The present community-based cross-sectional investigation was conducted among 300 unrelated (150 males; 150 females) indigenous Karbi tribal individuals (aged from 20-50 years) of Karbi Anglong district of Assam, Northeast India. Ethnically, the Karbi belongs to an endogamous Mongoloid tribal population which is a Tibeto-Burman linguistic family. They occupy the districts of Karbi Anglong, Dima Hasao, Marigaon, Kamrup, Golaghat, Nagaon, Lakhimpur, Cachar and Sonitpur of Assam and the states of Arunachal Pradesh, Nagaland and Meghalaya of Northeast India. However, they are mostly found to be concentrated in the district of Karbi Anglong (25°33/N to 26°35/N latitude and 92°10/E to 93°50/E longitude) [Table/Fig-1]. This district is the largest of 32 districts of the state of Assam and covers a total area of 10,434 sq km. According to the National Census of 2011, the district had a total population of 9,56,313 individuals (males: 4,90,167; females: 4,66,146) with an average literacy rate of 69.25% (males: 76,14%; females: 62.00%). The Karbi population was selected for the present investigation because of its widespread distribution of population in the district Karbi Anglong. Initially, a total of 340 Karbi individuals aged 20-50 years were approached by using random sampling for the necessary participation in the study. All the subjects included in the study were physically fit and free from any physical deformities of the limbs. The subjects were also interviewed informally regarding any previous physical injury and/or operation, or any trauma in the either hand which was also taken into consideration during the time of examination. The subjects who were having any kind of abnormalities, injuries or surgical episodes were excluded from the study. Out of 340, 40 individuals (11.76%) declined or avoided to take further part in the study due to limb deformities and physical deformities.

[Table/Fig-1]:

Showing the location of study area Karbi Anglong Assam, Northeast India.

A written consent was also obtained from the research participants who took part in the study before the collection of relevant information and anthropometric measurements. The participants were informed about the purpose and scope of the present study. The participation of the subjects in the present investigation was purely voluntary in nature. Necessary study permissions were also taken from the local village level authorities and traditional village headmen. The present study was conducted following the ethical guidelines of human experiments as laid down in the Helsinki Declaration [35]. The anthropometric data were collected during the period from September 2015 to August 2016. Moreover, all the individuals included in this study were engaged in agriculture, thereby controlling the twin factors of physical activity and nutritional condition to a large extent. The age of the participants was recorded and verified from the age certificates (e.g., birth certificates) and other age-related proofs (e.g., voter identification cards, driving license) issued by the Government authorities. The populations were identified as the Karbi tribal population by observing their physical appearance and cultural features and recording their surnames. These were subsequently verified from the official records. All the individuals were the residents of the Karbi dominated villages viz., Taralangso, Chephongkimi, Rongchedon, Rongkimi, Dikrenglangso and Rongjangphong of Diphu sub-division, Karbi Anglong district of Assam, Northeast India. These villages were selected due to homogeneity and easy accessibility by road. The villages were situated at a distance of approximately 25 km south-west of Diphu sub-division town of Karbi Anglong.

Five anthropometric measurements were recorded from the subjects following the standard anthropometric procedures [36,37]. The measurements taken are as follows:

Stature was recorded with the help of an anthropometric rod placed in the vertical distance from the vertex to floor with the head oriented in the Frankfurt horizontal plane. HL is measured from the midpoint of the distal transverse crease of the wrist (i.e., extending from ulnar to radial side) to the most anterior projection of the skin of the middle finger (e.g., inter stylion line). HB is measured as a distance between the radial side of the second metacarpophalyngeal joint (e.g., metacarpal rediale) and the ulnar side of the fifth metacarpophalangeal joint (e.g., metacarpal ulnale).

The descriptions of the relevant landmarks associated with HL and HB were shown in [Table/Fig-2]. A slide caliper was used to measure the length of LHL, RHL, LHB and RHB. All anthropometric measurements were recorded with accuracy nearest to 0.1 cm. The participants were barefooted at the time of recording the measurements. Owing to the diurnal variation in stature [38], the subjects were measured during the morning hours prior to leaving for their work. However, the subjects covered under this present study were also measured with ample precision to avoid any possible systematic errors and define the landmarks in the process of anthropometric data collection. However, repeated/multiple measurements were also performed to check the possible reliability of the anthropometric measurements in field situations by both the authors.

[Table/Fig-2]:

Showing the landmarks involved to Hand Length (HL) and Hand Breadth (HB).

The technical errors of measurement {TEM=√(ΣD²/2N)}, where D=Difference between the anthropometric measurements, N=the number of individuals measured) which is an accuracy index and measures the standard deviation between repeated measures which [39] have been determined to check the consistency of the data. Even though number of methods of measuring inconsistency are available, the preferred method involves calculation of relative TEM and subsequent determination of the coefficient of reliability [R={1-(TEM)²/SD²}], (SD=the standard deviation of all the measurements) [37]. The TEM were calculated from multiple measurements obtained from the 30 randomly selected Karbi individuals. Very high reliability (R) (R>0.98) was obtained in TEM analysis in height, LHL, RHL, LHB and RHB and values were found to be within the acceptable limits of R=0.95 [37]. Hence the measurements were expected to be reliable and TEM values were not incorporated into further statistical procedures.

All statistical analysis was done using the statistical package for social science (SPSS, Version, 17.0). The data obtained was statistically analysed using statistical constants which includes correlation, regression, paired sample t-test and one-way analysis of variance (ANOVA). Linear and step-wise multiple regression equations were formulated separately for each sex and entire sample together to find out whether a single equation could be used for all age groups or independent equation would be required separately for the individual age group to estimate stature using HL and HB. Efforts were also made to formulate multiplication factors based equations by using HL, HB and age as continuous variables to improve the accuracy for estimation of stature among individuals. Linear regression was used for estimating the stature from LHL, RHL, LHB and RHB. Step-wise multiple regressions were also done to estimate stature from a combination of these variables. ANOVA was utilised to assess bilateral differences in LHL, RHL, LHB and RHB. To understand the sex differences between Karbi male and female individuals in stature, LHL, RHL, LHB and RHB, ANOVA and paired t-test were utilised. A p-value of <0.05 and <0.01 were considered to be statistically significant, while p>0.05 was considered to be statistically not significant.

The sex-specific mean, standard deviation (SD) and range of stature, LHL, RHL, LHB and RHB among adult Karbi individuals are presented in [Table/Fig-3]. It was observed that in the case of the male individuals, LHL ranged from 10.59% to 13.87% of stature, LHB ranged from 4.85% to 5.89% of stature; RHL varied between 10.46% and 13.76% of stature and RHB varied between 4.79% and 5.73% of stature. It was also observed that LHL ranged from 10.50% to 11.12% of stature, LHB ranged from 4.62% to 5.37% of stature, RHL varied between 10.43% and 11.54% of stature and RHB varied between 4.66% and 5.37% of stature in the case of female individuals. Therefore, it is apparent that male Karbi individuals have longer HL related to their stature than female individuals, but when HB was taken into consideration, the females have broader HB as compared to male counterparts in relation to their stature. Following the classification of stature proposed by Martin and Saller [40]. It is observed that the majority of the Karbi male and female individuals fall in the ‘short to medium’ but the Karbi female individuals were found to be slightly shorter than the male Karbi counterparts [40]. The Pearson correlation coefficient analysis between the anthropometric variables showed statistically significant coefficient between the stature and LHL (r=0.694), RHL (r=0.693), LHB (r=0.621) and RHB (r=0.639) (p<0.01). Similarly, statistically significant correlation coefficient values were obtained between the stature and LHL (r=0.453), RHL (r=0.452), LHB (r=0.235) and RHB (r=0.239) (p<0.01) among males and stature and LHL (r=0.578), RHL (r=0.587), LHB (r=0.335) and RHB (r=0.329) (p<0.01) among female individuals.

It is evident that the females exhibit slightly smaller stature and smaller HL and HB with respect to LHL, RHL, LHB and RHB as compared to the males Karbi counterparts. Utilising the paired sample t-test, it was further observed that there were statistically significant differences (p<0.01) between the sexes, when HL, LHL, RHL, HB, LHB, RHB and stature are considered among Karbi individuals [Table/Fig-4]. Therefore, the results showed that adult Karbi male and female individuals significantly differed in HL rather than HB, thereby indicating sexual dimorphism (p<0.05).

The bilateral differences in HL and HB among males and females Karbi individuals were evaluated using paired sample t-test analysis and the results of the statistical analysis is depicted in [Table/Fig-5]. It is evident from the table that the bilateral differences in HL (in males) and HB (in both males and females) were found to be statistically significant (p<0.05). The bilateral differences were also found to be statistically significant for unknown sex hand dimensions of HL and HB among Karbi individuals (p<0.05).

Simple linear regression analysis was done to estimate the stature from the studied variables. Stature was taken as the dependent variable and HL, LHL, RHL, HB, LHB and RHB as the independent variables in linear regression analysis. The regression equations for the estimation of stature from HL, LHL, RHL, HB, LHB and RHB among Karbi male and female individuals in the present study are depicted in [Table/Fig-6]. In all the cases, the regression coefficients were statistically significant (p<0.05) and the values for the R and R² regression equations are also given in [Table/Fig-6]. Therefore, the stature was thus observed to be dependent on HL, LHL, RHL, HB, LHB and RHB.

The regression coefficient for stature on HL was 0.46 (t-value=6.22, d.f. 149; p<0.01), which was statistically significant. In case of stature on LHL, the regression coefficient of 0.45 was also statistically significant (t-value=6.18, d.f. 149; p<0.01) and stature on RHL, the regression coefficient of 0.45 was also found statistically significant (t-value=6.16, d.f. 149; p<0.01). In case of stature on HB was taken into consideration, the regression coefficient of 0.24 was statistically significant (t-value=3.05, d.f. 149; p<0.01). In case of stature on the LHB, the regression coefficient of 0.24 was again found statistically significant (t-value=2.49, d.f. 149; p<0.01) and stature on RHB, the regression coefficient of 0.24 was found statistically significant (t-value=2.99, d.f. 149; p<0.01).

The regression coefficient for stature on HL was 0.59 (t-value=8.85, d.f.149, p<0.01). In case of stature on LHL, the regression coefficient of 0.58 was also statistically significant (t-value=8.61, d.f. 149, p<0.01) and stature on RHL, the regression coefficient of 0.58 was found statistically significant (t-value=8.82, d.f. 149, p<0.01). In case of stature on HB, the regression coefficient was 0.34 (t-value=4.44, d.f. 149, p<0.01). In case of stature on the LHB, the regression coefficient of 0.34 is again found statistically significant (t-value=4.32, d.f. 149, p<0.01) and RHB on stature, the regression coefficient of 0.33 is also found statistically significant (t-value=4.24, d.f. 149, p<0.01).

The regression coefficient for stature on HL was 0.69 (t-value=16.76, d.f. 299, p<0.01). In case of stature on LHL, the regression coefficient of 0.69 was also statistically significant (t-value=16.62, d.f. 299, p<0.01) and stature on RHL, the regression coefficient of 0.69 was found statistically significant (t-value=16.59, d.f. 299, p<0.01). In case of stature on HB, the regression coefficient was 0.64 (t-value=14.28, d.f. 299, p<0.01). In case of stature on the LHB, the regression coefficient of 0.62 is again found statistically significant (t-value=13.67, d.f. 299, p<0.01) and RHB on stature, the regression coefficient of 0.64 is also found statistically significant (t-value=14.18, d.f. 299, p<0.01).

Step-wise multiple regression equations for stature, regression coefficient and standard error on both HL and HB among the adult Karbi individuals are presented in [Table/Fig-7]. In all the cases, the regression coefficients were found statistically significant (p<0.05). It was observed that the values of the correlation coefficient R and the coefficient of determination (R²) increased as the effects of both HL and HB came into force. The regression coefficients and R² were found to be higher on the right sides (e.g., RHL and RHB) in both male and female Karbi adult individuals.

The results of step-wise multiple regression equations for stature on both HL and HB among the adult Karbi individuals adding age as an independent variable and correlation coefficient and standard errors are depicted in [Table/Fig-8]. The regression coefficients were also found statistically significant (p<0.05). It was further observed that the values of the correlation coefficient R and the coefficient of determination (R²) increased as the effects of age came into force. Similarly, the regression coefficients and R² were found to be higher on the right sides (e.g., RHL and RHB) in both male and female Karbi adult individuals when age was introduced as independent variables with HL and HB (p<0.01).

India is known for its enormous ethnic population and genetic diversities are attributed to the existence of numbers of ethnic and/or genetic elements [41]. Stature/height varies with ethnic group/populations and is determined by the genetics of an individual, geographical location, environment and climatic conditions [8,9,13,37,42]. Therefore, a need for regional/ethnic or population specific studies on stature estimations are to be emphasised since long owing to the population/ethnic variations present in different parts of the world [10,43]. The anthropometric techniques are widely utilised to formulate the biological profile and personal identification despite of the several modern techniques (e.g., computer tomography and computer assisted superimposition). It is a single most portable, universally applicable, inexpensive and non-invasive technique widely used by biological anthropologists/forensic anthropologists and adopted by medical professionals to estimate body size and proportions for the purpose of identification of the body and skeletons [7-9,13,14,16,18-20,23,29,30,37,42,44]. This technique plays a pivotal role in the field of forensic anthropology and medico-legal cases in personal identification from human remains or population investigations. The estimation of stature from the different body segments includes long bones measurements, finger and phalanges and hand and foot dimensions have shown minimum errors and high predictability [8,15,42,44,45]. Several research studies have also estimated the stature in different ethnic groups/population using different hand dimensions in India [1,19-20,26,32]. There is almost a complete absence of forensic studies on stature estimation using hand dimensions among ethnic groups/populations of Assam, Northeast India. The detailed literature search using PUBMED/Google Scholar, has yielded very few research studies on the estimation of stature and sex from the index finger, ring finger and foot among population of Eastern/North-eastern India [8,9,11,18]. Moreover, results of the present investigation indicated that the hand dimensions (e.g., HL and HB) showed statistically significant positive correlation with stature in both sexes (p<0.05). The presence of a positive linearity between anthropometric variables (e.g., HL and HB) and the height/stature facilitates to the formulation of population-specific regression equations which can be utilised for the estimation of stature of an ethnic population. An individual materials (e.g., hand) or mutilated part when recovered and brought for forensic examination can provide valuable information about the age, sex, and stature of the person [19,42,45,46]. Several studies have been undertaken to establish correlation between stature and hand dimensions (e.g., HL and HB) from different ethnic communities/population [15,18-19,20,23,25-26,28,30,32,45]. This present study explored the anthropometric data using hand dimensions (e.g., HL and HB) and stature/height using linear and step-wise multiple regression models with HL, HB and age with the indicator among Karbi adults [Table/Fig-6,7 and 8]. These formulae are applicable to the ethnic groups/population from which the data have been collected due to inherent variations in these linear dimensions which may be attributed to biological and environmental factors [20,23]. The results showed that a statistically significant sex-differences were found between sexes in LHL, RHL, LHB, RHB and stature (p<0.05) [Table/Fig-4]. The results were in conformity with two studies conducted on the hand dimension in the different ethnic/populations [20,32], but sex-difference was observed to statistically significant in HL [20]. Krishan and Sharma reported a significant difference in HB, whereas no significant difference was observed in HL between the right and left sides [20]. Hence, it may be concluded that adult Karbi males and females differed with respect to RHL and LHB (p<0.05) but the differences were not found statistically significant in stature, LHB, and RHB (p>0.05), thereby also indicating sexual dimorphism respect to hand dimension [Table/Fig-5]. A significant difference was observed in HL among Indians females, west Europeans and natives of West Indies, while no such differences were found in HB [24]. Using paired sample t-test, the result of the present study indicated that there were statistically in significant differences observed in hand dimension in HL among females and total but the t-value was found to be statistically significant among males and female individuals in respect to HL (in males), HB (both males and females) Karbi individuals (p>0.05) [Table/Fig-5].

The preponderance of existing studies observed that women have smaller body dimensions than men individuals. Several studies have been already reported the sexual dimorphism existed in the HL and HB in the both sexes [1,20,25-26,32]. The results of the present study support the fact that male individuals are genetically taller than female counterparts. Age of puberty being two years later in males as compared to females give them an additional period of physical growth and maturation [20,25]. As far as bilateral asymmetry in the HL and HB are concerned that descriptive statistics showed significant asymmetry in both HL (except in females) and HB of mean value, standard deviation are similar and differences were found to be statistically significant (p>0.05) [Table/Fig-5]. Similar results were also obtained by Krishan and Sharma [20] and Agnihotri et al., [32] for HL dimension, but their study showed significant asymmetry in HB in the studies ethnic population (p<0.05). The present study has reported significant coefficients (R) when the regression of stature on LHL, RHL, LHB and RHB and regression of HL and HB is also considered for estimation of stature in both sexes (p<0.05). Stature is thus significantly dependent on LHL, RHL, LHB and RHB (p<0.05) [Table/Fig-6]. The predictability was increased when multiple regressions were done and also when age was added as another variable. It was further observed that the values of the correlation coefficient (R) and the coefficient of determination (R²) were found to be similar as the effects of both HL and HB came into force. For unknown cases, the values of R and R² also remained appreciably higher in both sexes in linear regressions when both HL and HB are coming into force. The results of the present study may conclude that the estimation of stature with LHL, RHL, LHB and RHB is possible in adults. In step-wise multiple regression equations, where age added as a predictor variable with HL and HB that showed a positive increase both R and R² in both male and female adult Karbi individuals. These estimations can, thus, play a prime role in case of identification cases involving the indigenous Karbi individuals and may apply for estimation in different ethnic communities in India [Table/Fig-6,7 and 8]. Present investigation has focused on the estimation of stature from population-specific equations (e.g., linear and multiple linear regression models) using hand dimension and the mean differences between the mean stature with estimated stature was observed to be statistically not significant (p>0.05) [Table/Fig-9].

The anthropometric measurements on hand dimensions in the present investigation were obtained from living individuals may not provide accurate estimation of stature on skeletal remains. This could be the one of the limitation of the present investigation; therefore similar equations could be developed and/or tested using skeletal remains to estimate the stature in population.

It is worthwhile to mention here that the present cross-sectional investigation is a pioneering study among indigenous Karbi tribal individuals of Karbi Anglong, Assam Northeast India. The present study is successful in estimating stature from LHL, RHL, LHB and RHB. It has also successfully observed the possible relationships between the stature and hand dimensions as well as estimated bilateral differences and sex-differences with respect to anthropometric variables (e.g., HL and HB) on hand dimensions. Furthermore, similar studies are also recommended for ethnic/population specific equation and/or utilise and validation of equations to estimate the stature from HL and HB. The present study equations may be used to obtain the approximate height of an individual/population in case of the practical problems in order to obtain direct measurements of the body remain (e.g., dismembered fragments or amputee or accidental cases).