05 and β = 0.80, a sample size of 11 per group was required. The investigators expected a dropout rate of 25% amongst pregnant women.
Thus, fifteen primigravid women and thirteen nulliparous controls participated in this study. Each control participant was matched to a pregnant participant based on her selfreported pre-pregnancy body weight.
Participant demographics are shown in Table 1. Control participants were significantly younger than the pregnant participants (p<0.001). Mass was significantly different between groups (p = 0.003), as pregnant participants were significantly heavier than the control participants in their second and third trimesters, but not in their first trimester or post-partum. Participants were recruited through advertisements placed around the university community and obstetrics healthcare facilities. Pregnant women were also recruited from the clinical obstetrics practice of one of our co-authors (PM). Participants were screened over the telephone to determine if they met the enrollment criteria for the study. Inclusion criteria included being between 18 and 45 years old, in a first pregnancy (pregnant group only), or never-pregnant (control group only). Each control participant must have also matched the self-reported pre-pregnancy weight (± 15N) of a pregnant participant.
Exclusion criteria included a history of severe lower limb injury (grade 3 sprain or higher, fracture, or surgery). Additionally, potential participants were excluded if they had lost more than 15% of body weight, or if they were considered by their obstetrician to have a complicated pregnancy (pregnant group only).

Experimental procedures
Data collection took place either at the obstetrics clinic or in our laboratory, based upon the participant's preference. At her initial visit, the experimental procedures were explained to each participant. Following this, written informed consent to participate was obtained from each subject in accordance with the Declaration of Helsinki. The study was approved by the university Institutional Review Board (Protocol # H-23586). Pregnant participants were asked to make a total of four visits: once during the last month of each trimester, and once one to three months post-partum. Control participants made one study visit. Arch index, an indirect indicator of arch height, was measured using an inked footprint according to the methods of Cavanagh and Rodgers (22). Specifically, each participant was asked to stand on the bathroom scale. She was instructed to place one foot on an inkpad (Aetrex Harris mat, Teaneck, NJ) located 15 cm medial to the scale. The inkpad is designed so that the participant's foot did not contact the ink; rather, the ink was located on the underside of a rubber bladder that contacts the paper when weight was placed on it. The participant was instructed to slowly transfer her weight to the inkpad until the scale read half of her weight. Following this, she lifted her foot from the inkpad and the inked footprint was removed from the mat. A line was drawn on the footprint from the most posterior aspect of the heel to the tip of the second toe. The distance from the anterior forefoot (not including the toes), to the posterior heel was measured along this line. The distance was divided into thirds and corresponding marks were made along the line on the footprint such that the foot was divided into the forefoot, midfoot, and rearfoot. The area of the midfoot and the area of the total foot were measured using Image J software (NIH, Bethesda, MD). Arch index is calculated as the area of the midfoot divided by the total area of the foot (22). Thus, a larger number represents a lower arch.
Arch index measures have intra-rater reliability of 0.92323).
The Arch Height Index Measurement System (AHIMS, JAKTOOL Engineered System, Cranbury, NJ), was used to assess arch height index, arch drop, and arch rigidity index according to the methods of Pohl and Farr (24). These measures have an intra-rater reliability between 0.87-0.92 (24,25). The first measurements were taken with the participant seated. The ankle position was standardized such that the line between the first metatarsal head and the lateral malleolus formed a 120° angle with the line from the lateral malleolus to the fibular head, as assessed with a goniometer. Using the AHIMS elevated on two wooden blocks to leave the medial longitudinal arch unsupported, three measurements were obtained on each foot: total foot length (foot length including the toes), truncated foot length (distance from the posterior heel to first metatarsal head), and dorsal height (height to the dorsum at one half of total foot length). The calipers in the AHIMS were then loosened to allow for the participant to stand such that half of her body weight was on the bathroom scale and half was on the wooden blocks. The same three measurements were then obtained in the standing position. Seated arch height index was calculated as seated dorsal height ÷ seated truncated foot length. Standing arch height index was calculated as standing dorsal height ÷ standing truncated foot length. A larger value indicates a higher medial longitudinal arch (24). Arch drop was calculated as seated dorsal height-standing dorsal height. Arch rigidity index was calculated as standing arch height index ÷ seated arch height index (24). An arch rigidity index of 1 indicates a perfectly rigid arch and lower numbers indicate a more flexible arch.
To assess the effect of added weight on these measurements of foot anthropometry, a weighted backpack (Jansport, Appleton, WI) worn on the participant's anterior trunk was used. Pregnant participants in their first two trimesters wore the pack such that total weight difference from their pre-pregnancy weight was 124N, which was based on the average weight gain of pregnant participants in a previous study (26). All standing measurements were repeated while the participants wore this pack. Third trimester participants did not wear a pack as they were at full-pregnancy weight.

Main effect of Added Weight
Arch drop and arch rigidity index were both significantly altered by adding weight to the participants. Arch drop increased by 18% (p = 0.007) and arch rigidity index decreased by 1% (p = 0.001) in the weighted condition across both the pregnant and control groups.
Data are provided in Table 2. No other variables were significantly different between the natural and weighted conditions.
Main effect of Leg Leg tested (left or right) was included in our statistical model because the legs were not independent observations. Also, we wanted to ascertain if the effect of added weight was the same bilaterally. No differences were noted between legs on any variables expect standing AHI (p = 0.04). The right standing AHI (0.319 ± 0.025) was lower than the left standing AHI (0.326 ± 0.026), indicating a slightly lower arch on the right leg in all participants. No interactions involving leg were significant (trimester x leg or weighted condition x leg). Thus, pregnant women did not demonstrate a greater asymmetry of standing AHI between the left and right legs than did the controls.

Interaction between Added Weight and Trimester
The condition (natural, weighted) x trimester interaction was significant in several variables ( Table 2). Increase in foot length with added weight was greater in pregnant participants vs controls, with the effect of the added weight increasing throughout pregnancy (p = 0.005), although the values were not statistically different between the trimesters/post-partum in the pregnant participants. Similarly, the condition x trimester interaction was significant in the foot width measure (p = 0.05) as the foot width was greater in the weighted condition in the latter trimesters such that in the weighted condition, the foot width of the pregnant women in the second and third trimesters and post-partum was significantly greater than control group. Foot width in the third trimester and post-partum was significantly greater than during the first trimester, and it was greater post-partum than it was in the third trimester. Arch index also saw a condition x trimester interaction (p = 0.037), such that when the participant was weighted, arch index was greater in the second and third trimesters and postpartum than in the controls.
Finally, a significant condition x trimester interaction was noted in the standing arch height index variable (p = 0.007). In the weighted condition, controls displayed a larger arch height index value than in the first and second trimesters and postpartum; pregnant participants in the weighted condition in their second and third trimesters and postpartum displayed a greater arch height index value than when they were in their first trimester, and their weighted arch height index values in their third trimester were less than in the second trimester. There were no significant interaction effects for arch drop or arch rigidity index.

Discussion
The purpose of this study was to examine the effect of added weight on foot anthropometry in pregnant women and non-pregnant controls. Results for the main effect of "trimester" have been published in (21). The current research paper is focused on the effect of added weight on foot anthropometry as well as the interaction between added weight and trimester. We hypothesized that added weight will result in increased foot length, width, arch index, and arch drop and decreased arch height index and arch rigidity index. We also hypothesized that these changes will be greater in the pregnant women, particularly in the third trimester, due to the combined effects of added weight and the pregnancy-related hormones.
Our results partially support these hypotheses. Simply adding weight to the participants via a weighted pack produced significant changes in arch drop and arch rigidity index. All participants displayed a greater arch drop while weighted, and a corresponding lower arch rigidity, indicating a more flexible arch, compared to when they were in the "natural" condition. Because these alterations were seen not only in the pregnant participants, but equally in the control participants who did not have the circulating pregnancy hormones such as relaxin, this indicates that marked weight gain alone will alter medial longitudinal arch anthropometry. Hills et al. reported increased contact of the midfoot with the ground in obese individuals (16).
Furthermore, added weight plus pregnancy, as in the condition x trimester interaction, was related to additional alterations to foot anthropometry. This suggests that other pregnancy-associated factors in addition to weight, such as increased hormone concentrations (e.g. relaxin), may play a role in foot anthropometry changes in pregnancy.
Foot length and width were significantly greater as pregnancy progressed (21), but this difference was exacerbated in the weighted condition in the current analysis.
Arch index was greater in the second and third trimesters and post-partum in the weighted condition compared to controls, indicating more contact area of the midfoot with the ground. Previously, we reported that arch index was only greater in the third trimester when compared to controls when examining the effects of trimester alone (21), and this effect was not seen post-partum; thus, the current study signifies that adding weight during pregnancy, when relaxin concentration is high, results in a greater arch index and a presumably lower medial longitudinal arch, than just pregnancy alone. While lower leg edema may affect the arch index measure during pregnancy, a greater arch index was also noted post-partum when swelling would not be a concern (27). The lower arch index is not due to a greater bodyweight after the pregnancy because by the time the participants were tested post-partum, their weight had returned to its pre-pregnancy value. Our data support the conclusion that the medial longitudinal arch is more lax following pregnancy, and the carrying of an additional weight may result in more arch deformity than prior to the pregnancy. As it would be expected that a post-partum women carry a child in her arms for the next few years, this increased laxity of her medial longitudinal arch may be a source of foot dysfunction or pain.
Arch height index is another measure that corresponds to the height of the medial longitudinal arch. A smaller height to the dorsum and/or a longer foot would result in a smaller arch height index, indicating a lower arch. In our previous study, we reported that no alterations were noted due to the main effects of trimester alone (21), despite the fact that foot length was increased during pregnancy. However, the current analysis finds that the condition x trimester interaction was significant such that, in the weighted condition, controls displayed a greater arch height index than the pregnant women in their first two trimesters and postpartum; first trimester participants displayed a greater arch height index than in the second and third trimesters and post-partum, and the third trimester values were greater than when the women were in their second trimester. We do not know what effect, if any, swelling had on our results. Swelling is very common, as more than eighty-five percent of women report swelling of the lower leg and foot during pregnancy (9). The incidence and magnitude of lower leg edema peaks in the third trimester (28). It may be that in their third trimesters, the feet of the pregnant women were swollen enough so that the dorsal height of the midfoot read higher on the AHIMS, thus reading as a higher arch. Perhaps the fact that the arch height index was lower in the weighted condition in the first and second trimesters and post-partum when compared to the third trimester condition was that swelling gave the foot more height, and this not seen in the other time points in pregnancy. Because no differences were noted between pregnancy trimesters in the natural condition (21), it could be that it took the additional weight to lower the arch in the other trimesters, but the arch could not get lower in the third trimester because of the swelling on the plantar aspect of the foot.
We could not assess any viscoelastic effects of the influence of added weight over a period of time, as in a real-life pregnancy that lasts nine months. Rather, we examined the short-term effects of adding a weight for a portion of the data collection session. The average pregnant participant in our study gained 16.2±6.3 kg by her third trimester visit.
It would not be feasible to continually add this amount of weight over the course of nine months to a non-pregnant study participant to determine the effects of added weight alone. This is a limitation of the study.
Our small sample size of 15 pregnant women and 13 nulligravida women is a limitation of the study. With this small number, we may not be able to accurately detect small, medium, or large between-group effects. Given the small sample size and our limited ability to detect effect sizes, some small group differences of 1% or 0.5mm that were statistically significant may not be clinically meaningful.
A further limitation of this study is that the pregnant participants were significantly older than the non-pregnant group. It was not feasible to match on age, as more than 80% of women have been pregnant by age 50, thus finding nulliparous women in that age group is difficult. However, previous work has demonstrated that there is no relationship between age and arch height or stiffness (29). Thus, we do not expect that the five year age difference between groups had a strong influence on our results.
Finally, these results are only generalizable to primigravida women. Segal et al. (14) reported that the greatest changes in arch height and rigidity occur during a first pregnancy. Thus it is likely that changes occurring during the first pregnancy persist, and thereafter there is little room for the arch to drop further.
It should be noted that our pregnant participants in this cohort reported little or no pain on a questionnaire asking them to rate their pain in the lumbar spine, posterior pelvis, and foot on a visual analogue scale over the course of their pregnancies (21). Given that fifty-six percent of pregnant women report lower leg and foot pain (20), and that this pain may be related to changes in foot anthropometry, it is feasible that our study examined a cohort with minimal foot alterations. Women who report significant pain may also develop more severe foot alterations and dysfunction as a result of the pregnancy. We are currently collecting data on the relationship of development of these anthropometric alterations to self-reported pain in pregnant women in their third trimester who report significant pain.
By the age of 50, more than 80% of American women have experienced at least one pregnancy (30). More than half of women report leg and foot pain during a pregnancy (9), with the incidence of pain increasing with the number of pregnancies (31). Pregnancy leads to long-term changes to foot structure and function, and these changes may be related to the incidence of pain (10,14,32). Thus, prevention of these physical changes to foot anthropometry may be of clinical concern. While we realize that we will not be able to, nor should we try to, alter serum hormone concentrations in pregnancy, it is important for us to understand the differential contribution of maternal weight gain and pregnancy related-hormones to changes to foot anthropometry and consequential pain so that we may be better able to prevent the pain. Additionally, an understanding that the interaction of weight gain and pregnancy hormones results in different variations of foot shape alterations than just weight gain alone will be valuable in prevention of long-term dysfunction and pain post-partum. For example, it would be beneficial to know if all pregnant women, or only those who have large weight gain during pregnancy, should be instructed to wear shoes with arch support or even prescribed orthotics. While the effect of orthotics on pain has not been evaluated in pregnant women, orthotics are known to reduce risk and severity of lower extremity pain and injury in other adult populations (33)(34)(35). Also, it would not be expected that pregnant women would experience the same changes to foot anthropometry that occur with load carriage or obesity, given that the interaction of pregnancy hormones and added weight results in different changes to foot dimensions than just added weight alone. The current results suggest clinicians should consider the additive effect of weight and pregnancy status when evaluating and treating changes in foot anthropometry in pregnant women.

Conclusions
The combination of the increased bodyweight associated with pregnancy and the increased concentration of serum relaxin are related to anthropometric alterations to the feet of pregnant women. Added weight alone resulted in increased arch drop and a decreased arch rigidity index. The interaction of added weight and trimester resulted in increased foot length and width, as well as a lower medial longitudinal arch, as indicated by the arch index and arch height index measures. These alterations persist post-partum, and are likely related to life-long foot pain in women who have experienced a pregnancy, which is a majority of the female population. Thus, this pain is of significant concern. An understanding of the differential contributions of the pregnancy related hormones and increased bodyweight may better enable clinicians to prevent and/or treat this pain.

Declarations
Ethics approval and consent to participate This study was approved by the WVU Institutional Review Board. [Protocol reference number: H-23586] Approved under the title: "Relationship between changes in foot and pelvis alignment and the incidence of lumbar, posterior pelvis, and foot pain during pregnancy." All study participants provided informed consent prior to their participation in the study.

Consent for publication
Not applicable. This manuscript does not contain data from any one individual person.

Availability of data and material
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing Interests
The authors declare that they have no competing interests.

Funding
This study was not funded. "weighted". † denotes significant difference between natural and weighted a significant difference from controls b significant difference from 1 st trimester weighted c significant difference from 2 nd trimester weighted d significant difference from 3 rd trimester e significant difference from post-partum weighted