Introduction

Reconstruction of stature is one of the important aspects of various parameters of identification for establishing individuality of the person. Under the circumstances, where mutilated, decomposed or fragmentary skeletal remains are recovered, the stature of an individual may be estimated by adopting Anatomical Method ₁ , if complete skeleton is available for examination or by following mathematical method where measurement of a single long bone may serve the purpose because there is a strong relation between skeletal element and stature. This means that measurement of any bone or combinations of bone reflect stature.

The forensic scientists are well aware of the fact that they have a very little choice of using anatomical method of stature reconstruction while dealing with skeletal remains. It is obvious that the complete skeleton is rarely available at the scene of crime and thus the scientists have no choice than to use a relatively less effective method of stature reconstruction, i.e. mathematical method, due to its obvious advantage that it is workable even if a single long bone of the upper or the lower extremity is available for examination. In most advanced countries documented skeletal remains are available to the forensic scientists, in India the situation is different and documented skeletal remains are not available for this purpose. In the absence of documented skeletal material the researchers have focused their attention towards living population groups of India and have taken relevant bone lengths over the skin ₂ and correlated them with the stature to find out the degree of relationship between them and subsequently formulated regression formulae for reconstruction of stature. Thus almost all the studies conducted by researchers in India pertains to the use of percutaneous measurements of body including upper and the lower extremity bone lengths for reconstruction of stature ₃,₄,₅,₆,₇,₈,₉,₁₀,₁₁,₁₂,₁₃,₁₄,₁₅,₁₆,₁₇ . All these studies have reported significant difference in the proportions of the limb bone dimensions, due to the environmental and genetic differentiation.

It is a known fact that the different population groups exhibit variation in their body proportions as a result of which correlation of one bone length to stature not only varies from population to population but also between sexes. Keeping this in mind an attempt has been made in the present study to reconstruct stature among male and female Shia Muslims of Delhi using five body measurements pertaining to lower extremity dimensions (femur length, tibial length, fibular length, foot length and foot breadth). It may be mentioned here that out of these lower limb measurements foot breadth (FB) have been used only in isolated studies.

Material And Methods

The present study is comprises of 503 male and 508 female Shia Muslims belonging to Delhi, India within the age range of 20 to 40 years. All the subjects were measured for five lower limb dimensions, viz. femur length, tibial length, fibular length, foot length and foot breadth in accordance with the standard measurement techniques recommended by Martin and Saller ₁ and Allbrook ₂ . All observations were recorded in centimeters (cm). Each subject was measured for the following percutaneous dimensions besides stature:

Stature (S): It is obtained as the projective distance between the standing surface and the highest point on the head (vertex) when the subject is standing in the standard standing position, using anthropometer.

Femur Length (FEML): It is measured when the subject stands erect with the left leg placed slightly ahead of the right one and the foot partly inverted to relax the soft tissues. The measurement is obtained as the distance from the upper most point on the greater trochanter to the lower most point palpable on the lateral femoral condyle, using rod compass.

Tibial Length (TIBL): The subject sat with left knee placed in the semi flexed position and the left foot partly inverted to relax the soft tissues and render bony landmarks prominent. The length of tibia from the medial condyle (as it becomes palpable and diverges anteriorly from the articulating femoral condyle) to the tip of the medial malleolus is measured, using rod compass

Fibular Length (FIBL): The subject sat with left knee placed in the semi flexed position and the left foot partly inverted to relax the soft tissue and render bony landmarks prominent. The distance between the upper most point palpable on fibular head, (little below the lateral margin of the knee) and the tip of the lateral malleolus is measured, using rod compass.

Foot Length (FL): The subject stands erect with left foot twelve inches forward than right foot with whole weight of body falling on left foot. The distance between acropodian and pternion is measured when the foot is fully stretched, using rod compass.

Foot Breadth (FB): It is obtained as a distance between metatarsal tibiale and metatarsal fibulare, using sliding caliper.

The data have been treated statistically using the standard programme of SPSS, to obtain mean, standard error of mean, test of significance and coefficient of correlation to assess the variations for all the lower limb measurements among male and female Shia Muslims. Besides these, multiplication factors and regression equations were also formulated for stature reconstruction.

Results And Discussion

Tables 1 and 2 list basic statistical constants for all the lower extremity measurements besides stature for male and female Shia Muslims of Delhi respectively. It is observed that the males have greater mean values for all the five measurements of the lower limb and stature than the females. The mean difference in the values of these measurements varies from 13.26 cm in case of stature to a minimum of 1.24 cm for foot breadth. However among the three lengths of the lower limb the fibular length exhibits maximum differences between the sexes followed by tibial length and femur length.

Figure 1

Table 1: Mean and other Statistical Constants of Lower Extremity Dimensions and Stature among Male Shia Muslims

Figure 2

Table 2: Mean and other Statistical Constants of Lower Extremity Dimensions and Stature among Female Shia Muslims

Table 3 displays sex differences in lower extremity measurements and stature among Shias of Delhi. It is perceptible from the table that the male Shias are not only taller than the females but also have longer dimensions for all the five lower limb measurements. The differential trends as assessed by means of t-test reveal highly significant sex differences (p<0.001) for all the lower extremity dimensions as well as stature. Owing to the highly significant sex differences the data on male and female Shias have been treated separately for further analysis pertaining to formulation of multiplication factors and regression equations for estimation of stature.

Figure 3

Table 3: Sex Differences in Different Measures of Lower Extremity and Stature among Shia Muslims of Delhi

Table 4 presents the mean M.Fs formulated, for male and female Shia Muslims for all the five lower limb dimensions. It is observed that the females exhibit greater mean M.Fs for foot length and foot breadth, while in case of femur length, tibial length and fibular length the males exhibit greater mean M.Fs thereby unfolding the fact that the males and females have variable proportions of their lower limb dimension with stature and also observed in case of other living populations.

Figure 4

Table 4: Multiplication Factors (M.Fs) of Lower Extremity Dimensions for Estimation of Stature among Male and Female Shia Muslims

Table 5 presents the linear regression equations for reconstruction of stature among male Shias from the five measurements of lower extremity i.e. femur length, tibial length, fibular length, foot length and foot breadth, along with standard error of estimate and coefficient of correlation (‘r') values between bone length and stature. It is perceptible from the table that all the measurements of the lower extremity exhibit correlation value above 0.5 except foot breadth. Tibial length exhibits the overall highest correlation (r = 0.765) with stature followed by fibular length, femur length and foot length, whereas foot breadth exhibits the least correlation with stature (r = 0.383). The standard error of estimate (SEE) is least with tibial length suggesting that the tibial length would provide the most dependable estimate of stature among male Shia Muslims. However the value of SEE is not much higher for fibular length and femur length and thus these measurements could also be used for obtaining a reasonably precise estimate of stature.

Figure 5

Table 5: Linear Regression Equations for Estimation of Stature Using Lower Extremity Dimensions among Male Shia Muslims

Table 6 presents the linear regression equations for reconstruction of stature among female Shia Muslims. In case of females it is observed that all the five measurements of the lower limb exhibit a comparatively low value of correlation with stature (between 0.267 – 0.742) as compared to the males. Femur length exhibits the highest correlation with stature followed by fibular length, tibial length and foot length whereas foot breadth exhibits the least correlation with stature. However the standard error of estimate (SEE) is least for femur length signifying that it would provide the most reliable estimate of stature among females. Despite a relatively high SEE for fibular length, tibial length and foot length they could also provide a convincingly accurate estimate of stature for Shia females.

Figure 6

Table 6: Linear Regression Equations for Estimation of Stature Using Lower Extremity Dimensions among Female Shia Muslims

In contrast to the long bone dimensions the foot length has a relatively high correlation with stature for both male (r = 0.546) and female (r = 0.538) Shias but the SEE is also high therefore it may not provide as reliable estimate as it accessible with tibial length for males and femur length for females. Thus foot length should be used for reconstruction of stature only in the absence of other dimensions of the lower extremity. However the estimated stature may not be quite reliable using foot breadth among both males and females as foot breadth has the least correlation with stature. But this dimension cannot be ignored due to low correlation value because if we come across a body part in fragmentary form, by measuring foot breadth an estimate of stature can be obtained.

An endeavor has also been made to appraise the variations in the estimated and actual stature using multiplication factors and regression equations, for all the dimensions for male and female Shias (Table 7). To observe the variations in the actual and the estimated stature using multiplication factors and regression equations approximately 10 percent of the total sample (50 males and 50 females) was obtained fresh and then subjected to analysis.

It is observed that the percentage error in the estimated stature diminishes adequately on using regression equations in contrast to the multiplication factors. The percentage error on using the multiplication factors varies from a minimum of 1.80% in case of femur length among females to a maximum of 4.54% in case of foot breadth among males. However, on using the linear regression equations the percentage error reduces effectively and ranges between 1.35% in case of femur length among females to 2.68% in case of foot breadth among males.

Figure 7

Table 7: Assessment of Percentage Error in Actual and Estimated Stature using Multiplication Factors and Linear Regression Equations

This clearly indicates that the dependability in the predicted stature would be better on using linear regression equations for any of these body dimensions as compared to the use of M.Fs for this purpose. It may also be emphasized that as the value of correlation is higher for all the lower limb measurements considered in the present study for males in contrast to the females and thus the consistency of the estimated stature would be greater among males than the one observed in case of females. It may also be emphasized that all the four length measurements exhibit a correlation, which is greater than 0.5 for both males and females and, thus, any of these lengths could offer a reliable estimate of stature for both male and female Shia Muslims of Delhi. Therefore depending upon the availability of the body part pertaining to the lower limb, stature may be estimated using linear regression equations or M.Fs with reasonable accuracy.

Table 8 presents the comparative account of M.Fs. for lower extremity bone dimensions i.e. FEML, TIBL, FIBL and FL among different living population groups of India.

A total of thirty-two M.F. values for FEML (including present study) have been reported of which thirteen pertains to males and eighteen to females.

It is evidenced from the table that the M.F. for males varies from a minimum of 3.16 among male Rajputs of Garhwal, U.P. ₁₈ to a maximum of 4.42 among male Rajput of Ranikhet, U.P. ₁₀ while among females the M.F. values varies from a minimum of 3.33 among female Brahmins of Garhwal, U.P. ₁₉ to a maximum of 5.27 among Warli females of Maharashtra ₂₀ . The male Shias of Delhi (Present Study) with the M.F. value as 4.02 fall very close to Rajput males (M.F. = 4.01) of Sirmour, H.P. ₂₁ . The female Shias of Delhi (Present study), on the other hand, show exactly the same value of 3.97 as reported for Jat females of Churu, Rajasthan ₂₂ .

A total of forty-six M.F. values for TIBL (including present study) have been reported among different living population groups, out of which twenty-one pertains to males and twenty-five to females.

It is observed from the table that the M.F. for males vary from a minimum of 4.00 among Yadav Males of Delhi ₂₃ (Nath et. al. 1999) to a maximum of 5.21 among Rajput males of Ranikhet, U.P. ₂₄ (Tiwary, 1986) while among females the M.F. varies from a minimum of 4.15 among Munda females of West Bengal ₂₅ (Duggal, 1985) to a maximum of 4.76 among Rajput females of Garhwal, U.P. ₁₈ .

Figure 8

Table 8: Multiplication Factors for Lower Extremity Dimensions among Different Living Indian Populations

The Shia males of Delhi (present study) with the M.F. value as 4.60 lie close to the Shia females of Delhi (Present study), Brahmin males of Garhwal, U.P. ₁₉ with a common value of M.F. as 4.59.

The Shia females, on the other hand, show exactly the same value of 4.59 as reported for Brahmin males of Garhwal, U.P. ₁₉

A total of eighteen M.F. values for FIBL (including present study) have been reported, among different living population groups of India. Out of which eight pertains to males and ten pertains to females.

It is evidenced from the table that the M.F. for males vary from a minimum of 4.11 among Jat males of Churu, Rajasthan ₂₂ to a maximum of 4.80 among Brahmin males of Ranikhet, U.P. and male Rajputs of Ranikhet, U.P. while among female the M.F. varies from a minimum of 4.13 among Jat females of Churu, Rajasthan ₂₂ to a maximum of 4.44 among Rajput females of Sirmour, H.P ₂₁ . The male and female Shia Muslims of Delhi (present study) with the M.F. values as 4.32 and 4.33 respectively lie close to the Jain females (M.F. = 4.34)of Delhi ₆ .

In case of FL, a total of forty-four M.F. values (twenty for males and twenty four for females) have been reported among different living population groups of India including the present study (Table 8).

It is evidenced from the table that M.F. for males vary from a minimum of 6.18 among male Brahmins of Garhwal, U.P. ₁₉ to a maximum of 7.46 among Rajput males of Ranikhet, U.P. while among females the M.F. varies from a minimum of 6.26 among Munda females of West Bengal ₂₆ to a maximum of 7.28 among Rajput females of Dehradun ₃₅ . The male Shia Mulsims of Delhi (present study) with the M.F. value as 6.76 fall very close to female Jains (M.F. = 6.75) of Delhi ₂₇ whereas the female Shia Muslims of Delhi (present study) with the M.F. value as 6.84 exhibit closeness with females Jains (M.F. = 6.83) of Delhi ₆ .

It is clear from this analysis that the values of M.F. of a particular bone length among living population vary from one group to another as well as from one sex to another, though certain groups may exhibit same M.F. value for a particular bone but it does not mean that the proportions of these bone length to stature are alike for these groups.

In connection with the formulation of multiplication factors Siddique and Shah ₃₆ commented that the study of one state is not necessarily applicable to another state. Singh and Sohal ₃₇ conducted a study on Punjabis of Amritsar and when their M.Fs. were compared with that of the ones formulated by Siddique and Shah ₃₆ on Punjabis of Lahore, a wide variation was observed in the multiplication factors for all the six long bones. This variation clearly indicates that the M.F. reported on a similar population group belonging to two different regions indicate variation in the M.F. values. A similar observation is noticed in the present study. Table 8 reveals that certain population groups like Brahmins and Rajputs inhabiting different geographical locations in U.P. and H.P. exhibit variation in their M.F. values. Similarly same population groups living in the same geographical location, like Jains and Punjabis of Delhi also show variation in their M.F. values. These variations clearly indicate that individuals belonging to the similar genetic composition but inhabiting different or similar geographical locations vary in their body proportions as a consequence of which the multiplication factor (i.e. Stature/ Bone length ratio) differs. This is a clear indication that secular trends also play an important role in this aspect and the multiplication factors formulated on a population needs to be revised at least once in a decade to have greater accuracy in the predicted stature among the living populations.

Table 9 presents a comparative account of regression equations formulated for FEML on different living Indian populations. It is evident from the table that there exists sufficient variation in the values of intercept and the slope of the prediction equations. The value of coefficient of correlation (r) (wherever listed) clearly describes the extent of relationship between femur length and stature in these population groups. The value of ‘r' varies from a minimum of 0.252 among Brahmin males of Garhwal, U.P. ₁₉ to a maximum of 0.979 among Brahmin females of Garhwal, U.P. ₁₉ . The male and female Shia Muslims of Delhi (present study) with the ‘r' values as 0.743 and 0.742 respectively fall well within the minimum and the maximum range of the coefficient of correlation. The value of ‘r' among male Shia Muslims of Delhi falls closest to female Shia Muslims of Delhi (present study) and vice versa.

Figure 9

Table 9: Regression Equations for FEML (Femur Length) among Different Living Indian Populations

The variation observed in all the thirty-six regression equations listed in the table clearly indicates that these are both population and sex specific like the multiplication factors. The prediction value of the regression equations for male and female Shia Muslims of Delhi based on femur length is on the higher side as both the sexes exhibit a high correlation between femur length and stature (Males:r=0.743;Females: r = 0.742) but the SEE is relatively high. Thus, femur length may be used to estimate stature for male and female Shia Muslims of Delhi.

Table 10 presents a comparative account of regression equations formulated for TIBL on different living Indian populations. The value of ‘r' varies from a minimum of 0.225 among Rajput females of Garhwal, U.P. ₁₈ to a maximum of 0.884 among Brahmin males of Sundernagar, H.P. The male and female Shia Muslims of Delhi (present study) with the ‘r' values as 0.765 and 0.717 respectively fall well within the minimum and the maximum range of the coefficient of correlation. The value of ‘r' among male Shia Muslims of Delhi lie close to Rajput males of Sirmour, H.P. ₂₁ where as for female Shia Muslims the value of ‘r' fall quite close to Hindu Baniya females of Delhi ₄

Figure 10

Table 10: Regression Equations for TIBL (Tibial Length) among Different Living Indian Populations

The variations observed in all the thirty-six regression equations listed in the table clearly indicates that these are both population and sex specific like the multiplication factors. The prediction value of the regression equations for male and female Shia Muslims of Delhi based on TIBL is on the higher side as both the sexes exhibit a high correlation between TIBL and stature (Males: r = 0.765; Females: r = 0.717) but the SEE is relatively high. Thus, TIBL may be used to estimate stature for male and female Shia Muslims of Delhi.

Table 11 presents a comparative account of regression equations formulated for FIBL on different living Indian populations. The value of ‘r' varies from a minimum of 0.540 among Brahmin females of Sampla, Haryana ₉ to a maximum of 0.90 among Rajput males of Ranikhet, U.P. The male and female Shia Muslims of Delhi (present study) with the ‘r' values as 0.758 and 0.718 respectively fall well within the minimum and the maximum range of the coefficient of correlation. The value of ‘r' among male Shia Muslims of Delhi lie close to Jain males of Delhi ₂₇ where as for female Shia Muslims the value of ‘r' fall quite close to Rajput males of Sirmour, H.P. ₂₁ .

Figure 11

Table 11: Regression Equations for FIBL (Fibular Length) among Different Living Indian Populations

The variations observed in all the sixteen regression equations listed in the table clearly indicate that these are both population and sex specific like the multiplication factors. The prediction value of the regression equations for male and female Shia Muslims of Delhi based on fibular length is on the higher side as both the sexes exhibit a high correlation between fibular length and stature (Males: r = 0.758; Females: r = 0.718) but the SEE is relatively high. Thus, fibular length may be used to estimate stature for male and female Shia Muslims of Delhi.

Table 12 presents a comparative account of regression equations formulated for FL on different living Indian populations. The value of ‘r' varies from a minimum of 0.330 among Rajput males of Garhwal, U.P. ₁₈ to a maximum of 0.921 among Brahmin females of Garhwal, U.P. ₁₉ . The male and female Shia Muslims of Delhi (present study) with the ‘r' values as 0.546 and 0.538 respectively fall well within the minimum and the maximum range of the coefficient of correlation. The value of ‘r' among male Shia Muslims of Delhi falls closest to female Shia Muslims of Delhi (present study) and vice versa.

Figure 12

Table 12: Regression Equations for FL (Foot Length) among Different Living Indian Populations

The variation observed in all the thirty-nine regression equations listed in the table clearly indicates that these are both population and sex specific like the multiplication factors. The prediction value of the regression equations for male and female Shia Muslims of Delhi based on foot length is on the higher side as both the sexes exhibit a high correlation between foot length and stature (Males: r = 0.546 ; Females: r = 0.538) but the SEE is relatively high. Thus, foot length may be used to estimate stature for male and female Shia Muslims of Delhi.

Conclusions

• Male Shias exhibit greater dimensions than the females for all the lower limb measurements and stature.

• t-test reveals highly significant sex differences for all the measurements including stature.

• Males exhibit greater mean M.Fs for long bones of the lower limbs (FEML, TIBL, and FIBL). However for remaining body dimensions of the lower limb i.e. FL and FB, the females show greater M.F. values.

• Among males, TIBL exhibits the highest correlation value of 0.765, while FB exhibits the lowest correlation value of 0.383.

• FEML among females, reveals the highest correlation (r =0.742), whereas FB exhibits the lowest correlation value of 0.267.

• Male Shias exhibit greater correlation values for all the lower limb dimensions than the females.

• In case of the long bones of the lower limbs, among males, TIBL exhibits the highest value of correlation (r =0.765), followed by FIBL (r =0.758) and FEML (r =0.743). While among females, FEML provides the highest correlation (r =0.742), followed by FIBL (r =0.718), and TIBL (r =0.717).

• Foot breadth shows correlation value less than 0.5 for both the sexes. Thus, the stature estimate obtained using this may not be as reliable as obtained through long bones of the lower limb.

• The females exhibit lesser average difference and the percentage error between observed and the estimated stature, than the males, on using multiplication factors and regression equations for almost all the lower limb measurements.

• Multiplication factors formulated for Shia Muslims show variation from that of the multiplication factors on other Indian populations.

• The present study also indicate that the secular trends play an important role and the multiplication factors formulated on a population needs to be revised at least once in a decade to have greater accuracy in the prediction of stature among the living populations.

Correspondence to

Prof. Surinder Nath, PhD Department of Anthropology, University of Delhi, Delhi-110007 INDIA E-mail: snath31@rediffmail.com