BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
Lifelong Chinese Bound Feet – A Quantitative Ultrasound and Lifestyle Questionnaire Study in Postmenopausal Women
For peer review only Journal: BMJ Open
Manuscript ID: bmjopen-2014-006521
Article Type: Research
Date Submitted by the Author: 15-Sep-2014
Complete List of Authors: Qin, Ling; The Chinese University of Hong Kong, ; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology Pan, Yi; Second Affiliated Hospital of Kunming Medical University, Department of Endocrinology Zhang, Ming; The Hong Kong Polytechnic University, Department of Mechanical Engineering Xu, Mian; Second Affiliated Hospital of Kunming Medical University, Department of Endocrinology Lao, Hanchang; Second Affiliated Hospital of Kunming Medical University, Department of Orthopaedics O’Laughlin, Michael; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology
Tong, Shan; The Hong Kong Polytechnic University, Department of http://bmjopen.bmj.com/ Mechanical Engineering Zhao, Yanling; Health Promotion Foundation of China, Training Department of International Osteoporosis Diagnosis and Treatment Cheng, JCY; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology Guo, Xia; The Hong Kong Polytechnic University, Department of Rehabilitation Sciences
Primary Subject
Sports and exercise medicine on September 29, 2021 by guest. Protected copyright. Heading:
Secondary Subject Heading: Sports and exercise medicine
ORTHOPAEDIC & TRAUMA SURGERY, Adult orthopaedics < ORTHOPAEDIC Keywords: & TRAUMA SURGERY, Foot & ankle < ORTHOPAEDIC & TRAUMA SURGERY
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 Lifelong Chinese Bound Feet – A Quantitative Ultrasound and Lifestyle Questionnaire 6 7 8 Study in Postmenopausal Women 9 10 1* 2 3 2 4 11 Ling Qin PhD; Yi Pan MD; Ming Zhang PhD; Mian Xu MD; Hanchang Lao MD; Michael 12 13 C. O’Laughlin1 BSc; Shan Tong3 MPhil; Yanling Zhao5 MD; Cheng JCY1 MD; Xia Guo6 MD 14 15 For peer review only 16 17 1. Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 18 19 the Chinese University of Hong Kong, Hong Kong SAR, China. 20 21 2. Department of Endocrinology, the Second Affiliated Hospital of Kunming Medical 22 23 24 University, Kunming, Yunnan Province, China 25 26 3. Department of Mechanical Engineering, the Hong Kong Polytechnic University, Hong Kong 27 28 SAR, China. 29 30 31 4. Department of Orthopaedics, the Second Affiliated Hospital of Kunming Medical University, 32 33
Kunming, Yunnan Province, China http://bmjopen.bmj.com/ 34 35 5. Training Department of International Osteoporosis Diagnosis and Treatment, Health 36 37 38 Promotion Foundation of China, Beijing China 39 40 6. Department of Rehabilitation Sciences, the Hong Kong Polytechnic University, Hong Kong 41 42 on September 29, 2021 by guest. Protected copyright. 43 SAR, China. 44 45 46 *: Corresponding author: Ling QIN (PhD), e mail: [email protected] 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 Abstract 3 4 5 6 Objective The phenomenon of foot binding or also known as "lotus feet" has an enduring and 7 8 influential history in China. Due to man made smaller foot size, the life long foot binding may 9 10 11 impose adverse effect on skeletons and we investigated the bone properties in postmenopausal 12 13 14 women with bound feet that may provide new information to develop appreciate 15 For peer review only 16 countermeasures for prevention of fragility fractures. 17 18 19 Design Population based cohort study. 20 21 22 Participants This study involved 254 postmenopausal women aged between 65 and 80, including 23 24 25 172 with bounded feet and 82 age and gender matched control subjects living in a remote region 26 27 of China. 28 29 30 Outcomes Anthropometric data, SF 36 life style questionnaire, and heel quantitative ultrasound 31 32 33 (QUS) were collected for the entire study population. A small subset was also invited for http://bmjopen.bmj.com/ 34 35 assessment of bone mineral and microarchitecture at distal tibiae using high resolution peripheral 36 37 38 quantitative computed tomography (HR pQCT), gait and balance. 39 40 41 Results Women with bound feet had significantly lower QUS value as compared to age matched 42 on September 29, 2021 by guest. Protected copyright. 43 women with normal feet, supported by HR pQCT data as well. However, SF 36 questionnaire 44 45 46 results did not reveal statistically significant differences in any categorical responses, including 47 48 physical functioning, general health vitality and physical component summary score, number of 49 50 previous fractures. No impaired body balance was found in the small set of group. 51 52 53 Conclusions The man made changes of foot binding led to reduced physical activity and 54 55 56 therefore were prone to osteoporosis. A compensation mechanism for women with bound feet 57 58 with no difference in fragility fracture might be related on not affected capability in body balance 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 tested for the small subset, implying the importance of improving or maintaining body balance in 3 4 overall prevention strategy of fragility fractures confirmed in this unique foot binding study. 5 6 7 8 Introduction 9 10 11 Few phenomena in history have been brought to such a high level of esteem and popularity with 12 13 14 a perplexing origin, and even fewer have shared such extensive, enduring impacts. For over a 15 For peer review only 16 millennium, Chinese women of all ages and social classes bounded their feet to a fraction of its 17 18 original size, suffering great pain and hardship during the process. Literature documented that 19 20 21 this tradition began in the closing years of the tenth century by either dancers during the Song 22 23 dynasty that valued the aesthetic beauty and graceful movements of small feet or a hedonistic 24 25 26 emperor who ordered his concubines to bind their feet to resemble the shape of a “golden lotus” 27 1,2 28 leaf. One aspect beyond debate, however, is the widespread influence of foot binding and the 29 30 longevity of its practice: spanning over 1000 years and four dynasties in China. The tradition 31 32 th 33 reached its apex during the 18 century, where around over 50% of all women and close to 100% http://bmjopen.bmj.com/ 34 35 of women in upper class societies bound their feet.2,3 36 37 38 Foot binding was first outlawed in 1912 but persisted in various rural regions of China. Women 39 40 41 with small feet were associated with restricted labor and physical activity that promoted 42 on September 29, 2021 by guest. Protected copyright. 1 3 43 feministic behavior and increased one’s likelihood of marriage. Nowadays, older women aged 44 45 65 or above have become the very last generation still living with bound feet. These women 46 47 48 mainly reside in rural and impoverished provinces of China. However, while the phenomenon is 49 50 well documented, very few previous reports have collected scientific data to study the direct 51 52 effect of lifelong foot binding to one’s bone statues and fragility fractures. 53 54 55 56 One of the observable consequences of lifelong foot binding is the reduced foot size and 57 58 accordingly significant decrease in support area of the foot that greatly hinders the ability for one 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 to participate in both social and physical activities throughout one’s lifetime (Figure 1). Physical 3 4 exercise has previously been shown with strong evidence of enhancing bone health.4 7 The 5 6 7 decreased support given by the smaller foot area would be likely to reduce balancing capabilities, 8 9 lead to a greater risk factor of falling and cause higher incidence of fragility fracture due to 10 11 7 9 12 restricted physical activity. We therefore put forward our study hypothesis that women with 13 14 bound feet had adverse impact on musculoskeletal health due to diminished lifelong tendencies 15 For peer review only 16 of physical activity, decrease in ability to balance as a result of smaller foot support area and 17 18 19 therefore higher fall frequency and fall related fragility fractures. 20 21 22 23 24 25 Methods 26 27 28 29 Participants 30 31 32 This was a project conducted by both clinical scientists and bioengineers from Hong Kong and 33 http://bmjopen.bmj.com/ 34 physicians from Mainland China to reveal the social and physiological impacts of lifelong foot 35 36 37 binding with participants recruited from Luliang County of Yunan Province, China (Supplement 38 39 I). Participants who had any diseases or drug use that were known to affect metabolism of 40 41
musculoskeletal system as well as recent fracture (within 1 year) were excluded. The study was on September 29, 2021 by guest. Protected copyright. 42 43 44 supported by the Chinese Health Promotion Foundation (Appendix A) and human ethics 45 46 approval was obtained from the Second Affiliated Hospital of Kunming Medical Collage for the 47 48 49 part of study conducted in Kunming (Appendix B) and the Chinese University of Hong Kong 50 51 for the part conducted in Hong Kong (Appendix C). 52 53 54 A total of 254 subjects were included into the study. Of the 254 subjects, 172 subjects had bound 55 56 feet (mean age 74.6±3.5, range 65 80 years) and 82 subjects had normal feet (mean 74.5±4.0, 57 58 59 range 69 87 years). All subjects had their weight and height measured and body mass index 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 (BMI) calculated. Previous history of fractures, number of children, and years since menopause 3 4 (YSM) was also documented. 5 6 7 8 A modified 36 item short form health survey (SF 36) was given to all participants after 9 10 translation into Chinese that was validated by one of our previous studies.10 The survey was 11 12 organized into eight multi item scales assessing: physical functioning (PF), physical role (PR), 13 14 15 bodily pain (BP),For general health peer (GH), vitality review (VT), social functioning only (SF), emotional role (ER) 16 17 on daily activities and mental health (MH) that comprehensively determine the quality of life, 18 19 20 physical, mental, and psychosocial health (Appendix D). Of the eight categories, five categories 21 22 that govern the physical well being of individuals, including PF, PR, BP, GH and VT, were 23 24 linearly transformed to a 0 100 scale using a standardized three step algorithm and subsequently 25 26 11,12 27 scaled by computing the mean values for each category. Mean values for each parameter 28 29 between the two groups were compared to determine differences between bound feet and control 30 31 subjects. Finally, scaled scores in each category were combined to produce a higher order 32 33 http://bmjopen.bmj.com/ 34 physical component summary (PCS) score. The PCS score was standardized with the 1998 U.S. 35 36 general population database with a mean score of 50 and S.D. of 10.13 37 38 39 Heel quantitative ultrasound (QUS) 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 All 254 recruited subjects participated in quantitative ultrasound (QUS) evaluation at heel in 44 45 Luliang county using Lunar Achilles Plus (GE Healthcare, Milwaukee, WI, USA) (Figure 2). 46 47 Speed of sound (SOS) (m/s) and broadband ultrasound attenuation (BUA) (Db/MHz cm) were 48 49 50 recorded. A stiffness index (SI) was derived from SOS and BUA based on following formula: 51 52 SI = (0.67 × BUA) + (0.28 × SOS) – 420.14 QUS T scores were calculated using built in Chinese 53 54 female reference population. Z values were subsequently calculated with the same formula, 55 56 57 expressed as SD above or below the mean SI values of an age, gender and weight matched 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 reference database. Instead of a conventional T score ≤ 2.5 SD, a T score ≤ 1.8 SD was used as 3 4 cutoff value for defining osteoporosis, as previous studies demonstrated that the T score 5 6 7 threshold of 2.5 SD might lead to underestimation of the prevalence of osteoporosis when QUS 8 9 was used for bone measurement at the heel 15,16 while QUS T score threshold of 1.8 SD 10 11 12 identified the same percentage of persons with osteoporosis as the World Health Organization 13 14 (WHO) threshold for bone mineral density (BMD) measurements using Dual energy X ray 15 For peer review only 16 absorptiometry (DXA). 16,17 17 18 19 20 Due to significant physical distance between Luliang County and Hong Kong (Supplement I) 21 22 and lacking financial support for international traveling, only two women with bound feet with 23 24 mean QUS values representative of the bound feet group were invited to Hong Kong for case 25 26 27 analysis using below specified advanced assessments, including 3 D bone densitometry 28 29 (Supplement II), gait and body balance analysis relevant tests for predicting fall and fragility 30 31 fractures (Supplement III). 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 Statistical analysis 39 40 41 We used Statistical Package for Social Sciences (SPSS for Windows version 19.0, SPSS Inc., 42 on September 29, 2021 by guest. Protected copyright. 43 Chicago, IL, USA) for statistical analysis. Quantitative data was expressed as mean±SD for 44 45 46 continuous data and counts by percentage for discrete data. Student’s t test was used to compare 47 48 parameters between two groups, including age, BMI, size of foot, center of gravity, number of 49 50 children, YSM, previous fractures, and rate of T score based osteoporosis. Chi square test was 51 52 53 used to study the association between previous fracture rate and bound feet status and between 54 55 T score of QUS data and bound feet status. For SF 36 questionnaire data, we sued Student’s t 56 57 58 test to determine if statistical significance existed between categorical mean values of subjects. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 Chi square test was used for testing statistical difference in rate of osteoporosis and incidence of 3 4 low energy fractures between two groups. Values of P < 0.05 were considered statistically 5 6 7 significant. 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 Results 3 4 5 Anthropometric data and SF-36 questionnaire 6 7 8 Table 1 shows no statistically difference in age between postmenopausal women with and 9 10 without bound feet, implying a homogenous grouping for comparison. Significant differences are 11 12 13 however seen in body weight, BMI, length and width of feet, number of children, and T score 14 15 based rate of osteoporosisFor betweenpeer the two review groups. SF 36 questionnaire only results are summarized 16 17 in Table 2 where no statistically significant differences are found in any categorical responses 18 19 20 between women with and without bound feet except the vitality score that shows a noticeable but 21 22 not statistically significant decrease in bound feet subjects. 23 24 25 Heel quantitative ultrasound (QUS) 26 27 28 Figure 5A shows T and Z scores of heel QUS bone stiffness index. The mean T and Z scores of 29 30 31 women with bound feet are both significant lower than that of the age and gender matched 32 33
control subjects (p<0.001 for both) (Figure 3). T score based calculation shows 32.51% http://bmjopen.bmj.com/ 34 35 36 significantly more osteoporotic cases found in women with bound feet (165 out of 172, i.e. 37 38 95.93%) compared with age and gender matched control subjects (52 out of 82, i.e. 63.42%) 39 40 (p<0.01). 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 Discussion 3 4 5 Lifelong foot binding is a widespread tradition of ancient China that until this day still affects a 6 7 significant number of elderly women living in rural regions. We investigated potential adverse 8 9 impact of foot binding on musculoskeletal system, including bone mineral and structural statues 10 11 12 along with major fracture risk factors and fracture incidence. Our findings confirmed our 13 14 hypothesis that women with bound feet had adverse impact on bone health but disproved our 15 For peer review only 16 concerns on osteoporosis associated increase in fragility fractures. 17 18 19 20 Due to availability, we only used a portable heel QUS for bone scan in Luliang County, a remote 21 22 mountain area of Kunming province to quantify bone statues of 172 postmenopausal women 23 24 with bound feet and their peers. However, calcaneus bone is a skeletal site that is directly 25 26 27 affected by foot binding and associated alterations in physical activities. Our QUS measurement 28 29 showed 1/3 higher osteoporotic rate in bound feet women than that of the controls. HR pQCT 30 31 used for a small subset evaluation for the current study is an advanced technology that allows 32 33 http://bmjopen.bmj.com/ 34 insight into volumetric BMD (vBMD) and microarchitecture. Trabecular microarchitecture 35 36 parameters of the representative subjects indicated that bone quality is inferior in bound feet 37 38 39 women with on average 20% less trabecular number and half of its vBMD in core region of the 40 41 distal tibia (Supplement II), implying systemic adverse impact of foot binding on weight 42 on September 29, 2021 by guest. Protected copyright. 43 bearing bones. 44 45 46 Previous QUS studies in normal postmenopausal women demonstrated the value of QUS for 47 48 17 19 49 evaluation of osteoporosis and predicting fragility fractures. As fragility fractures were 50 51 associated with both skeletal and non skeletal factors, we adopted 36 questionnaires, a validated 52 53 54 instrument for evaluation of relationship between physical activity and osteoporosis and/or 55 13, 21 56 fragility fractures into the current study. Impaired body balance has been a known risk factor 57 58 of fall associated fragility fractures in women with normal feet. 8,22,23 However, our analysis only 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 showed non detectable inferior vitality and physical functioning in women with bound feet and 3 4 the differences in accumulated plantar pressure with more pressure placed on the calcaneus of 5 6 7 the heel and less force distributed over the forefoot inbound feet (Supplement II), implying an 8 9 compensation mechanism in human kinesiology after foot binding in later life. Our data did not 10 11 12 show the difference in number of previous fractures in women with or without foot binding. 13 14 15 An early as inFor 1997, Cummingspeer and review co workers reported only a first Dual energy X ray 16 17 Absorptiometry (DXA) study to measure BMD at spine and hip in post menopausal women with 18 19 20 bound feet in Beijing, the capital city of China, and reported lifelong foot binding had significant 21 22 adverse effects on BMD at spine BMD and higher fragility fracture rate when compared to the 23 24 control subjects.24 The direct adverse impact to the bound feet is the significant reduced foot size 25 26 27 which is described as “golden lotus”. The associated consequence is an increased tendency of 28 29 falling and increased risk of fall induced fractures in later life. The overall higher fracture rate in 30 31 Beijing study was 15% while significant lower fracture incidence with 7.5% was found in our 32 33 http://bmjopen.bmj.com/ 34 current subjects that might be explained by different lifestyles between women living in urban 35 36 areas in Beijing of Cumming’s study24 and rural regions of our current one conducted in Luliang, 37 38 a remote county in Yunnan province. There were no balance and gait data available in Beijing’s 39 40 41 study. The bound feet subjects in Beijing were mostly descendants of the upper class families 42 on September 29, 2021 by guest. Protected copyright. 43 and little physical work was required in their later life, with speculated impairment in 44 45 46 musculoskeletal coordination to explain both higher osteoporosis and fall associated fragility 47 1,3,24,25 48 fracture rate. As our study subjects in remote Luliang county were working class women 49 50 and often had to perform mandatory labor work in old age, this might explain that the balance 51 52 53 analysis and life style questionnaire did not show difference between women with and without 54 55 bound feet evaluated in a small subset of subjects (Supplement III). The associated active 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 lifestyle might have led to different adaptations of the body balance throughout one’s lifetime to 3 4 compensate smaller foot support area. 5 6 7 8 The findings of our study on heel QUS and lifestyle in terms of SF 36 questionnaire offer a 9 10 unique look into the effects of foot binding. The significance of our current study in social and 11 12 health care aspects was the mobilization of the local government to support our project. Our 13 14 15 findings generatedFor from thepeer current study review raised the awareness only of osteoporosis prevention 16 17 measures in rural regions of China at one site and at another. The importance of physical 18 19 20 activities and/or exercise that are specifically relevant for balance maintenance as this has been a 21 7,10,26,27 22 known and crucial factor for prevention of fall and fall induced fragility fractures. 23 24 25 Our current study had a few limitations apart from the cross sectional study in nature: 1) 26 27 Inhomogeneous number in study subjects: the higher percentage of bound feet subjects was due 28 29 30 to the greater percentage of elderly women with bound feet living in the location of study; 2) due 31 32 to limited budget and long distance between remote Luliang country and Hong Kong, only few 33 http://bmjopen.bmj.com/ 34 35 subjects were invited to fly to Hong Kong for HR pQCT, gait and balance tests (Supplements II 36 37 and III); 3) As only fracture history could be recalled for analysis, no reliable fall frequency 38 39 could be provided for analysis. 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 Conclusion 47 48 49 Foot binding had life long negative impact on bone health. However, the neuromusculoskeletal 50 51 52 coordination of the bound feet postmenopausal women with active life style might be 53 54 compensated as no difference was revealed in incidence of fragility fractures, implying the 55 56 importance of body balance maintenance in prevention of fragility fractures. 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 Funding This study was supported by the China Health Promotion Foundation (Ref. 2009 01) 5 6 7 and Contemporary Orthopaedic Research and Education Fund of the Chinese University of Hong 8 9 Kong (CUHK#6903361). We also appreciate technical support from Miss Fong TN and Miss 10 11 12 Hung WY of Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 13 14 The Chinese University of Hong Kong for their assistance in HR pQCT measurements. 15 For peer review only 16 17 18 19 20 Competing interests: The authors declare no competing interests. 21 22 23 24 Contribution statement: Ling Qin designed and organized this project, participated in data 25 26 27 collection, analysis and interpretation, and guided the preparation of this Manuscript; Yi Pan, 28 29 Mian Xu, and Hanchang Lao recruited subjects, performed questionnaires and conducted QUS 30 31 scanning at Luliang county; Ming Zhang and Shan Cong performed gait analysis, and Xia Guo 32 33 http://bmjopen.bmj.com/ 34 performed balance test and related analysis in Hong Kong. Yanling Zhao co organized the study 35 36 and jointly applied for research grant for this project; Michael O’Laughlin did a part of the data 37 38 analysis and assisted in drafting the manuscript. JCY Cheng contributed to planning and 39 40 41 preparation of the manuscript. 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 Data sharing: No additional data available 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 References 3 4 5 1. Levy HS. Chinese Footbinding: The history of a curious erotic custom. New York, NY: W. 6 7 8 Rawls, 1966 9 10 2. Wang P. Aching for beauty: Foot binding in China. Minneapolis: University of Minnesota 11 12 Press, 2000 13 14 15 3. Fairbank JK.For The Great peerChinese Revolution: review 1800 1985 New only York, Harper & Row, 1986 16 17 4. Gerdhem P, Akesson K, Obrant KJ. Effect of previous and present physical activity on bone 18 19 20 mass in elderly women. Osteoporosis Int 2003; 14:208–12. 21 22 5. Cooper C, Harvey NC. Osteoporosis risk assessment. BMJ 2012; 344:e3427. 23 24 6. Qin L, Au SZ, Choy YW, Leung PC, Neff M, KM Lee, MC Lau, Woo J, Chan KM. Regular 25 26 27 Tai Chi exercise may retard bone loss in postmenopausal women – A case control study. 28 29 Arch Phys Med Reha 2002; 83(10):1355 9. 30 31 7. Qin L, Choy WY, Leung KS, Leung PC, Au SK, Hung WY, Dambacher MA, Chan KM. 32 33 http://bmjopen.bmj.com/ 34 Beneficial effects of regular Tai Chi exercise on musculoskeletal system. J Bone Miner 35 36 Metabol 2005; 23(2):186 90. 37 38 8. Edwards MH, Jameson K, Denison H, Harvey NC, Sayer AA, Dennison EM, Cooper C. 39 40 41 Clinical risk factors, bone density and fall history in the prediction of incident fracture among 42 on September 29, 2021 by guest. Protected copyright. 43 men and women. Bone 2013; 52(2):541 7. 44 45 46 9. Hippisley Cox J, Coupland C. Derivation and validation of updated QFracture algorithm to 47 48 predict risk of osteoporotic fracture in primary care in the United Kingdom: prospective open 49 50 cohort study. BMJ 2012; 344:e4191; 1 16. 51 52 53 10. Chan KM, Qin L, Lau M, Woo J, Au SK, Choy WY, Lee KM, Lee SH. A randomized, 54 55 prospective study of the effects of Tai Chi Chun exercise on bone mineral density in 56 57 postmenopausal women. Arch Phys Med Rehab 2004; 85:717 22. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 11. Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF 36). I. Conceptual 3 4 framework and item selection. Med Care 1992; 30:473–83. 5 6 7 12. Ware JE, Kosinski M, Bayliss MS, McHormey, Rogers WH, Raczek A. Comparison of 8 9 methods for the scoring and statistical analysis of SF 36 health profile and summary 10 11 12 measures: summary of results from the Medical Outcomes Study. Med Care 1995; 33(4): 13 14 264 79. 15 For peer review only 16 13. Maglinte GA, Hays RD, Kaplan RM. US general population norms for telephone 17 18 19 administration of the SF 36v2. J Clin Epidemiol 2012; 65(5): 497–502. 20 21 14. Economos CD, Sacheck JM, Wacker W, Shea K, Naumova EN Precision of Lunar Achilles+ 22 23 bone quality measurements: time dependency and multiple machine use in field studies. Br J 24 25 26 Radiol 2007; 80:919 25. 27 28 15. Vu TT, Nguyen CK, Nguyen TL, Le BM, Nguyen Trung Le Dm Bui TN, Bakamori M, 29 30 31 Kunii D, Sakai T, Yamamoto S. Determining the prevalence of osteoporosis and related 32 33 factors using quantitative ultrasound in Vietnamese adult women. Am J Epidemiol 2005; http://bmjopen.bmj.com/ 34 35 161(9):824 30. 36 37 38 16. Frost ML, Blake GM, Fogelman I. Can the WHO criteria for diagnosing osteoporosis be 39 40 applied to calcaneal quantitative ultrasound? Osteoporos Int 2000; 11:321–30. 41 on September 29, 2021 by guest. Protected copyright. 42 17. Liu JM, Ma LY, Bi YF, Huang Y, Xu M, Zhao HY, Sun LH, Tao B, Li XY, Wang WQ, 43 44 45 Ning G. A population based study examining Calcaneus quantitative ultrasound and its 46 47 optimal cut points to discriminate osteoporotic fractures among 9352 Chinese women and 48 49 men. J Clin Endocrinol Metab 2012: 97(3):800 9. 50 51 52 18. Hans D, Dargent Molina P, Schott AM, Sebert JL, Cormier C, Kotzki PO, Delmas PD, 53 54 Pouilles JM, Breart G, Meunier PJ. Ultrasonographic heel measurements to predict hip 55 56 57 fracture in elderly women: the EPIDOS prospective study. Lancet 1996; 348:511–14. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 19. Glueer CC. Quantitative ultrasound techniques for the assessment of osteoporosis: expert 3 4 agreement on current status. J Bone Miner Res 1997; 12:1280–8. 5 6 7 20. Nayak S, Olkin I, Liu H, Grabe M, Gould MK, Allen IE, Owens DK, Bravata DM. Meta 8 9 analysis: accuracy of quantitative ultrasound for identifying patients with osteoporosis. Ann 10 11 12 Intern Med 2006; 6;144(11):832 41. 13 14 21. Lai BM, Tsang SW, Lam CL, Kung AW. Validation of the Quality of Life Questionnaire of 15 For peer review only 16 the European Foundation for Osteoporosis (QUALEFFO 31) in Chinese. Clin Rheumatol 17 18 19 2010; 29(9):965 72. 20 21 22. Kanis JA. Diagnosis of osteoporosis and assessment of fracture risk. Lancet 2002; 22 23 24 359(9321):1929 36. 25 26 23. Kern LM, Powe NR, Levine MA, Fitzpatrick AL, Harris TB, Robbins J, Fried LP. 27 28 Association between screening for osteoporosis and the incidence of hip fracture. Ann Intern 29 30 31 Med 2005; 142(3):173 81. 32 33 24. Cummings S, Ling X, Stone K. Consequences of foot binding among older women in Beijing, http://bmjopen.bmj.com/ 34 35 China. Am J Pub Health 1997; 87(10): 1677 9. 36 37 38 25. U.S. Preventive Services Task Force. Screening for osteoporosis: U.S. preventive services 39 40 task force recommendation statement. Ann Intern Med 2011; 1;154(5):356 64. 41 on September 29, 2021 by guest. Protected copyright. 42 26. Trombetti A, Hars M, Herrmann F, Rizzoli R, Ferrari S. Effect of a multifactorial fall and 43 44 45 fracture risk assessment and management program on gait and balance performances and 46 47 disability in hospitalized older adults: a controlled study. Osteoporosis Int 2013; 24(3):867 48 49 50 76. 51 52 27. Edwards BJ, Song J, Dunlop DD, Fink HA, Cauley JA. Functional decline after incident 53 54 wrist fractures Study of Osteoporotic Fractures: prospective cohort study. BMJ 55 56 57 2010;341:c3324. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 Figure 1: A side-by-side aerial view drawing shows the physical difference between normal feet 37 38 (A) and bound feet (B). Lateral X-rays compared between postmenopausal women of normal 39 40 41 foot (C) and with bound foot (D). 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 2: Heel quantitative ultrasound (QUS) measurement for a women with bound feet. 25 26 27 28 29 30 31 QUS T score QUS Z score 32 3 33 http://bmjopen.bmj.com/ 34 2 BFL Controls BFL Controls 35 0.68 36 1 37 0 38 39 -1 -0.33 40 41 -2 42 -1.96 on September 29, 2021 by guest. Protected copyright. -3 43 -3.00 ** 44 -4 45 46 -5 47 ** 48 49 50 51 Figure 3: QUS calcaneal bone stiffness index-based T- and Z-scores of bound feet ladies (BFL) 52 53 (n=172) compared to the controls (n=82). **:p<0.01. 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 Table 1: Anthropometric and lifestyle data compared between women with or without bound feet 4 5 (Data in mean ± SD) 6 7 Bound feet subjects Control subjects Difference (%) P- value 8 (N=172) (N=82) 9 10 11 Anthropometric data 12 13 Age (years) 74.6±3.5 74.5±4.0 0.20% 0.756 14 15 Weight (kg) For peer46.1±7.4 review48.7±8.9 only-5.3% 0.017* 16 17 BMI (kg/m2) 20.8±3.2 21.9±3.5 -5.1% 0.013* 18 19 20 Length of right foot (cm) 222.7±18.4 254.0±7.3 -12.3% 0.000** 21 22 Width of right foot (cm) 74.3±8.4 99.8±7.0 -25.6% 0.000** 23 24 Lifestyle data 25 26 YSM (years) 26.41±5.21 27.49±5.98 3.93% 0.142 27 28 29 Number of Children 7.02±2.63 6.21±2.13 13.00% 0.007** 30 31 Number subject with 13/172 (7.56%) 6/82 (7.30 %) 0.26% 0.846 32 previous fractures (%) 33 http://bmjopen.bmj.com/ 34 Osteoporosise (%) 165/172 (95.93%) 52/82 (63.42%) 32.51% 0.000** 35 36 (QUS-Stiffness Index) # 37 38 *: p<0.05, **: p<0.01 (unpaired T-test); #: T-score set at -1.8 SD and chi-square test used for 39 statistical analysis. 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 Table 2. Descriptive statistics of 5 scaled categories used to assess physical health in SF-36 4 5 questionnaire (Values given as mean ± SD) 6 Bound Feet Ladies Controls P-value 7 Categories (N= 172) (N= 82) 8 9 Physical 10 Functioning 65.40±22.54 67.53±24.03 0.494 11 Role Physical 45.76±41.24 47.81±40.41 0.712 12 Bodily Pain 64.05±22.53 61.75±21.86 0.448 13 14 General Health 56.65±18.41 54.69±22.40 0.495 15 Vitality For peer61.32±15.01 review64.81±15.44 only 0.090 16 17 PCS 42.9 42.4 18 PCS: physical component summary Score. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 Supplement I 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 Figure 1 Elderly women with bounded feet mainly reside in rural and impoverished 19 provinces of China. 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 Figure 2 A collaborative research project conducted by both clinical scientists and 35 bioengineers from Hong Kong and physicians from Mainland China to investigate the 36 social and physiological impacts of lifelong foot binding on female musculoskeletal 37 38 health in Luliang County of Yunnan Province of Yunnan Province, China. 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 Figure 3 Due to significant physical distance between Luliang County and Hong 27 28 Kong, only few bound feet women with mean QUS values representative of the 29 bound feet group were invited to travel long distance to Hong Kong for assessment of 30 31 bone microstructure, gait and balance using advanced biomedical technologies, i.e. 32 the invited women with bound feet needed to take bus from Luliang County to travel 33 http://bmjopen.bmj.com/ 34 around 200 kilometers to the capital city Kunming of Yunnan Province, China and 35 then took flight to travel over 1200 kilometers to Hong Kong. 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 Supplement II 4 5 6 1. Measurement of volumetric bone mineral density (vBMD) and microarchitecture 7 8 using high-resolution peripheral quantitative computed tomography (HR-pQCT) 9 vBMD and microstructure of the non-dominant distal tibia of the two women with 10 bound feet were evaluated using HR-pQCT (XtremeCT, Scanco Medical AG, 11 12 Switzerland). A standard scanning program was used and the respective region of 13 interest (ROI) of the scanned tibia were automatically separated into cortical and 14 15 trabecularFor compartments peer for calculating review bone mineral density only and microarchitecture 16 of both trabecular bone and cortical bone (Figure 1). Hong Kong female Chinese 17 1 18 population specific T-score at -2.5SD was used for diagnosis of osteoporosis. 19 Analysis: For HR-pQCT, both T and Z scores from distal tibia of the two women with 20 21 bound feet were averaged for comparison with an age-matched reference database. 22 Mean values for each SPS balancing tests parameter was also calculated by averaging 23 results of the two bound feet women that were subsequently compared to age-matched 24 25 controls from population database of HR-pQCT. 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 A B C 38 39 Figure 1 Bone assessment. Heel QUS measurement for a woman with bound feet (A), 40 41 HR- pQCT scanning on lower limb (B), and 3-D HR-pQCT images of distal tibia on September 29, 2021 by guest. Protected copyright. 42 showing inferior bone structural and density in woman with bound feet (C) and age- 43 44 and gender-matched control with normal feet (D). 45 46 47 2. HR-pQCT Findings 48 Mean HR-pQCT values of volumetric BMD and microarchitecture in non-dominant 49 50 distal tibia of two women with bound feet are compared with reference population 51 database to calculate T- and Z-score and their percentage difference (Table 1, Figure 52 2). All parameters show inferior density and structural values in bound foot women, 53 54 with large BMD differences in total trabecular BMD (Dtrab) (42.7%) and inner 55 trabecular BMD (Dinn) (-90.5%) and large bone structural data in total trabecular 56 57 bone volume fraction (tBV/TV) and total trabecular separation (tTb.Sp). Another 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 striking structural difference is seen in the total trabecular number (tTb.N) (36.0%), 4 5 with a much lower value in women with bound feet. 6 7 8 Table 1 Tibia HR-pQCT values obtained for bound feet women (BFW) compared to 9 age and gender matched reference database values for bone geometry and 10 microarchitecture 11 12 Reference Database Difference 13 BFW (Age 60-79) (%) 14 15 Bone geometryFor peer review only 16 Total Area (mm2) 509.35 622.11±145.45 -18.1% 17 2 18 Ct. Area (mm ) 71.65 88.08±32.76 -18.7% 19 Tb. Area (mm2) 426.85 520.55±152.19 -18.0% 20 21 Bone mineral density and microarchitecture 22 D100 (mg HA/cm³) 184.7 230.68±74.05 -19.9% 23 24 Dcomp. (mg HA/cm³) 768.10 782.58±90.89 -18.5% 25 Ct.Th. (mm) 0.815 0.91±0.38 -10.4% 26 27 Ct.Pm. (mm) 88.85 97.66±11.26 -9.0% 28 Dtrab. (mg HA/cm³) 66.85 116.65±42.19 -42.7% 29 Dmeta. (mg HA/cm³) 155.90 194.10±47.43 -19.7% 30 31 Dinn. (mg HA/cm³) 6.05 63.96±43.34 -90.5% 32 tBV/TV 0.06 0.10±0.03 -42.3% 33 http://bmjopen.bmj.com/ 34 tTb.N 0.80 1.25±0.40 -36.0% 35 tTb.Th 0.07 0.08±0.03 -11.9% 36 37 tTb.Sp 1.20 0.77±0.32 56.3% 38 BF: Bound Foot; Ct. Area: Corticular Area; Tb. Area: Trabecular Area; D100: average 39 40 vBMD; Dcomp: cortical density; Ct.Th: cortical thickness; Ct.Pm: cortical 41 perimeter; Dtrab: trabecular density; Dmeta: meta-trabecular density; Dinn: 42 on September 29, 2021 by guest. Protected copyright. 43 inner-trabecular density (calculated from central 50% of the total trabecular bone 44 area); tBV/TV total trabecular bone volume fraction; tTb.N: total trabecular 45 number; tTb.Th: total trabecular thickness; tTb.Sp: total trabecular separation. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 HR-pQCT T score HR-pQCT Z score 6 3 7 8 2 Controls BFL Controls 9 BFL 10 1 11 12 0 13 -0.26 14 -1 15 For peer review-0.88 only 16 -2 17 18 -3 -2.38 19 -3.02 20 -4 21 22 -5 23 24 25 Figure 2 T- and Z-scores of (total or inner) trabecular BMD measured by HR-pQCT 26 compared between the bound feet women (BFW) (mean of the two subjects) 27 28 compared to the controls from the reference database. Osteoporosis (%): 2/2 (100%) 29 in HR-pQCT trabecular BMD with T-scores set at -2.5 SD. 30 31 32 Key Reference 33 http://bmjopen.bmj.com/ 34 1. Tang XL, Qin L, Kwok AW, Zhu TY, Kun EW, Hung VW, Griffith JF, Leung PC, 35 36 Li EK, Tam LS. Alterations of bone geometry, density, microarchitecture, and 37 biomechanical properties in systemic lupus erythematosus on long-term 38 glucocorticoid: a case-control study using HR-pQCT. Osteoporos Int 2013; 39 40 24(6):1817-26. 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 Supplement III 4 5 1. Methods: Gait and balance analysis 6 7 8 Gait analysis: Tekscan Walkway (Tekscan Inc., USA) was used to measure dynamic 9 plantar pressure distributions during walking for test subjects (Figure 1). This 10 2 11 walkway was 3 meters long, consisting of 4 sensors per cm sampling at a rate of 100 12 Hz. Subjects were asked to walk at a comfortable, self-chosen pattern and allowed to 13 familiarize themselves with the procedure and equipment before data collection. Peak 14 15 pressure Forof the whole peer foot and pressure review of two major RIOs: only forefoot and rearfoot were 16 collected during the stance phase of walking. Pressure values were also recorded for 17 1 18 analysis suing published protocol. 19 Balance analysis: SMART EquiTest® (NeuroCom International, Inc., USA) was used 20 21 for assessing static postural stability (SPS) or balancing capabilities of the subjects 22 that were compared with built-in normal database values of Chinese females.2 Briefly, 23 24 a force platform within the system recorded the displacement of center of pressure 25 (CoP) where the subjects were required to complete three stance conditions, i.e. (a) 26 27 bipedal stance with eyes open (BiSEO), (b) bipedal stance with eyes closed (BiSEC) 28 and (c) single-leg stance with eyes open (SLSEO), with bare feet placed on the 29 markings of the force platform. Data were acquired at 100Hz over a period of 20 30 31 seconds in conditions (a) and (b), and 10 seconds in condition (c). Each condition was 32 performed thrice and maximum CoP displacements in antero-posterior (AP) and 33 http://bmjopen.bmj.com/ 34 lateral directions were extracted (Figure 2). In conditions (a) and (b), participants 35 stood at shoulder width; in condition (c), participants performed single-leg stance tests 36 37 on dominant leg. 38 Analysis: Plantar pressure distribution charts were directly compared between the 39 40 right foot of the bound feet lady and that of a weight adjusted normal control of 41 similar age. 42 on September 29, 2021 by guest. Protected copyright.
43 44 45 46 47 48 49 50 51 52 53 A B 54 55 56 Figure 1 Walking gait patterns: Tekscan Walkway System was used to analyze lady 57 with bound feet (A) and a typical example of a three step gait of a woman with bound 58 feet (B). 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 A B 13 14 Figure 2 Static Posture Stability (SPS) assessment using SMART EquiTest® 15 For peer review only 16 (NeuroCom International, Inc., USA) (A) and the displacement of center of pressure 17 (B). 18
19 20 2. Findings 21 22 Gait analysis Figure 3 displays feet force during stance phase. The force during 23 24 propulsion of the stance (gait) phase for control subject is shown to occur at the 25 forefoot while for women with bound feet, the force occurs at the rearfoot while the 26 27 forefoot contributes only a minor force in its total pressure exerted. Foot binding thus 28 shifts the center of gravity towards the heel. Moreover, the rearfoot contact time is 29 found prolonged in bound feet women. For the control subjects, the rearfoot contact 30 31 begins with the heel strike during the beginning of the stance phase (0%) and ends at 32 the heel rise (80% of stance phase). For bound feet women, the heel contact time is 33 http://bmjopen.bmj.com/ 34 prolonged to over 90% of the stance phase. The forefoot contribution to the 35 propulsion is also only seen at the very end of stance phase. The difference in 36 37 accumulated plantar pressure distribution between a normal and a woman with bound 38 foot is demonstrated where the bound feet lady shows a much larger degree of 39 40 pressure placed on the calcaneus of the heel and less force distributed over the 41 forefoot in comparison to a normal foot (Figure 4). 42 on September 29, 2021 by guest. Protected copyright. Balance test Figure 5 summarizes SPS measurements of two women with bound feet 43 44 in comparison to data from the gender- and age-matched control database. Overall, 45 the results indicate no obvious differences in all three categories of anterior-posterior 46 47 and lateral sway compared with between women with bound feet and the controls, 48 including degree of sway in bipedal stance with eyes open (BiSEO), bipedal stance 49 50 with eyes closed (BiSEC), and single-leg stance with eyes open (SLSEO). 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 28 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 Figure 3 Representative foot force compared between subjects with bound feet and 18 gender- and age-matched control, showing that a bound feet lady places almost the 19 20 entirety of her foot force on the rearfoot for the predominant part of the stance phase 21 with prolonged contact time, while the forefoot contributes to the propulsion only at 22 23 the very end of the stance phase (Left). A gender- and age-matched control subject 24 shows a more balanced pressure distribution of the rear and forefoot (Right). 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 Figure 4: Comparison of representative plantar pressure distribution in women with 47 normal feet and bound feet: regions in red represent areas of highest pressure (>231 48 49 kPa), while regions in blue represent areas of lowest pressure (>0 kPa). 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 28 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 3.5 Control Bound feet 5 6 3 7 8 9 2.5 10 11 2 12 13 1.5 14 (cm)Sway 15 For1 peer review only 16 17 0.5 18 19 0 20 BiSEO (cm) BiSEC (cm) SLSEO (cm) BiSEO (cm) BiSEC (cm) SLSEO (cm) 21 22 Anterior-Posterior Sway Lateral Sway 23 24 Figure 5: Static postural stability (SPS) balance test shows no obvious difference in 25 26 both anterior-posterior and lateral sway categories compared for the mean of two 27 women with bound feet and the gender- and age-matched reference population 28 29 database. BiSEO: bipedal stance with eyes open; BiSEC: bipedal stance with eyes 30 closed; and SLSEO: single leg stance with eyes open. 31 32 33 Key Reference: http://bmjopen.bmj.com/ 34 35 1. Mak AF, Zhang M, Boone DA. State-of-the-art research in lower-limb prosthetic 36 biomechanics-socket interface: a review. J Rehabil Res Dev. 2001; 38(2):161–74. 37 38 2. Cheing GL, Chau RM, Kwan RL, Choi CH, Zheng YP. Do the biomechanical 39 properties of the ankle-foot complex influence postural control for people with 40 41 Type 2 diabetes? Clin Biomech 2013; 28(1):88-92. 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
Lifelong Chinese Bound Feet – A Quantitative Ultrasound and Lifestyle Questionnaire Study in Postmenopausal Women
For peer review only Journal: BMJ Open
Manuscript ID: bmjopen-2014-006521.R1
Article Type: Research
Date Submitted by the Author: 28-Jan-2015
Complete List of Authors: Qin, Ling; The Chinese University of Hong Kong, ; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology Pan, Yi; Second Affiliated Hospital of Kunming Medical University, Department of Endocrinology Zhang, Ming; The Hong Kong Polytechnic University, Department of Mechanical Engineering Xu, Mian; Second Affiliated Hospital of Kunming Medical University, Department of Endocrinology Lao, Hanchang; Second Affiliated Hospital of Kunming Medical University, Department of Orthopaedics O’Laughlin, Michael; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology
Tong, Shan; The Hong Kong Polytechnic University, Department of http://bmjopen.bmj.com/ Mechanical Engineering Zhao, Yanling; Health Promotion Foundation of China, Training Department of International Osteoporosis Diagnosis and Treatment Cheng, JCY; The Chinese University of Hong Kong, Department of Orthopaedics and Traumatology Guo, Xia; The Hong Kong Polytechnic University, Department of Rehabilitation Sciences Hung, Vivian; The Chinese University of Hong Kong, Department of Orthopedics and Traumatology on September 29, 2021 by guest. Protected copyright. Primary Subject Sports and exercise medicine Heading:
Secondary Subject Heading: Sports and exercise medicine
ORTHOPAEDIC & TRAUMA SURGERY, Adult orthopaedics < ORTHOPAEDIC Keywords: & TRAUMA SURGERY, Foot & ankle < ORTHOPAEDIC & TRAUMA SURGERY
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 1 1 2 3 1 4 5 6 2 Lifelong Chinese Bound Feet – A Quantitative Ultrasound and Lifestyle Questionnaire 7 8 3 Study in Postmenopausal Women 9 10 11 4 Ling Qin1*PhD; Yi Pan2# MD; Ming Zhang3* PhD; Mian Xu2 MD; Hanchang Lao4 MD; Michael 12 13 1 3 5 1 1 14 5 C. O’Laughlin BSc; Shan Tong MPhil; Yanling Zhao MD; Hung VWY MPhil; Cheng JCY 15 For peer review only 16 6 MD; Xia Guo6* MD 17 18 19 7 1. Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 20 21 22 8 the Chinese University of Hong Kong, Hong Kong SAR, China. 23 24 9 2. Department of Endocrinology, the Second Affiliated Hospital of Kunming Medical 25 26 10 27 University, Kunming, Yunnan Province, China 28 29 11 3. Department of Mechanical Engineering, the Hong Kong Polytechnic University, Hong Kong 30 31 12 SAR, China. 32 33 http://bmjopen.bmj.com/ 34 13 4. Department of Orthopaedics, the Second Affiliated Hospital of Kunming Medical University, 35 36 14 Kunming, Yunnan Province, China 37 38 15 5. Training Department of International Osteoporosis Diagnosis and Treatment, Health 39 40 41 16 Promotion Foundation of China, Beijing China 42 on September 29, 2021 by guest. Protected copyright. 43 17 6. Department of Rehabilitation Sciences, the Hong Kong Polytechnic University, Hong Kong 44 45 18 SAR, China. 46 47 48 49 19 *: Corresponding authors: Ling QIN (PhD) for study design and implementation, e mail: 50 20 [email protected]; Ming ZHANG (PhD) for gait analysis, e mail: [email protected]; 51 52 21 Xia GUO (MD) for balance evaluation, e mail: [email protected] 53 54 55 22 #: Co first author 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 2 1 2 3 1 Abstract 4 5 6 2 Objective The phenomenon of foot binding or also known as "lotus feet" has an enduring and 7 8 9 3 influential history in China. Due to man made smaller foot size, the life long foot binding may 10 11 4 impose adverse effect on skeletons and we investigated the bone properties in postwomen with 12 13 5 bound feet that may provide new information to develop appreciate countermeasures for 14 15 For peer review only 16 6 prevention of fragility fractures. 17 18 7 Design Population based cohort study. 19 20 8 Participants This study involved 254 postmenopausal women aged between 65 and 80, 21 22 23 9 including 172 with bounded feet and 82 age and gender matched control subjects living in a 24 25 10 remote region of China. 26 27 11 28 Outcomes Anthropometric data, SF 36 life style questionnaire, and heel quantitative ultrasound 29 30 12 (QUS) were collected for the entire study population. A small subset of 2 cases was also invited 31 32 13 for assessment of bone mineral and microarchitecture at distal tibiae using high resolution 33 http://bmjopen.bmj.com/ 34 35 14 peripheral quantitative computed tomography (HR pQCT), gait and balance. 36 37 15 Results Women with bound feet had significantly lower QUS value as compared to age matched 38 39 16 women with normal feet, supported by HR pQCT data as well. However, SF 36 questionnaire 40 41 42 17 results did not reveal statistically significant differences in any categorical responses, including on September 29, 2021 by guest. Protected copyright. 43 44 18 physical functioning, general health vitality and physical component summary score, number of 45 46 19 previous fractures. No impaired body balance was found in the small set of group. 47 48 49 20 Conclusions The man made changes of foot binding led to reduced physical activity and 50 51 21 therefore were prone to osteoporosis. Women with bound feet and osteoporosis did not have a 52 53 22 54 higher incidence of fragility fracture when compared to controls. This might be explained by 55 56 23 compensation in physical activity that improves body balance, implying the importance of 57 58 24 improving or maintaining body balance in overall prevention strategy of fragility fractures. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 3 1 2 3 1 4 5 2 Introduction 6 7 8 9 3 Few phenomena in history have been brought to such a high level of esteem and popularity with 10 11 4 a perplexing origin, and even fewer have shared such extensive, enduring impacts. For over a 12 13 5 millennium, Chinese women of all ages and social classes bounded their feet to a fraction of its 14 15 For peer review only 16 6 original size, suffering great pain and hardship during the process. Literature documented that 17 18 7 this tradition began in the closing years of the tenth century by either dancers during the Song 19 20 8 dynasty that valued the aesthetic beauty and graceful movements of small feet or a hedonistic 21 22 23 9 emperor who ordered his concubines to bind their feet to resemble the shape of a “golden lotus” 24 25 10 leaf.1,2 One aspect beyond debate, however, is the widespread influence of foot binding and the 26 27 11 28 longevity of its practice: spanning over 1000 years and four dynasties in China. The tradition 29 th 30 12 reached its apex during the 18 century, where around over 50% of all women and close to 100% 31 32 13 of women in upper class societies bound their feet.2,3 33 http://bmjopen.bmj.com/ 34 35 14 36 Foot binding was first outlawed in 1912 but persisted in various rural regions of China. Women 37 38 15 with small feet were associated with restricted labor and physical activity that promoted 39 40 16 feministic behavior and increased one’s likelihood of marriage.1 3 Nowadays, older women aged 41 42 on September 29, 2021 by guest. Protected copyright. 43 17 65 or above have become the very last generation still living with bound feet. These women 44 45 18 mainly reside in rural and impoverished provinces of China. However, while the phenomenon is 46 47 19 well documented, very few previous reports have collected scientific data to study the direct 48 49 50 20 effect of lifelong foot binding to one’s bone statues and fragility fractures. 51 52 53 21 One of the observable consequences of lifelong foot binding is the reduced foot size and 54 55 22 accordingly significant decrease in support area of the foot that greatly hinders the ability for one 56 57 58 23 to participate in both social and physical activities throughout one’s lifetime (Figure 1). Physical 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 4 1 2 4 7 3 1 exercise has previously been shown with strong evidence of enhancing bone health. The 4 5 2 decreased support given by the smaller foot area would be likely to reduce balancing capabilities, 6 7 3 lead to a greater risk factor of falling and cause higher incidence of fragility fracture due to 8 9 7 9 10 4 restricted physical activity. We therefore put forward our study hypothesis that women with 11 12 5 bound feet had adverse impact on musculoskeletal health due to diminished lifelong tendencies 13 14 6 of physical activity, decrease in ability to balance as a result of smaller foot support area and 15 For peer review only 16 17 7 therefore higher fall frequency and fall related fragility fractures. 18 19 20 8 21 22 23 9 Methods 24 25 26 27 10 Participants 28 29 30 11 This was a project conducted by both clinical scientists and bioengineers from Hong Kong and 31 32 12 physicians from Mainland China to reveal the social and physiological impacts of lifelong foot 33 http://bmjopen.bmj.com/ 34 35 13 binding with participants recruited from Luliang County of Yunan Province, China (Supplement 36 37 14 I). Participants who had any diseases or drug use that were known to affect metabolism of 38 39 15 40 musculoskeletal system as well as recent fracture (within 1 year) were excluded. The study was 41 42 16 supported by the Chinese Health Promotion Foundation (Appendix A) and human ethics on September 29, 2021 by guest. Protected copyright. 43 44 17 approval was obtained from the Second Affiliated Hospital of Kunming Medical Collage for the 45 46 47 18 part of study conducted in Kunming (Appendix B) and the Chinese University of Hong Kong 48 49 19 for the part conducted in Hong Kong (Appendix C). 50 51 52 20 A total of 254 subjects were included into the study. Of the 254 subjects, 172 subjects had bound 53 54 55 21 feet (mean age 74.6±3.5, range 65 80 years) and 82 subjects had normal feet (mean 74.5±4.0, 56 57 22 range 69 87 years). All subjects had their weight and height measured and body mass index 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 1 2 3 1 (BMI) calculated. Previous history of fractures, number of children, and years since menopause 4 5 2 (YSM) was also documented. 6 7 8 3 A modified 36 item short form health survey (SF 36) was given to all participants after 9 10 10 11 4 translation into Chinese that was validated by one of our previous studies. The survey was 12 13 5 organized into eight multi item scales assessing: physical functioning (PF), physical role (PR), 14 15 6 bodily pain (BP),For general health peer (GH), vitality review (VT), social functioning only (SF), emotional role (ER) 16 17 18 7 on daily activities and mental health (MH) that comprehensively determine the quality of life, 19 20 8 physical, mental, and psychosocial health (Appendix D). Of the eight categories, five categories 21 22 9 that govern the physical well being of individuals, including PF, PR, BP, GH and VT, were 23 24 25 10 linearly transformed to a 0 100 scale using a standardized three step algorithm and subsequently 26 27 11 scaled by computing the mean values for each category.11,12 Mean values for each parameter 28 29 12 30 between the two groups were compared to determine differences between bound feet and control 31 32 13 subjects. Finally, scaled scores in each category were combined to produce a higher order 33 http://bmjopen.bmj.com/ 34 14 physical component summary (PCS) score. The PCS score was standardized with the 1998 U.S. 35 36 13 37 15 general population database with a mean score of 50 and S.D. of 10. 38 39 40 16 Heel quantitative ultrasound (QUS) 41 42 on September 29, 2021 by guest. Protected copyright. 43 17 All 254 recruited subjects participated in quantitative ultrasound (QUS) evaluation at heel in 44 45 18 Luliang county using Lunar Achilles Plus (GE Healthcare, Milwaukee, WI, USA) (Figure 2). 46 47 48 19 Speed of sound (SOS) (m/s) and broadband ultrasound attenuation (BUA) (Db/MHz cm) were 49 50 20 recorded. A stiffness index (SI) was derived from SOS and BUA based on following formula: 51 52 21 14 53 SI = (0.67 × BUA) + (0.28 × SOS) – 420. QUS T scores were calculated using built in Chinese 54 55 22 female reference population. Z values were subsequently calculated with the same formula, 56 57 23 expressed as SD above or below the mean SI values of an age, gender and weight matched 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 6 1 2 3 1 reference database. Instead of a conventional T score ≤ 2.5 SD, a T score ≤ 1.8 SD was used as 4 5 2 cutoff value for defining osteoporosis, as previous studies demonstrated that the T score 6 7 3 threshold of 2.5 SD might lead to underestimation of the prevalence of osteoporosis when QUS 8 9 15,16 10 4 was used for bone measurement at the heel while QUS T score threshold of 1.8 SD 11 12 5 identified the same percentage of persons with osteoporosis as the World Health Organization 13 14 6 (WHO) threshold for bone mineral density (BMD) measurements using Dual energy X ray 15 For peer review only 16 16,17 17 7 absorptiometry (DXA). 18 19 20 8 Due to significant physical distance between Luliang County and Hong Kong (Supplement I) 21 22 9 and lacking financial support for international traveling, only two women with bound feet with 23 24 25 10 mean QUS values representative of the bound feet group were invited to Hong Kong for case 26 27 11 analysis using below specified advanced assessments, including 3 D bone densitometry 28 29 12 30 (Supplement II), gait and body balance analysis relevant tests for predicting fall and fragility 31 32 13 fractures (Supplement III). 33 http://bmjopen.bmj.com/ 34 35 14 36 37 38 15 Statistical analysis 39 40 41
16 on September 29, 2021 by guest. Protected copyright. 42 We used Statistical Package for Social Sciences (SPSS for Windows version 19.0, SPSS Inc., 43 44 17 Chicago, IL, USA) for statistical analysis. Quantitative data was expressed as mean±SD for 45 46 18 continuous data and counts by percentage for discrete data. Student’s t test was used to compare 47 48 49 19 parameters between two groups, including age, BMI, size of foot, center of gravity, number of 50 51 20 children, YSM, previous fractures, and rate of T score based osteoporosis. Chi square test was 52 53 21 used to study the association between previous fracture rate and bound feet status and between 54 55 56 22 T score of QUS data and bound feet status. For SF 36 questionnaire data, we sued Student’s t 57 58 23 test to determine if statistical significance existed between categorical mean values of subjects. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 7 1 2 3 1 Chi square test was used for testing statistical difference in rate of osteoporosis and incidence of 4 5 2 low energy fractures between two groups. Values of P < 0.05 were considered statistically 6 7 3 significant. 8 9 10 11 4 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 8 1 2 3 1 Results 4 5 2 Anthropometric data and SF-36 questionnaire 6 7 8 3 Table 1 shows no statistically difference in age between postmenopausal women with and 9 10 11 4 without bound feet, implying a homogenous grouping for comparison. Significant differences are 12 13 5 however seen in body weight, BMI, length and width of feet, number of children, and T score 14 15 6 based rate of osteoporosisFor betweenpeer the two review groups. SF 36 questionnaire only results are summarized 16 17 18 7 in Table 2 where no statistically significant differences are found in any categorical responses 19 20 8 between women with and without bound feet except the vitality score that shows a noticeable but 21 22 9 23 not statistically significant decrease in bound feet subjects. 24 25 26 10 Heel quantitative ultrasound (QUS) 27 28 29 11 Figure 3 shows T and Z scores of heel QUS bone stiffness index. The mean T and Z scores of 30 31 12 women with bound feet are both significant lower than that of the age and gender matched 32 33 http://bmjopen.bmj.com/ 34 13 control subjects (p<0.001 for both) (Figure 3). T score based calculation shows 32.51% 35 36 14 significantly more osteoporotic cases found in women with bound feet (165 out of 172, i.e. 37 38 15 95.93%) compared with age and gender matched control subjects (52 out of 82, i.e. 63.42%) 39 40 41 16 (p<0.01). 42 on September 29, 2021 by guest. Protected copyright. 43 44 17 45 46 47 18 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 9 1 2 3 1 Discussion 4 5 2 Lifelong foot binding is a widespread tradition of ancient China that until this day still affects a 6 7 3 significant number of elderly women living in rural regions. We investigated potential adverse 8 9 10 4 impact of foot binding on musculoskeletal system, including bone mineral and structural statues 11 12 5 along with major fracture risk factors and fracture incidence. Our findings confirmed our 13 14 6 hypothesis that women with bound feet had adverse impact on bone health but disproved our 15 For peer review only 16 17 7 concerns on osteoporosis associated increase in fragility fractures. 18 19 20 8 Due to availability, we only used a portable heel QUS for bone scan in Luliang County, a remote 21 22 9 23 mountain area of Kunming province to quantify bone statues of 172 postmenopausal women 24 25 10 with bound feet and their peers. However, calcaneus bone is a skeletal site that is directly 26 27 11 affected by foot binding and associated alterations in physical activities. Our QUS measurement 28 29 30 12 showed 1/3 higher osteoporotic rate in bound feet women than that of the controls. HR pQCT 31 32 13 used for a small subset evaluation for the current study is an advanced technology that allows 33 http://bmjopen.bmj.com/ 34 14 insight into volumetric BMD (vBMD) and microarchitecture. Trabecular microarchitecture 35 36 37 15 parameters of the representative subjects indicated that bone quality was inferior in bound feet 38 39 16 women with on average 20% less trabecular number and 42.7% of its vBMD in core region of 40 41
17 the distal tibia although the cortical vBMD (Dcomp) was much less affected with only 18.1% on September 29, 2021 by guest. Protected copyright. 42 43 44 18 lower mean value as compared with the age and gender matched reference population database 45 46 19 (Supplement II), implying systemic adverse impact of foot binding on weight bearing bones. If 47 48 20 49 such difference would be independent of weight bearing, measurements at non weight bearing 50 51 21 skeletal sites such as at distal radius of a sizable study subgroup of future studies might provide 52 53 22 relevant evidences to support this hypothesis. 54 55 56 23 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 10 1 2 3 1 Previous QUS studies in normal postmenopausal women demonstrated the value of QUS for 4 5 2 evaluation of osteoporosis and predicting fragility fractures.17 19 As fragility fractures were 6 7 3 associated with both skeletal and non skeletal factors, we adopted 36 questionnaires, a validated 8 9 10 4 instrument for evaluation of relationship between physical activity and osteoporosis and/or 11 12 5 fragility fractures into the current study.13, 21 Impaired body balance has been a known risk factor 13 14 6 of fall associated fragility fractures in women with normal feet. 8,22,23 However, our analysis only 15 For peer review only 16 17 7 showed non detectable inferior vitality and physical functioning in women with bound feet and 18 19 8 the differences in accumulated plantar pressure with more pressure placed on the calcaneus of 20 21 9 22 the heel and less force distributed over the forefoot inbound feet (Supplement II), implying an 23 24 10 compensation mechanism in human kinesiology after foot binding in later life. Our data did not 25 26 11 show the difference in number of previous fractures in women with or without foot binding. 27 28 29 12 30 An early as in 1997, Cummings and co workers reported a first Dual energy X ray 31 32 13 Absorptiometry (DXA) study to measure BMD at spine and hip in post menopausal women with 33 http://bmjopen.bmj.com/ 34 14 bound feet in Beijing, the capital city of China, and reported lifelong foot binding had significant 35 36 37 15 adverse effects on BMD at spine BMD and higher fragility fracture rate when compared to the 38 39 16 control subjects.24 The direct adverse impact to the bound feet is the significant reduced foot size 40 41 17 which is described as “golden lotus”. The associated consequence is an increased tendency of on September 29, 2021 by guest. Protected copyright. 42 43 44 18 falling and increased risk of fall induced fractures in later life. The overall higher fracture rate in 45 46 19 Beijing study was 15% while significant lower fracture incidence with 7.5% was found in our 47 48 20 current subjects that might be explained by different lifestyles between women living in urban 49 50 24 51 21 areas in Beijing of Cumming’s study and rural regions of our current one conducted in Luliang, 52 53 22 a remote county in Yunnan province. There were no balance and gait data available in Beijing’s 54 55 23 56 study. The bound feet subjects in Beijing were mostly descendants of the upper class families 57 58 24 and little physical work was required in their later life, with speculated impairment in 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 11 1 2 3 1 musculoskeletal coordination to explain both higher osteoporosis and fall associated fragility 4 5 2 fracture rate. 1,3,24,25 As our study subjects in remote Luliang county were working class women 6 7 3 and often had to perform mandatory labor work in old age, this might explain that the balance 8 9 10 4 analysis and life style questionnaire did not show difference between women with and without 11 12 5 bound feet evaluated in a small subset of subjects (Supplement III). The associated active 13 14 6 lifestyle might have led to different adaptations of the body balance throughout one’s lifetime to 15 For peer review only 16 17 7 compensate smaller foot support area. 18 19 20 8 The findings of our study on heel QUS and lifestyle in terms of SF 36 questionnaire offer a 21 22 9 unique look into the effects of foot binding. The significance of our current study in social and 23 24 25 10 health care aspects was the mobilization of the local government to support our project. Our 26 27 11 findings generated from the current study raised the awareness of osteoporosis prevention 28 29 12 30 measures in rural regions of China at one site and at another. Apart from difference in 31 26,27 32 13 nutrition, the importance of physical activities and/or exercise that are specifically relevant 33 http://bmjopen.bmj.com/ 34 14 for balance maintenance as this has been a known and crucial factor for prevention of fall and 35 36 7,10,28,29 37 15 fall induced fragility fractures. 38 39 40 16 Our current study had a few limitations apart from the cross sectional study in nature: 1) 41 42 17 Inhomogeneous number in study subjects: the higher percentage of bound feet subjects was due on September 29, 2021 by guest. Protected copyright. 43 44 18 45 to the greater percentage of elderly women with bound feet living in the location of study; 2) due 46 47 19 to limited budget and long distance between remote Luliang country and Hong Kong, only few 48 49 20 subjects were invited to fly to Hong Kong for HR pQCT, gait and balance tests (Supplements II 50 51 52 21 and III); 3) As only fracture history could be recalled for analysis, no reliable fall frequency 53 54 22 could be provided for analysis. 55 56 57 23 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 12 1 2 3 1 Conclusion 4 5 6 2 Foot binding had life long negative impact on bone health. However, the neuromusculoskeletal 7 8 9 3 coordination of the bound feet postmenopausal women with active life style might be 10 11 4 compensated as no difference was revealed in incidence of fragility fractures, implying the 12 13 5 importance of body balance maintenance in prevention of fragility fractures. 14 15 For peer review only 16 17 6 18 19 7 Funding This study was supported by the China Health Promotion Foundation (Ref. 2009 01) 20 21 8 and Contemporary Orthopaedic Research and Education Fund of the Chinese University of Hong 22 23 24 9 Kong (CUHK#6903361). We also appreciate technical support from Miss Fong TN and Miss 25 26 10 Hung WY of Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 27 28 11 The Chinese University of Hong Kong for their assistance in HR pQCT measurements. 29 30 31 32 12 33 http://bmjopen.bmj.com/ 34 13 Competing interests: The authors declare no competing interests. 35 36 14 37 38 39 15 Contribution statement: Ling Qin designed and organized this project, participated in data 40 41 16 collection, analysis and interpretation, and guided the preparation of this Manuscript; Yi Pan, 42 on September 29, 2021 by guest. Protected copyright. 43 17 Mian Xu, and Hanchang Lao recruited subjects, performed questionnaires and conducted QUS 44 45 46 18 scanning at Luliang county; Ming Zhang and Shan Cong performed gait analysis, and Xia Guo 47 48 19 performed balance test and related analysis in Hong Kong. Yanling Zhao co organized the study 49 50 20 51 and jointly applied for research grant for this project; Michael O’Laughlin did a part of the data 52 53 21 analysis and assisted in drafting the manuscript. WY Hong did the HR pQCT measurement and 54 55 22 its data analysis; JCY Cheng contributed to planning and preparation of the manuscript. 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 13 1 2 3 1 Data sharing: Authors agree with the data sharing as specified in the policy of BMJ Open to 4 5 2 make all submitted materials, including supplementary files available to the public. 6 7 8 3 Table 1: Anthropometric and lifestyle data compared between women with or without bound feet 9 10 4 (Data in mean ± SD) 11 12 Bound feet subjects Control subjects Difference (%) P- value 13 (N=172) (N=82) 14 Anthropometric data 15 For peer review only 16 17 Age (years) 74.6±3.5 74.5±4.0 0.20% 0.756 18 Weight (kg) 46.1±7.4 48.7±8.9 5.3% 0.017* 19 BMI (kg/m2) 20.8±3.2 21.9±3.5 5.1% 0.013* 20 Length of right foot (cm) 222.7±18.4 254.0±7.3 12.3% 0.000** 21 22 Width of right foot (cm) 74.3±8.4 99.8±7.0 25.6% 0.000** 23 Lifestyle data 24 25 YSM (years) 26.41±5.21 27.49±5.98 3.93% 0.142 26 Number of Children 7.02±2.63 6.21±2.13 13.00% 0.007** 27 Number subject with 13/172 (7.56%) 6/82 (7.30 %) 0.26%@ 0.846 28 previous fractures (%) 29 30 Osteoporosis (%) (QUS 165/172 (95.93%) 52/82 (63.42%) 32.51%@@ 0.000** 31 Stiffness Index) # 32 Odds ratio 33 5 *: p<0.05, **: p<0.01 (unpaired T test); #: T score set at 1.8 SD and chi square test used for http://bmjopen.bmj.com/ 34 35 6 36 statistical analysis; Odds ratio (OR): @, OR=1.036 (p>0.05); @@, OR=13,60 (p<0.01). 37 38 39 7 References 40 41 42 8 1. Levy HS. Chinese Footbinding: The history of a curious erotic custom. New York, NY: W. on September 29, 2021 by guest. Protected copyright. 43 44 9 Rawls, 1966 45 46 47 10 2. Wang P. Aching for beauty: Foot binding in China. Minneapolis: University of Minnesota 48 49 11 Press, 2000 50 51 12 3. Fairbank JK. The Great Chinese Revolution: 1800 1985 New York, Harper & Row, 1986 52 53 54 13 4. Gerdhem P, Akesson K, Obrant KJ. Effect of previous and present physical activity on bone 55 56 14 mass in elderly women. Osteoporosis Int 2003; 14:208–12. 57 58 15 59 5. Cooper C, Harvey NC. Osteoporosis risk assessment. BMJ 2012; 344:e3427. 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 14 1 2 3 1 6. Qin L, Au SZ, Choy YW, Leung PC, Neff M, KM Lee, MC Lau, Woo J, Chan KM. Regular 4 5 2 Tai Chi exercise may retard bone loss in postmenopausal women – A case control study. 6 7 3 Arch Phys Med Reha 2002; 83(10):1355 9. 8 9 10 4 7. Qin L, Choy WY, Leung KS, Leung PC, Au SK, Hung WY, Dambacher MA, Chan KM. 11 12 5 Beneficial effects of regular Tai Chi exercise on musculoskeletal system. J Bone Miner 13 14 6 Metabol 2005; 23(2):186 90. 15 For peer review only 16 17 7 8. Edwards MH, Jameson K, Denison H, Harvey NC, Sayer AA, Dennison EM, Cooper C. 18 19 8 Clinical risk factors, bone density and fall history in the prediction of incident fracture among 20 21 9 22 men and women. Bone 2013; 52(2):541 7. 23 24 10 9. Hippisley Cox J, Coupland C. Derivation and validation of updated QFracture algorithm to 25 26 11 predict risk of osteoporotic fracture in primary care in the United Kingdom: prospective open 27 28 29 12 cohort study. BMJ 2012; 344:e4191; 1 16. 30 31 13 10. Chan KM, Qin L, Lau M, Woo J, Au SK, Choy WY, Lee KM, Lee SH. A randomized, 32 33 14 prospective study of the effects of Tai Chi Chun exercise on bone mineral density in http://bmjopen.bmj.com/ 34 35 36 15 postmenopausal women. Arch Phys Med Rehab 2004; 85:717 22. 37 38 16 11. Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF 36). I. Conceptual 39 40 17 framework and item selection. Med Care 1992; 30:473–83. 41 42 on September 29, 2021 by guest. Protected copyright. 43 18 12. Ware JE, Kosinski M, Bayliss MS, McHormey, Rogers WH, Raczek A. Comparison of 44 45 19 methods for the scoring and statistical analysis of SF 36 health profile and summary 46 47 20 48 measures: summary of results from the Medical Outcomes Study. Med Care 1995; 33(4): 49 50 21 264 79. 51 52 22 13. Maglinte GA, Hays RD, Kaplan RM. US general population norms for telephone 53 54 55 23 administration of the SF 36v2. J Clin Epidemiol 2012; 65(5): 497–502. 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 15 1 2 3 1 14. Economos CD, Sacheck JM, Wacker W, Shea K, Naumova EN Precision of Lunar Achilles+ 4 5 2 bone quality measurements: time dependency and multiple machine use in field studies. Br J 6 7 3 Radiol 2007; 80:919 25. 8 9 10 4 15. Vu TT, Nguyen CK, Nguyen TL, Le BM, Nguyen Trung Le Dm Bui TN, Bakamori M, 11 12 5 Kunii D, Sakai T, Yamamoto S. Determining the prevalence of osteoporosis and related 13 14 6 factors using quantitative ultrasound in Vietnamese adult women. Am J Epidemiol 2005; 15 For peer review only 16 17 7 161(9):824 30. 18 19 8 16. Frost ML, Blake GM, Fogelman I. Can the WHO criteria for diagnosing osteoporosis be 20 21 9 22 applied to calcaneal quantitative ultrasound? Osteoporos Int 2000; 11:321–30. 23 24 10 17. Liu JM, Ma LY, Bi YF, Huang Y, Xu M, Zhao HY, Sun LH, Tao B, Li XY, Wang WQ, 25 26 11 Ning G. A population based study examining Calcaneus quantitative ultrasound and its 27 28 29 12 optimal cut points to discriminate osteoporotic fractures among 9352 Chinese women and 30 31 13 men. J Clin Endocrinol Metab 2012: 97(3):800 9. 32 33 14 18. Hans D, Dargent Molina P, Schott AM, Sebert JL, Cormier C, Kotzki PO, Delmas PD, http://bmjopen.bmj.com/ 34 35 36 15 Pouilles JM, Breart G, Meunier PJ. Ultrasonographic heel measurements to predict hip 37 38 16 fracture in elderly women: the EPIDOS prospective study. Lancet 1996; 348:511–14. 39 40 17 19. Glueer CC. Quantitative ultrasound techniques for the assessment of osteoporosis: expert 41 42 on September 29, 2021 by guest. Protected copyright. 43 18 agreement on current status. J Bone Miner Res 1997; 12:1280–8. 44 45 19 20. Nayak S, Olkin I, Liu H, Grabe M, Gould MK, Allen IE, Owens DK, Bravata DM. Meta 46 47 20 48 analysis: accuracy of quantitative ultrasound for identifying patients with osteoporosis. Ann 49 50 21 Intern Med 2006; 6;144(11):832 41. 51 52 22 21. Lai BM, Tsang SW, Lam CL, Kung AW. Validation of the Quality of Life Questionnaire of 53 54 55 23 the European Foundation for Osteoporosis (QUALEFFO 31) in Chinese. Clin Rheumatol 56 57 24 2010; 29(9):965 72. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 16 1 2 3 1 22. Kanis JA. Diagnosis of osteoporosis and assessment of fracture risk. Lancet 2002; 4 5 2 359(9321):1929 36. 6 7 3 8 23. Kern LM, Powe NR, Levine MA, Fitzpatrick AL, Harris TB, Robbins J, Fried LP. 9 10 4 Association between screening for osteoporosis and the incidence of hip fracture. Ann Intern 11 12 5 Med 2005; 142(3):173 81. 13 14 6 15 24. Cummings S,For Ling X, Stonepeer K. Consequences review of foot binding only among older women in Beijing, 16 17 7 China. Am J Pub Health 1997; 87(10): 1677 9. 18 19 8 25. U.S. Preventive Services Task Force. Screening for osteoporosis: U.S. preventive services 20 21 22 9 task force recommendation statement. Ann Intern Med 2011; 1;154(5):356 64. 23 24 10 26. Grigg M, Arora M, Diwan A. Role of nutritional supplementation in elderly patients with hip 25 26 11 fractures. J Orthopaedic Translation 2014; 2(1): 26 34. 27 28 29 12 27. Qin L, Choy W, Hung V, Au S, Chan K, Leung K. Age related vessel calcification at distal 30 31 13 extremities is a risk factor of osteoporosis. J Orthopaedic Translation 2014; 2(1): 43 48. 32 33 14 http://bmjopen.bmj.com/ 34 28. Trombetti A, Hars M, Herrmann F, Rizzoli R, Ferrari S. Effect of a multifactorial fall and 35 36 15 fracture risk assessment and management program on gait and balance performances and 37 38 16 disability in hospitalized older adults: a controlled study. Osteoporosis Int 2013; 24(3):867 39 40 41 17 76. 42 on September 29, 2021 by guest. Protected copyright. 43 18 29. Edwards BJ, Song J, Dunlop DD, Fink HA, Cauley JA. Functional decline after incident 44 45 19 wrist fractures Study of Osteoporotic Fractures: prospective cohort study. BMJ 2010; 46 47 48 20 341:c3324. 49 50 21 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 1 1 2 3 1 4 5 6 2 Lifelong Chinese Bound Feet – A Quantitative Ultrasound and Lifestyle Questionnaire 7 8 3 Study in Postmenopausal Women 9 10 11 4 Ling Qin1*PhD; Yi Pan2# MD; Ming Zhang3* PhD; Mian Xu2 MD; Hanchang Lao4 MD; Michael 12 13 1 3 5 1 1 14 5 C. O’Laughlin BSc; Shan Tong MPhil; Yanling Zhao MD; Hung VWY MPhil; Cheng JCY 15 For peer review only 16 6 MD; Xia Guo6* MD 17 18 19 7 1. Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 20 21 22 8 the Chinese University of Hong Kong, Hong Kong SAR, China. 23 24 9 2. Department of Endocrinology, the Second Affiliated Hospital of Kunming Medical 25 26 10 27 University, Kunming, Yunnan Province, China 28 29 11 3. Department of Mechanical Engineering, the Hong Kong Polytechnic University, Hong Kong 30 31 12 SAR, China. 32 33 http://bmjopen.bmj.com/ 34 13 4. Department of Orthopaedics, the Second Affiliated Hospital of Kunming Medical University, 35 36 14 Kunming, Yunnan Province, China 37 38 15 5. Training Department of International Osteoporosis Diagnosis and Treatment, Health 39 40 41 16 Promotion Foundation of China, Beijing China 42 on September 29, 2021 by guest. Protected copyright. 43 17 6. Department of Rehabilitation Sciences, the Hong Kong Polytechnic University, Hong Kong 44 45 18 SAR, China. 46 47 48 49 19 *: Corresponding authors: Ling QIN (PhD) for study design and implementation, e mail: 50 20 [email protected]; Ming ZHANG (PhD) for gait analysis, e mail: [email protected]; 51 52 21 Xia GUO (MD) for balance evaluation, e mail: [email protected] 53 54 55 22 #: Co first author 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 2 1 2 3 1 Abstract 4 5 6 2 Objective The phenomenon of foot binding or also known as "lotus feet" has an enduring and 7 8 9 3 influential history in China. Due to man made smaller foot size, the life long foot binding may 10 11 4 impose adverse effect on skeletons and we investigated the bone properties in postwomen with 12 13 5 bound feet that may provide new information to develop appreciate countermeasures for 14 15 For peer review only 16 6 prevention of fragility fractures. 17 18 7 Design Population based cohort study. 19 20 8 Participants This study involved 254 postmenopausal women aged between 65 and 80, 21 22 23 9 including 172 with bounded feet and 82 age and gender matched control subjects living in a 24 25 10 remote region of China. 26 27 11 28 Outcomes Anthropometric data, SF 36 life style questionnaire, and heel quantitative ultrasound 29 30 12 (QUS) were collected for the entire study population. A small subset of 2 cases was also invited 31 32 13 for assessment of bone mineral and microarchitecture at distal tibiae using high resolution 33 http://bmjopen.bmj.com/ 34 35 14 peripheral quantitative computed tomography (HR pQCT), gait and balance. 36 37 15 Results Women with bound feet had significantly lower QUS value as compared to age matched 38 39 16 women with normal feet, supported by HR pQCT data as well. However, SF 36 questionnaire 40 41 42 17 results did not reveal statistically significant differences in any categorical responses, including on September 29, 2021 by guest. Protected copyright. 43 44 18 physical functioning, general health vitality and physical component summary score, number of 45 46 19 previous fractures. No impaired body balance was found in the small set of group. 47 48 49 20 Conclusions The man made changes of foot binding led to reduced physical activity and 50 51 21 therefore were prone to osteoporosis. Women with bound feet and osteoporosis did not have a 52 53 22 54 higher incidence of fragility fracture when compared to controls. This might be explained by 55 56 23 compensation in physical activity that improves body balance, implying the importance of 57 58 24 improving or maintaining body balance in overall prevention strategy of fragility fractures. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 3 1 2 3 1 4 5 2 Introduction 6 7 8 9 3 Few phenomena in history have been brought to such a high level of esteem and popularity with 10 11 4 a perplexing origin, and even fewer have shared such extensive, enduring impacts. For over a 12 13 5 millennium, Chinese women of all ages and social classes bounded their feet to a fraction of its 14 15 For peer review only 16 6 original size, suffering great pain and hardship during the process. Literature documented that 17 18 7 this tradition began in the closing years of the tenth century by either dancers during the Song 19 20 8 dynasty that valued the aesthetic beauty and graceful movements of small feet or a hedonistic 21 22 23 9 emperor who ordered his concubines to bind their feet to resemble the shape of a “golden lotus” 24 25 10 leaf.1,2 One aspect beyond debate, however, is the widespread influence of foot binding and the 26 27 11 28 longevity of its practice: spanning over 1000 years and four dynasties in China. The tradition 29 th 30 12 reached its apex during the 18 century, where around over 50% of all women and close to 100% 31 32 13 of women in upper class societies bound their feet.2,3 33 http://bmjopen.bmj.com/ 34 35 14 36 Foot binding was first outlawed in 1912 but persisted in various rural regions of China. Women 37 38 15 with small feet were associated with restricted labor and physical activity that promoted 39 40 16 feministic behavior and increased one’s likelihood of marriage.1 3 Nowadays, older women aged 41 42 on September 29, 2021 by guest. Protected copyright. 43 17 65 or above have become the very last generation still living with bound feet. These women 44 45 18 mainly reside in rural and impoverished provinces of China. However, while the phenomenon is 46 47 19 well documented, very few previous reports have collected scientific data to study the direct 48 49 50 20 effect of lifelong foot binding to one’s bone statues and fragility fractures. 51 52 53 21 One of the observable consequences of lifelong foot binding is the reduced foot size and 54 55 22 accordingly significant decrease in support area of the foot that greatly hinders the ability for one 56 57 58 23 to participate in both social and physical activities throughout one’s lifetime (Figure 1). Physical 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 4 1 2 4 7 3 1 exercise has previously been shown with strong evidence of enhancing bone health. The 4 5 2 decreased support given by the smaller foot area would be likely to reduce balancing capabilities, 6 7 3 lead to a greater risk factor of falling and cause higher incidence of fragility fracture due to 8 9 7 9 10 4 restricted physical activity. We therefore put forward our study hypothesis that women with 11 12 5 bound feet had adverse impact on musculoskeletal health due to diminished lifelong tendencies 13 14 6 of physical activity, decrease in ability to balance as a result of smaller foot support area and 15 For peer review only 16 17 7 therefore higher fall frequency and fall related fragility fractures. 18 19 20 8 21 22 23 9 Methods 24 25 26 27 10 Participants 28 29 30 11 This was a project conducted by both clinical scientists and bioengineers from Hong Kong and 31 32 12 physicians from Mainland China to reveal the social and physiological impacts of lifelong foot 33 http://bmjopen.bmj.com/ 34 35 13 binding with participants recruited from Luliang County of Yunan Province, China (Supplement 36 37 14 I). Participants who had any diseases or drug use that were known to affect metabolism of 38 39 15 40 musculoskeletal system as well as recent fracture (within 1 year) were excluded. The study was 41 42 16 supported by the Chinese Health Promotion Foundation (Appendix A) and human ethics on September 29, 2021 by guest. Protected copyright. 43 44 17 approval was obtained from the Second Affiliated Hospital of Kunming Medical Collage for the 45 46 47 18 part of study conducted in Kunming (Appendix B) and the Chinese University of Hong Kong 48 49 19 for the part conducted in Hong Kong (Appendix C). 50 51 52 20 A total of 254 subjects were included into the study. Of the 254 subjects, 172 subjects had bound 53 54 55 21 feet (mean age 74.6±3.5, range 65 80 years) and 82 subjects had normal feet (mean 74.5±4.0, 56 57 22 range 69 87 years). All subjects had their weight and height measured and body mass index 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 1 2 3 1 (BMI) calculated. Previous history of fractures, number of children, and years since menopause 4 5 2 (YSM) was also documented. 6 7 8 3 A modified 36 item short form health survey (SF 36) was given to all participants after 9 10 10 11 4 translation into Chinese that was validated by one of our previous studies. The survey was 12 13 5 organized into eight multi item scales assessing: physical functioning (PF), physical role (PR), 14 15 6 bodily pain (BP),For general health peer (GH), vitality review (VT), social functioning only (SF), emotional role (ER) 16 17 18 7 on daily activities and mental health (MH) that comprehensively determine the quality of life, 19 20 8 physical, mental, and psychosocial health (Appendix D). Of the eight categories, five categories 21 22 9 that govern the physical well being of individuals, including PF, PR, BP, GH and VT, were 23 24 25 10 linearly transformed to a 0 100 scale using a standardized three step algorithm and subsequently 26 27 11 scaled by computing the mean values for each category.11,12 Mean values for each parameter 28 29 12 30 between the two groups were compared to determine differences between bound feet and control 31 32 13 subjects. Finally, scaled scores in each category were combined to produce a higher order 33 http://bmjopen.bmj.com/ 34 14 physical component summary (PCS) score. The PCS score was standardized with the 1998 U.S. 35 36 13 37 15 general population database with a mean score of 50 and S.D. of 10. 38 39 40 16 Heel quantitative ultrasound (QUS) 41 42 on September 29, 2021 by guest. Protected copyright. 43 17 All 254 recruited subjects participated in quantitative ultrasound (QUS) evaluation at heel in 44 45 18 Luliang county using Lunar Achilles Plus (GE Healthcare, Milwaukee, WI, USA) (Figure 2). 46 47 48 19 Speed of sound (SOS) (m/s) and broadband ultrasound attenuation (BUA) (Db/MHz cm) were 49 50 20 recorded. A stiffness index (SI) was derived from SOS and BUA based on following formula: 51 52 21 14 53 SI = (0.67 × BUA) + (0.28 × SOS) – 420. QUS T scores were calculated using built in Chinese 54 55 22 female reference population. Z values were subsequently calculated with the same formula, 56 57 23 expressed as SD above or below the mean SI values of an age, gender and weight matched 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 6 1 2 3 1 reference database. Instead of a conventional T score ≤ 2.5 SD, a T score ≤ 1.8 SD was used as 4 5 2 cutoff value for defining osteoporosis, as previous studies demonstrated that the T score 6 7 3 threshold of 2.5 SD might lead to underestimation of the prevalence of osteoporosis when QUS 8 9 15,16 10 4 was used for bone measurement at the heel while QUS T score threshold of 1.8 SD 11 12 5 identified the same percentage of persons with osteoporosis as the World Health Organization 13 14 6 (WHO) threshold for bone mineral density (BMD) measurements using Dual energy X ray 15 For peer review only 16 16,17 17 7 absorptiometry (DXA). 18 19 20 8 Due to significant physical distance between Luliang County and Hong Kong (Supplement I) 21 22 9 and lacking financial support for international traveling, only two women with bound feet with 23 24 25 10 mean QUS values representative of the bound feet group were invited to Hong Kong for case 26 27 11 analysis using below specified advanced assessments, including 3 D bone densitometry 28 29 12 30 (Supplement II), gait and body balance analysis relevant tests for predicting fall and fragility 31 32 13 fractures (Supplement III). 33 http://bmjopen.bmj.com/ 34 35 14 36 37 38 15 Statistical analysis 39 40 41
16 on September 29, 2021 by guest. Protected copyright. 42 We used Statistical Package for Social Sciences (SPSS for Windows version 19.0, SPSS Inc., 43 44 17 Chicago, IL, USA) for statistical analysis. Quantitative data was expressed as mean±SD for 45 46 18 continuous data and counts by percentage for discrete data. Student’s t test was used to compare 47 48 49 19 parameters between two groups, including age, BMI, size of foot, center of gravity, number of 50 51 20 children, YSM, previous fractures, and rate of T score based osteoporosis. Chi square test was 52 53 21 used to study the association between previous fracture rate and bound feet status and between 54 55 56 22 T score of QUS data and bound feet status. For SF 36 questionnaire data, we sued Student’s t 57 58 23 test to determine if statistical significance existed between categorical mean values of subjects. 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 7 1 2 3 1 Chi square test was used for testing statistical difference in rate of osteoporosis and incidence of 4 5 2 low energy fractures between two groups. Values of P < 0.05 were considered statistically 6 7 3 significant. 8 9 10 11 4 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 8 1 2 3 1 Results 4 5 2 Anthropometric data and SF-36 questionnaire 6 7 8 3 Table 1 shows no statistically difference in age between postmenopausal women with and 9 10 11 4 without bound feet, implying a homogenous grouping for comparison. Significant differences are 12 13 5 however seen in body weight, BMI, length and width of feet, number of children, and T score 14 15 6 based rate of osteoporosisFor betweenpeer the two review groups. SF 36 questionnaire only results are summarized 16 17 18 7 in Table 2 where no statistically significant differences are found in any categorical responses 19 20 8 between women with and without bound feet except the vitality score that shows a noticeable but 21 22 9 23 not statistically significant decrease in bound feet subjects. 24 25 26 10 Heel quantitative ultrasound (QUS) 27 28 29 11 Figure 3 shows T and Z scores of heel QUS bone stiffness index. The mean T and Z scores of 30 31 12 women with bound feet are both significant lower than that of the age and gender matched 32 33 http://bmjopen.bmj.com/ 34 13 control subjects (p<0.001 for both) (Figure 3). T score based calculation shows 32.51% 35 36 14 significantly more osteoporotic cases found in women with bound feet (165 out of 172, i.e. 37 38 15 95.93%) compared with age and gender matched control subjects (52 out of 82, i.e. 63.42%) 39 40 41 16 (p<0.01). 42 on September 29, 2021 by guest. Protected copyright. 43 44 17 45 46 47 18 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 9 1 2 3 1 Discussion 4 5 2 Lifelong foot binding is a widespread tradition of ancient China that until this day still affects a 6 7 3 significant number of elderly women living in rural regions. We investigated potential adverse 8 9 10 4 impact of foot binding on musculoskeletal system, including bone mineral and structural statues 11 12 5 along with major fracture risk factors and fracture incidence. Our findings confirmed our 13 14 6 hypothesis that women with bound feet had adverse impact on bone health but disproved our 15 For peer review only 16 17 7 concerns on osteoporosis associated increase in fragility fractures. 18 19 20 8 Due to availability, we only used a portable heel QUS for bone scan in Luliang County, a remote 21 22 9 23 mountain area of Kunming province to quantify bone statues of 172 postmenopausal women 24 25 10 with bound feet and their peers. However, calcaneus bone is a skeletal site that is directly 26 27 11 affected by foot binding and associated alterations in physical activities. Our QUS measurement 28 29 30 12 showed 1/3 higher osteoporotic rate in bound feet women than that of the controls. HR pQCT 31 32 13 used for a small subset evaluation for the current study is an advanced technology that allows 33 http://bmjopen.bmj.com/ 34 14 insight into volumetric BMD (vBMD) and microarchitecture. Trabecular microarchitecture 35 36 37 15 parameters of the representative subjects indicated that bone quality was inferior in bound feet 38 39 16 women with on average 20% less trabecular number and 42.7% of its vBMD in core region of 40 41
17 the distal tibia although the cortical vBMD (Dcomp) was much less affected with only 18.1% on September 29, 2021 by guest. Protected copyright. 42 43 44 18 lower mean value as compared with the age and gender matched reference population database 45 46 19 (Supplement II), implying systemic adverse impact of foot binding on weight bearing bones. If 47 48 20 49 such difference would be independent of weight bearing, measurements at non weight bearing 50 51 21 skeletal sites such as at distal radius of a sizable study subgroup of future studies might provide 52 53 22 relevant evidences to support this hypothesis. 54 55 56 23 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 10 1 2 3 1 Previous QUS studies in normal postmenopausal women demonstrated the value of QUS for 4 5 2 evaluation of osteoporosis and predicting fragility fractures.17 19 As fragility fractures were 6 7 3 associated with both skeletal and non skeletal factors, we adopted 36 questionnaires, a validated 8 9 10 4 instrument for evaluation of relationship between physical activity and osteoporosis and/or 11 12 5 fragility fractures into the current study.13, 21 Impaired body balance has been a known risk factor 13 14 6 of fall associated fragility fractures in women with normal feet. 8,22,23 However, our analysis only 15 For peer review only 16 17 7 showed non detectable inferior vitality and physical functioning in women with bound feet and 18 19 8 the differences in accumulated plantar pressure with more pressure placed on the calcaneus of 20 21 9 22 the heel and less force distributed over the forefoot inbound feet (Supplement II), implying an 23 24 10 compensation mechanism in human kinesiology after foot binding in later life. Our data did not 25 26 11 show the difference in number of previous fractures in women with or without foot binding. 27 28 29 12 30 An early as in 1997, Cummings and co workers reported a first Dual energy X ray 31 32 13 Absorptiometry (DXA) study to measure BMD at spine and hip in post menopausal women with 33 http://bmjopen.bmj.com/ 34 14 bound feet in Beijing, the capital city of China, and reported lifelong foot binding had significant 35 36 37 15 adverse effects on BMD at spine BMD and higher fragility fracture rate when compared to the 38 39 16 control subjects.24 The direct adverse impact to the bound feet is the significant reduced foot size 40 41 17 which is described as “golden lotus”. The associated consequence is an increased tendency of on September 29, 2021 by guest. Protected copyright. 42 43 44 18 falling and increased risk of fall induced fractures in later life. The overall higher fracture rate in 45 46 19 Beijing study was 15% while significant lower fracture incidence with 7.5% was found in our 47 48 20 current subjects that might be explained by different lifestyles between women living in urban 49 50 24 51 21 areas in Beijing of Cumming’s study and rural regions of our current one conducted in Luliang, 52 53 22 a remote county in Yunnan province. There were no balance and gait data available in Beijing’s 54 55 23 56 study. The bound feet subjects in Beijing were mostly descendants of the upper class families 57 58 24 and little physical work was required in their later life, with speculated impairment in 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 11 1 2 3 1 musculoskeletal coordination to explain both higher osteoporosis and fall associated fragility 4 5 2 fracture rate. 1,3,24,25 As our study subjects in remote Luliang county were working class women 6 7 3 and often had to perform mandatory labor work in old age, this might explain that the balance 8 9 10 4 analysis and life style questionnaire did not show difference between women with and without 11 12 5 bound feet evaluated in a small subset of subjects (Supplement III). The associated active 13 14 6 lifestyle might have led to different adaptations of the body balance throughout one’s lifetime to 15 For peer review only 16 17 7 compensate smaller foot support area. 18 19 20 8 The findings of our study on heel QUS and lifestyle in terms of SF 36 questionnaire offer a 21 22 9 unique look into the effects of foot binding. The significance of our current study in social and 23 24 25 10 health care aspects was the mobilization of the local government to support our project. Our 26 27 11 findings generated from the current study raised the awareness of osteoporosis prevention 28 29 12 30 measures in rural regions of China at one site and at another. Apart from difference in 31 26,27 32 13 nutrition, the importance of physical activities and/or exercise that are specifically relevant 33 http://bmjopen.bmj.com/ 34 14 for balance maintenance as this has been a known and crucial factor for prevention of fall and 35 36 7,10,28,29 37 15 fall induced fragility fractures. 38 39 40 16 Our current study had a few limitations apart from the cross sectional study in nature: 1) 41 42 17 Inhomogeneous number in study subjects: the higher percentage of bound feet subjects was due on September 29, 2021 by guest. Protected copyright. 43 44 18 45 to the greater percentage of elderly women with bound feet living in the location of study; 2) due 46 47 19 to limited budget and long distance between remote Luliang country and Hong Kong, only few 48 49 20 subjects were invited to fly to Hong Kong for HR pQCT, gait and balance tests (Supplements II 50 51 52 21 and III); 3) As only fracture history could be recalled for analysis, no reliable fall frequency 53 54 22 could be provided for analysis. 55 56 57 23 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 12 1 2 3 1 Conclusion 4 5 6 2 Foot binding had life long negative impact on bone health. However, the neuromusculoskeletal 7 8 9 3 coordination of the bound feet postmenopausal women with active life style might be 10 11 4 compensated as no difference was revealed in incidence of fragility fractures, implying the 12 13 5 importance of body balance maintenance in prevention of fragility fractures. 14 15 For peer review only 16 17 6 18 19 7 Funding This study was supported by the China Health Promotion Foundation (Ref. 2009 01) 20 21 8 and Contemporary Orthopaedic Research and Education Fund of the Chinese University of Hong 22 23 24 9 Kong (CUHK#6903361). We also appreciate technical support from Miss Fong TN and Miss 25 26 10 Hung WY of Bone Quality and Health Centre, Department of Orthopaedics and Traumatology, 27 28 11 The Chinese University of Hong Kong for their assistance in HR pQCT measurements. 29 30 31 32 12 33 http://bmjopen.bmj.com/ 34 13 Competing interests: The authors declare no competing interests. 35 36 14 37 38 39 15 Contribution statement: Ling Qin designed and organized this project, participated in data 40 41 16 collection, analysis and interpretation, and guided the preparation of this Manuscript; Yi Pan, 42 on September 29, 2021 by guest. Protected copyright. 43 17 Mian Xu, and Hanchang Lao recruited subjects, performed questionnaires and conducted QUS 44 45 46 18 scanning at Luliang county; Ming Zhang and Shan Cong performed gait analysis, and Xia Guo 47 48 19 performed balance test and related analysis in Hong Kong. Yanling Zhao co organized the study 49 50 20 51 and jointly applied for research grant for this project; Michael O’Laughlin did a part of the data 52 53 21 analysis and assisted in drafting the manuscript. WY Hong did the HR pQCT measurement and 54 55 22 its data analysis; JCY Cheng contributed to planning and preparation of the manuscript. 56 57 58 23 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 13 1 2 3 1 Data sharing: Authors agree with the data sharing as specified in the policy of BMJ Open to 4 5 2 make all submitted materials, including supplementary files available to the public. 6 7 8 3 Table 1: Anthropometric and lifestyle data compared between women with or without bound feet 9 10 4 (Data in mean ± SD) 11 12 Bound feet subjects Control subjects Difference (%) P- value 13 (N=172) (N=82) 14 Anthropometric data 15 For peer review only 16 17 Age (years) 74.6±3.5 74.5±4.0 0.20% 0.756 18 Weight (kg) 46.1±7.4 48.7±8.9 5.3% 0.017* 19 BMI (kg/m2) 20.8±3.2 21.9±3.5 5.1% 0.013* 20 Length of right foot (cm) 222.7±18.4 254.0±7.3 12.3% 0.000** 21 22 Width of right foot (cm) 74.3±8.4 99.8±7.0 25.6% 0.000** 23 Lifestyle data 24 25 YSM (years) 26.41±5.21 27.49±5.98 3.93% 0.142 26 Number of Children 7.02±2.63 6.21±2.13 13.00% 0.007** 27 Number subject with 13/172 (7.56%) 6/82 (7.30 %) 0.26%@ 0.846 28 previous fractures (%) 29 30 Osteoporosis (%) (QUS 165/172 (95.93%) 52/82 (63.42%) 32.51%@@ 0.000** 31 Stiffness Index) # 32 Odds ratio 33 5 *: p<0.05, **: p<0.01 (unpaired T test); #: T score set at 1.8 SD and chi square test used for http://bmjopen.bmj.com/ 34 6 35 statistical analysis; Odds ratio (OR): @, OR=1.036 (p>0.05); @@, OR=13,60 (p<0.01). 36 37 7 38 39 40 8 References 41 42 on September 29, 2021 by guest. Protected copyright. 43 9 1. Levy HS. Chinese Footbinding: The history of a curious erotic custom. New York, NY: W. 44 45 10 46 Rawls, 1966 47 48 11 2. Wang P. Aching for beauty: Foot binding in China. Minneapolis: University of Minnesota 49 50 12 Press, 2000 51 52 53 13 3. Fairbank JK. The Great Chinese Revolution: 1800 1985 New York, Harper & Row, 1986 54 55 14 4. Gerdhem P, Akesson K, Obrant KJ. Effect of previous and present physical activity on bone 56 57 15 mass in elderly women. Osteoporosis Int 2003; 14:208–12. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 14 1 2 3 1 5. Cooper C, Harvey NC. Osteoporosis risk assessment. BMJ 2012; 344:e3427. 4 5 2 6. Qin L, Au SZ, Choy YW, Leung PC, Neff M, KM Lee, MC Lau, Woo J, Chan KM. Regular 6 7 3 Tai Chi exercise may retard bone loss in postmenopausal women – A case control study. 8 9 10 4 Arch Phys Med Reha 2002; 83(10):1355 9. 11 12 5 7. Qin L, Choy WY, Leung KS, Leung PC, Au SK, Hung WY, Dambacher MA, Chan KM. 13 14 6 Beneficial effects of regular Tai Chi exercise on musculoskeletal system. J Bone Miner 15 For peer review only 16 17 7 Metabol 2005; 23(2):186 90. 18 19 8 8. Edwards MH, Jameson K, Denison H, Harvey NC, Sayer AA, Dennison EM, Cooper C. 20 21 9 22 Clinical risk factors, bone density and fall history in the prediction of incident fracture among 23 24 10 men and women. Bone 2013; 52(2):541 7. 25 26 11 9. Hippisley Cox J, Coupland C. Derivation and validation of updated QFracture algorithm to 27 28 29 12 predict risk of osteoporotic fracture in primary care in the United Kingdom: prospective open 30 31 13 cohort study. BMJ 2012; 344:e4191; 1 16. 32 33 14 10. Chan KM, Qin L, Lau M, Woo J, Au SK, Choy WY, Lee KM, Lee SH. A randomized, http://bmjopen.bmj.com/ 34 35 36 15 prospective study of the effects of Tai Chi Chun exercise on bone mineral density in 37 38 16 postmenopausal women. Arch Phys Med Rehab 2004; 85:717 22. 39 40 17 11. Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF 36). I. Conceptual 41 42 on September 29, 2021 by guest. Protected copyright. 43 18 framework and item selection. Med Care 1992; 30:473–83. 44 45 19 12. Ware JE, Kosinski M, Bayliss MS, McHormey, Rogers WH, Raczek A. Comparison of 46 47 20 48 methods for the scoring and statistical analysis of SF 36 health profile and summary 49 50 21 measures: summary of results from the Medical Outcomes Study. Med Care 1995; 33(4): 51 52 22 264 79. 53 54 55 23 13. Maglinte GA, Hays RD, Kaplan RM. US general population norms for telephone 56 57 24 administration of the SF 36v2. J Clin Epidemiol 2012; 65(5): 497–502. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 15 1 2 3 1 14. Economos CD, Sacheck JM, Wacker W, Shea K, Naumova EN Precision of Lunar Achilles+ 4 5 2 bone quality measurements: time dependency and multiple machine use in field studies. Br J 6 7 3 Radiol 2007; 80:919 25. 8 9 10 4 15. Vu TT, Nguyen CK, Nguyen TL, Le BM, Nguyen Trung Le Dm Bui TN, Bakamori M, 11 12 5 Kunii D, Sakai T, Yamamoto S. Determining the prevalence of osteoporosis and related 13 14 6 factors using quantitative ultrasound in Vietnamese adult women. Am J Epidemiol 2005; 15 For peer review only 16 17 7 161(9):824 30. 18 19 8 16. Frost ML, Blake GM, Fogelman I. Can the WHO criteria for diagnosing osteoporosis be 20 21 9 22 applied to calcaneal quantitative ultrasound? Osteoporos Int 2000; 11:321–30. 23 24 10 17. Liu JM, Ma LY, Bi YF, Huang Y, Xu M, Zhao HY, Sun LH, Tao B, Li XY, Wang WQ, 25 26 11 Ning G. A population based study examining Calcaneus quantitative ultrasound and its 27 28 29 12 optimal cut points to discriminate osteoporotic fractures among 9352 Chinese women and 30 31 13 men. J Clin Endocrinol Metab 2012: 97(3):800 9. 32 33 14 18. Hans D, Dargent Molina P, Schott AM, Sebert JL, Cormier C, Kotzki PO, Delmas PD, http://bmjopen.bmj.com/ 34 35 36 15 Pouilles JM, Breart G, Meunier PJ. Ultrasonographic heel measurements to predict hip 37 38 16 fracture in elderly women: the EPIDOS prospective study. Lancet 1996; 348:511–14. 39 40 17 19. Glueer CC. Quantitative ultrasound techniques for the assessment of osteoporosis: expert 41 42 on September 29, 2021 by guest. Protected copyright. 43 18 agreement on current status. J Bone Miner Res 1997; 12:1280–8. 44 45 19 20. Nayak S, Olkin I, Liu H, Grabe M, Gould MK, Allen IE, Owens DK, Bravata DM. Meta 46 47 20 48 analysis: accuracy of quantitative ultrasound for identifying patients with osteoporosis. Ann 49 50 21 Intern Med 2006; 6;144(11):832 41. 51 52 22 21. Lai BM, Tsang SW, Lam CL, Kung AW. Validation of the Quality of Life Questionnaire of 53 54 55 23 the European Foundation for Osteoporosis (QUALEFFO 31) in Chinese. Clin Rheumatol 56 57 24 2010; 29(9):965 72. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 16 1 2 3 1 22. Kanis JA. Diagnosis of osteoporosis and assessment of fracture risk. Lancet 2002; 4 5 2 359(9321):1929 36. 6 7 3 8 23. Kern LM, Powe NR, Levine MA, Fitzpatrick AL, Harris TB, Robbins J, Fried LP. 9 10 4 Association between screening for osteoporosis and the incidence of hip fracture. Ann Intern 11 12 5 Med 2005; 142(3):173 81. 13 14 6 15 24. Cummings S,For Ling X, Stonepeer K. Consequences review of foot binding only among older women in Beijing, 16 17 7 China. Am J Pub Health 1997; 87(10): 1677 9. 18 19 8 25. U.S. Preventive Services Task Force. Screening for osteoporosis: U.S. preventive services 20 21 22 9 task force recommendation statement. Ann Intern Med 2011; 1;154(5):356 64. 23 24 10 26. Grigg M, Arora M, Diwan A. Role of nutritional supplementation in elderly patients with hip 25 26 11 fractures. J Orthopaedic Translation 2014; 2(1): 26 34. 27 28 29 12 27. Qin L, Choy W, Hung V, Au S, Chan K, Leung K. Age related vessel calcification at distal 30 31 13 extremities is a risk factor of osteoporosis. J Orthopaedic Translation 2014; 2(1): 43 48. 32 33 14 http://bmjopen.bmj.com/ 34 28. Trombetti A, Hars M, Herrmann F, Rizzoli R, Ferrari S. Effect of a multifactorial fall and 35 36 15 fracture risk assessment and management program on gait and balance performances and 37 38 16 disability in hospitalized older adults: a controlled study. Osteoporosis Int 2013; 24(3):867 39 40 41 17 76. 42 on September 29, 2021 by guest. Protected copyright. 43 18 29. Edwards BJ, Song J, Dunlop DD, Fink HA, Cauley JA. Functional decline after incident 44 45 19 wrist fractures Study of Osteoporotic Fractures: prospective cohort study. BMJ 2010; 46 47 48 20 341:c3324. 49 50 21 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 Figure 1: A side-by-side aerial view drawing shows the physical difference between normal feet (A) and 36 bound feet (B). Lateral X-rays compared between postmenopausal women of normal foot (C) and with bound foot (D). 37 143x118mm (150 x 150 DPI) 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 45 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 Figure 2: Heel quantitative ultrasound (QUS) measurement for a women with bound feet. 80x75mm (96 x 96 DPI) 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 45 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 3: QUS calcaneal bone stiffness index-based T- and Z-scores of bound feet ladies (BFL) (n=172) 25 compared to the controls (n=82). **:p<0.01. 26 233x118mm (150 x 150 DPI) 27 28 29 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40 41 42 on September 29, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 45
1 2 3 4 Supplement I BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 Figure 1 Elderly women with bounded feet mainly reside in rural and impoverished 20 21 provinces of China. 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 Figure 2 A collaborative research project conducted by both clinical scientists and 37
bioengineers from Hong Kong and physicians from Mainland China to investigate the http://bmjopen.bmj.com/ 38 39 social and physiological impacts of lifelong foot binding on female musculoskeletal 40 41 health in Luliang County of Yunnan Province of Yunnan Province, China. 42 43 44 45
46 on September 29, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 45 BMJ Open
1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 Figure 3 Due to significant physical distance between Luliang County and Hong 29 30 Kong, only few bound feet women with mean QUS values representative of the 31 bound feet group were invited to travel long distance to Hong Kong for assessment of 32 33 bone microstructure, gait and balance using advanced biomedical technologies, i.e. 34 35 the invited women with bound feet needed to take bus from Luliang County to travel 36 around 200 kilometers to the capital city Kunming of Yunnan Province, China and 37 38 then took flight to travel over 1200 kilometers to Hong Kong. http://bmjopen.bmj.com/ 39 40 41 42 43 44 45
46 on September 29, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 45
1 1 2 3 4 Supplement II BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 7 1. Measurement of volumetric bone mineral density (vBMD) and microarchitecture 8 using high-resolution peripheral quantitative computed tomography (HR-pQCT) 9 10 vBMD and microstructure of the non-dominant distal tibia of the two women with 11 bound feet were evaluated using HR-pQCT (XtremeCT, Scanco Medical AG, 12 13 Switzerland). A standard scanning program was used and the respective region of 14 interest (ROI) of the scanned tibia were automatically separated into cortical and 15 16 trabecular compartments for calculating bone mineral density and microarchitecture 17 For peer review only 18 of both trabecular bone and cortical bone (Figure 1). Hong Kong female Chinese 19 population specific T-score at -2.5SD was used for diagnosis of osteoporosis.1 20 21 Analysis: For HR-pQCT, both T and Z scores from distal tibia of the two women with 22 bound feet were averaged for comparison with an age-matched reference database. 23 24 Mean values for each SPS balancing tests parameter was also calculated by averaging 25 results of the two bound feet women that were subsequently compared to age-matched 26 27 controls from population database of HR-pQCT. 28 29 30 31 32 33 34 35 36 37 38 http://bmjopen.bmj.com/ A 39 A B C 40 41 42 Figure 1 Bone assessment. Heel QUS measurement for a woman with bound feet (A), 43 44 HR- pQCT scanning on lower limb (B), and 3-D HR-pQCT images of distal tibia 45 showing inferior bone structural and density in woman with bound feet (C) and age- 46 on September 29, 2021 by guest. Protected copyright. 47 and gender-matched control with normal feet (D). 48 49 50 2. HR-pQCT Findings 51 Mean HR-pQCT values of volumetric BMD and microarchitecture in non-dominant 52 53 distal tibia of two women with bound feet is compared with reference population 54 55 database to calculate T- and Z-score and their percentage difference (Table 1, Figure 56 2). All parameters show inferior density and structural values in bound foot women, 57 58 with large BMD differences in total trabecular BMD (Dtrab) (42.7%) and inner 59 trabecular BMD (Dinn) (-90.5%) and large bone structural data in total trabecular 60 bone volume fraction (tBV/TV) and total trabecular separation (tTb.Sp). Another
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 45 BMJ Open
1 2 2 3 4 striking structural difference is seen in the total trabecular number (tTb.N) (36.0%), BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 with a much lower value in women with bound feet. To be noted that the cortical 6 7 vBMD (Dcomp) was much less affected as there was only 18.1% lower mean value as 8 compared with the age and gender matched reference population database. 9 10 11 12 13 Table 1 Distal tibia HR-pQCT values obtained for bound feet women (BFW) 14 compared to age and gender matched reference database values for bone geometry 15 16 and microarchitecture 17 For peer review only 18 Reference Database Difference 19 Distal Tibia BFW (Age 60-79) (%) 20 21 Bone geometry 22 Total Area (mm2) 509.35 622.11±145.45 -18.1% 23 2 24 Ct. Area (mm ) 71.65 88.08±32.76 -18.7% 25 Tb. Area (mm2) 426.85 520.55±152.19 -18.0% 26 27 Bone mineral density and microarchitecture 28 29 D100 (mg HA/cm³) 184.7 230.68±74.05 -19.9% 30 Dcomp. (mg HA/cm³) 768.10 782.58±90.89 -18.5% 31 32 Ct.Th. (mm) 0.815 0.91±0.38 -10.4% 33 Ct.Pm. (mm) 88.85 97.66±11.26 -9.0% 34 35 Dtrab. (mg HA/cm³) 66.85 116.65±42.19 -42.7% 36 Dmeta. (mg HA/cm³) 155.90 194.10±47.43 -19.7% 37 38 Dinn. (mg HA/cm³) 6.05 63.96±43.34 -90.5% http://bmjopen.bmj.com/ 39 tBV/TV 0.06 0.10±0.03 -42.3% 40 41 tTb.N 0.80 1.25±0.40 -36.0% 42 tTb.Th 0.07 0.08±0.03 -11.9% 43 44 tTb.Sp 1.20 0.77±0.32 56.3% 45 BF: Bound Foot; Ct. Area: Corticular Area; Tb. Area: Trabecular Area; D100: average 46 on September 29, 2021 by guest. Protected copyright. 47 vBMD; Dcomp: cortical density; Ct.Th: cortical thickness; Ct.Pm: cortical 48 49 perimeter; Dtrab: trabecular density; Dmeta: meta-trabecular density; Dinn: 50 inner-trabecular density (calculated from central 50% of the total trabecular bone 51 52 area); tBV/TV total trabecular bone volume fraction; tTb.N: total trabecular 53 number; tTb.Th: total trabecular thickness; tTb.Sp: total trabecular separation. 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 45
1 3 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 HR-pQCT T score HR-pQCT Z score 6 3 7 8 2 9 BFL Controls BFL Controls 10 11 1 12 13 0 14 -0.26 15 -1 16 -0.88 17 For peer review only -2 18 19 -2.38 20 -3 21 -3.02 22 -4 23 24 -5 25 26 27 Figure 2 T- and Z-scores of (total or inner) trabecular BMD measured by HR-pQCT 28 compared between the bound feet women (BFW) (mean of the two subjects) 29 30 compared to the controls from the reference database. Osteoporosis (%): 2/2 (100%) 31 in HR-pQCT trabecular BMD with T-scores set at -2.5 SD. 32 33 34 35 Key Reference 36 37 1. Tang XL, Qin L, Kwok AW, Zhu TY, Kun EW, Hung VW, Griffith JF, Leung PC, 38 Li EK, Tam LS. Alterations of bone geometry, density, microarchitecture, and http://bmjopen.bmj.com/ 39 40 biomechanical properties in systemic lupus erythematosus on long-term 41 glucocorticoid: a case-control study using HR-pQCT. Osteoporos Int 2013; 42 43 24(6):1817-26. 44 45
46 on September 29, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 45 BMJ Open
1 1 2 3 Supplement III 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 1. Methods: Gait and balance analysis 7 8 9 Gait analysis: Tekscan Walkway (Tekscan Inc., USA) was used to measure dynamic 10 plantar pressure distributions during walking for test subjects (Figure 1). This 11 2 12 walkway was 3 meters long, consisting of 4 sensors per cm sampling at a rate of 100 13 Hz. Subjects were asked to walk at a comfortable, self-chosen pattern and allowed to 14 15 familiarize themselves with the procedure and equipment before data collection. Peak 16 pressure of the whole foot and pressure of two major RIOs: forefoot and rearfoot were 17 For peer review only 18 collected during the stance phase of walking. Pressure values were also recorded for 19 analysis suing published protocol.1 20 21 Balance analysis: SMART EquiTest® (NeuroCom International, Inc., USA) was used 22 23 for assessing static postural stability (SPS) or balancing capabilities of the subjects 24 that were compared with built-in normal database values of Chinese females.2 Briefly, 25 26 a force platform within the system recorded the displacement of center of pressure 27 (CoP) where the subjects were required to complete three stance conditions, i.e. (a) 28 29 bipedal stance with eyes open (BiSEO), (b) bipedal stance with eyes closed (BiSEC) 30 and (c) single-leg stance with eyes open (SLSEO), with bare feet placed on the 31 32 markings of the force platform. Data were acquired at 100Hz over a period of 20 33 seconds in conditions (a) and (b), and 10 seconds in condition (c). Each condition was 34 35 performed thrice and maximum CoP displacements in antero-posterior (AP) and 36 lateral directions were extracted (Figure 2). In conditions (a) and (b), participants 37 38 stood at shoulder width; in condition (c), participants performed single-leg stance tests http://bmjopen.bmj.com/ 39 40 on dominant leg. 41 Analysis: Plantar pressure distribution charts were directly compared between the 42 43 right foot of the bound feet lady and that of a weight adjusted normal control of 44 similar age. What to note is that only two women with bound feet from Luliang to be 45
46 invited to Hong Kong and data were compared with Hong Kong local women. The on September 29, 2021 by guest. Protected copyright. 47 accuracy of the dada could be confounded by lifestyle difference where more active 48 49 lifestyle is normally associated with better gait and balance. However, the major 50 difference should be explained by foot anatomy, with significant small foot size in 51 52 women with bound feet as compared with that of women with normal foot. 53 Figure 1 Walking gait patterns: Tekscan 54 55 Walkway System was used to analyze lady 56 57 with bound feet (A) and a typical example of 58 a three step gait of a woman with bound feet 59 60 (B).
A B For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 45
1 2 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 7 8 9 10 11 12 13 14 A B 15 16 Figure 2 Static Posture Stability (SPS) assessment using SMART EquiTest® 17 (NeuroCom ForInternational, peer Inc., USA) review (A) and the displacement only of center of pressure 18 19 (B). 20 21 22 2. Findings 23 24 Gait analysis Figure 3 displays feet force during stance phase. The force during 25 26 propulsion of the stance (gait) phase for control subject is shown to occur at the 27 forefoot while for women with bound feet, the force occurs at the rearfoot while the 28 29 forefoot contributes only a minor force in its total pressure exerted. Foot binding thus 30 shifts the center of gravity towards the heel. Moreover, the rearfoot contact time is 31 32 found prolonged in bound feet women. For the control subjects, the rearfoot contact 33 begins with the heel strike during the beginning of the stance phase (0%) and ends at 34 35 the heel rise (80% of stance phase). For bound feet women, the heel contact time is 36 prolonged to over 90% of the stance phase. The forefoot contribution to the 37 38 propulsion is also only seen at the very end of stance phase. The difference in http://bmjopen.bmj.com/ 39 40 accumulated plantar pressure distribution between a normal and a woman with bound 41 foot is demonstrated where the bound feet lady shows a much larger degree of 42 43 pressure placed on the calcaneus of the heel and less force distributed over the 44 forefoot in comparison to a normal foot (Figure 4). 45
46 Balance test Figure 5 summarizes SPS measurements of two women with bound feet on September 29, 2021 by guest. Protected copyright. 47 in comparison to data from the gender- and age-matched control database. Overall, 48 49 the results indicate no obvious differences in all three categories of anterior-posterior 50 and lateral sway compared with between women with bound feet and the controls, 51 52 including degree of sway in bipedal stance with eyes open (BiSEO), bipedal stance 53 with eyes closed (BiSEC), and single-leg stance with eyes open (SLSEO). 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 43 of 45 BMJ Open
1 3 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 Figure 3 Representative foot force compared between subjects with bound feet and 20 gender- and age-matched control, showing that a bound feet lady places almost the 21 22 entirety of her foot force on the rearfoot for the predominant part of the stance phase 23 with prolonged contact time, while the forefoot contributes to the propulsion only at 24 25 the very end of the stance phase (Left). A gender- and age-matched control subject 26 shows a more balanced pressure distribution of the rear and forefoot (Right). 27 28 29 30 31 32 33 34 35 36 37 38 http://bmjopen.bmj.com/ 39 40 41 42 43 44 45 46 on September 29, 2021 by guest. Protected copyright. 47 48 49 50 Figure 4: Comparison of representative plantar pressure distribution in women with 51 normal feet and bound feet: regions in red represent areas of highest pressure (>231 52 53 kPa), while regions in blue represent areas of lowest pressure (>0 kPa). 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 44 of 45
1 4 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 3.5 Control Bound feet 6 7 3 8 9 10 2.5
11 12 2 13 14 1.5
15 (cm)Sway 16 17 1 For peer review only 18 19 0.5 20 21 0 22 BiSEO (cm) BiSEC (cm) SLSEO (cm) BiSEO (cm) BiSEC (cm) SLSEO (cm) 23 24 Anterior-Posterior Sway Lateral Sway 25 26 Figure 5: Static postural stability (SPS) balance test shows no obvious difference in 27 28 both anterior-posterior and lateral sway categories compared for the mean of two 29 women with bound feet and the gender- and age-matched reference population 30 31 database. BiSEO: bipedal stance with eyes open; BiSEC: bipedal stance with eyes 32 33 closed; and SLSEO: single leg stance with eyes open. 34 35 36 Specification of the relevant anthropometric information of the subset of the women 37 with bound feet for gait and balance tests: both of them were also within the range but 38 http://bmjopen.bmj.com/ 39 slightly above the average of women with bound feet with BW 48.2 and 47.5 kg; 40 while with BMI of 21.0 and 20.6, respectively. This number is close to the mean of 41 42 the control group (refer to Table 1). 43 44 45 A small subset was tested for gait and balance using state-of-the-art facilities and the 46 data suggested that in spite of variation in gait patterns, the balance capability were on September 29, 2021 by guest. Protected copyright. 47 48 similar irrespective of size of the feet, i.e. significantly smaller in manmade small feet 49 in women with bound feet. Since the subjects were old and were not able to delineate 50 51 if the previous fracture was a result of fall or spontaneous fracture due to poor bone 52 53 quality, the overall findings however might still imply the essential role of body 54 balance in prevention of fragility fracture. 55 56 57 Key Reference: 58 59 1. Mak AF, Zhang M, Boone DA. State-of-the-art research in lower-limb prosthetic 60 biomechanics-socket interface: a review. J Rehabil Res Dev. 2001; 38(2):161–74.
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 45 of 45 BMJ Open
1 5 2 3 2. Cheing GL, Chau RM, Kwan RL, Choi CH, Zheng YP. Do the biomechanical 4 BMJ Open: first published as 10.1136/bmjopen-2014-006521 on 17 March 2015. Downloaded from 5 properties of the ankle-foot complex influence postural control for people with 6 Type 2 diabetes? Clin Biomech 2013; 28(1):88-92. 7 8 9 10 11 12 13 14 15 16 17 For peer review only 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 http://bmjopen.bmj.com/ 39 40 41 42 43 44 45
46 on September 29, 2021 by guest. Protected copyright. 47 48 49 50 51 52 53 54 55 56 57 58 59 60
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml