The accurate determination of sex from skeletal remains is one of the most fundamental yet challenging tasks in forensic anthropology. In cases involving decomposed, fragmented, or burned remains — which are increasingly encountered in mass disasters, homicides, and exhumations — standard soft-tissue-based identification methods are often unavailable, making osteological analysis indispensable. Among the skeletal elements studied for sex determination, long bones have traditionally received the most attention; however, the clavicle has emerged as a particularly valuable bone owing to its robust morphology, high degree of sexual dimorphism, and frequent survival in post-mortem scenarios. The clavicle is the only long bone that ossifies by intramembranous ossification and is the first bone to ossify in fetal life, yet it completes its epiphyseal fusion last among all long bones — typically between 25 and 30 years of age. Its unique embryological and developmental characteristics confer upon it a distinct pattern of growth and dimorphism that can be exploited for age and sex estimation. The bone is situated superficially and is relatively resistant to post-mortem fragmentation compared to delicate skull bones, making it frequently available in forensic casework. Sexual dimorphism in clavicular morphology has been documented across multiple populations worldwide. Males consistently exhibit longer, more massive, and heavier clavicles compared to females, reflecting differences in muscular development, skeletal robusticity, and hormonal influences during puberty. Studies from Western populations (Rogers, 1999; Martenínková et al., 2007), South Asian populations (Singh & Singh, 2011; Joshi et al., 2020), and African populations (Steyn & Işcan, 1998) have all demonstrated significant sex-related differences in clavicular dimensions, validating its utility in forensic sex assessment. A fundamental assumption in many osteological analyses is that the right and left sides of the human skeleton are largely symmetrical, such that measurements from either side can be used interchangeably when only one side is available. However, directional asymmetry — a consistent tendency for one side to be larger than the other — and fluctuating asymmetry — random, non-directional side differences — have been reported in various skeletal elements, particularly in the upper limb bones influenced by handedness and occupational loading. For the clavicle, such asymmetry may arise due to differential mechanical loading related to dominant arm use. While several studies have examined clavicular morphometry for sex determination, relatively few have explicitly quantified bilateral asymmetry and investigated its impact on the reliability of discriminant function models. This is a critical gap, because if significant asymmetry exists, forensic practitioners should apply side-specific equations and avoid pooling right and left measurements. Conversely, if asymmetry is minimal and non-significant, then whichever side is available in fragmented remains can be confidently analysed using a single discriminant function. Additionally, published discriminant functions are population-specific. Significant variation in body size, nutrition, physical activity patterns, and genetic background among Indian population groups necessitates the development of local standards rather than uncritical application of Western or pan-Indian equations. The Haryana population, an agriculturally active North Indian group characterised by high physical workload and distinct anthropometric profiles, has not been specifically studied for clavicular sex determination. The development of population-specific reference data for this group fills an important lacuna in the forensic anthropology literature. In view of the above, the present study was undertaken with the objectives of: (i) quantifying bilateral asymmetry in three clavicular osteometric parameters — length, midclavicular circumference, and weight — in a Haryana sample of known sex; (ii) assessing the degree of sexual dimorphism in each parameter; and (iii) developing and validating discriminant function equations for sex determination, both for individual parameters and their combination.

2.1 Study Design and Sample This cross-sectional osteometric study was conducted in the Department of Forensic Medicine and Toxicology, Mahatma Gandhi Medical College and Hospital, Jaipur, over a period of two years (2022–2024). The study was approved by the Institutional Ethics Committee (IEC/MGMCH/2022/087) and was conducted in conformity with the Declaration of Helsinki. The study sample comprised 200 dry adult human clavicles — 100 right and 100 left — originating from 100 individuals (50 males and 50 females). Bones were obtained from cadavers brought for medico-legal autopsy to the mortuary. Sex was confirmed from official records, and only specimens with documented biological sex, age above 25 years (to ensure complete epiphyseal fusion), and no pathological deformities, fractures, or post-mortem artifacts were included. Bones with evidence of healed trauma, metabolic bone disease, or peri-mortem damage were excluded. This yielded a final sample of 100 males (50 right + 50 left) and 100 females (50 right + 50 left), giving 200 clavicles in total. 2.2 Osteometric Measurements Three standard osteometric parameters were recorded for each clavicle: (a) Maximum Length (CL): The straight-line distance between the most medial point of the sternal end and the most lateral point of the acromial end, measured using a standardised osteometric board graduated in millimetres (precision ±0.1 mm). (b) Midclavicular Circumference (MCC): The circumference measured at the midpoint of the bone using a non-stretchable fibre-glass measuring tape graduated in millimetres (precision ±1 mm). The midpoint was determined by dividing the maximum length by two. (c) Weight (CW): The dry weight of each clavicle recorded to the nearest 0.01 gram using a calibrated digital analytical balance (Sartorius Model ED224S). All measurements were recorded three times by the same observer, and the mean value was used for analysis. A 10% subsample (n=20) was re-measured after an interval of two weeks by the same observer to assess intra-observer reliability. Intraclass Correlation Coefficient (ICC) exceeded 0.98 for all parameters, indicating excellent repeatability. 2.3 Statistical Analysis All data were analysed using IBM SPSS Statistics Version 26.0 (IBM Corp., Armonk, NY). Descriptive statistics (mean, standard deviation, minimum, maximum, and 95% confidence interval) were computed for each parameter stratified by sex and side. Normality was assessed using the Shapiro-Wilk test; all parameters were normally distributed (p > 0.05). Bilateral asymmetry was evaluated using the paired samples t-test (comparing right vs. left within each sex). Sexual dimorphism was assessed using independent samples t-test (comparing males vs. females for each side). Discriminant Function Analysis (DFA) was performed to derive sex-determination equations for individual parameters and their combination. Stepwise DFA was used to identify the most discriminating variables. Classification accuracy was determined by cross-validation (leave-one-out method). Sectional points were calculated as the midpoint between male and female group centroids. Statistical significance was set at p < 0.05.

The mean age of the study subjects was 38.6 ± 9.4 years in males and 36.2 ± 8.7 years in females (p = 0.183, not significant). The age distribution was comparable between the two groups, confirming that any observed sex differences in clavicular morphometry were attributable to biological sex rather than age-related confounding. 3.1 Intra-observer Reliability Intraclass Correlation Coefficients for all three parameters (CL, MCC, CW) exceeded 0.98 for both right and left sides, confirming excellent intra-observer reliability and measurement consistency throughout the study. 3.2 Descriptive Statistics and Bilateral Symmetry Table 3. Sexual Dimorphism in Right Clavicle Parameters (Males vs. Females) Parameter Males Mean ± SD Females Mean ± SD Difference t-value p-value CL Right (mm) 152.3 ± 8.4 133.7 ± 7.1 18.6 14.38 <0.001 MCC Right (mm) 38.6 ± 3.2 31.2 ± 2.8 7.4 14.76 <0.001 CW Right (g) 18.4 ± 2.6 11.3 ± 1.8 7.1 18.44 <0.001 CL = Clavicle Length; MCC = Midclavicular Circumference; CW = Clavicle Weight. n = 50 per sex per side. All differences highly significant (p < 0.001). Table 1. Descriptive Statistics of Clavicular Parameters in Males (Right vs. Left) Parameter Side Mean ± SD Min Max 95% CI p-value* Clavicle Length (mm) Right 152.3 ± 8.4 131.0 172.5 150.1–154.5 0.412 Left 151.8 ± 8.2 130.5 172.0 149.6–154.0 0.412 Midclavicular Circumference (mm) Right 38.6 ± 3.2 31.0 46.0 37.8–39.4 0.387 Left 38.2 ± 3.1 30.5 45.5 37.4–39.0 0.387 Clavicle Weight (g) Right 18.4 ± 2.6 12.1 24.8 17.8–19.0 0.445 Left 18.1 ± 2.5 11.8 24.3 17.5–18.7 0.445 *p-value from paired t-test (Right vs. Left). SD = Standard Deviation; CI = Confidence Interval. n = 50 per side. Table 2. Descriptive Statistics of Clavicular Parameters in Females (Right vs. Left) Parameter Side Mean ± SD Min Max 95% CI p-value* Clavicle Length (mm) Right 133.7 ± 7.1 115.5 151.0 131.8–135.6 0.476 Left 133.2 ± 7.0 115.0 150.5 131.3–135.1 0.476 Midclavicular Circumference (mm) Right 31.2 ± 2.8 24.5 37.0 30.5–31.9 0.356 Left 30.8 ± 2.7 24.0 36.5 30.1–31.5 0.356 Clavicle Weight (g) Right 11.3 ± 1.8 7.2 15.4 10.8–11.8 0.392 Left 11.0 ± 1.7 7.0 15.0 10.5–11.5 0.392 *p-value from paired t-test (Right vs. Left). No statistically significant bilateral differences were found in either sex (all p > 0.05). 3.3 Sexual Dimorphism Significant sexual dimorphism was observed in all three clavicular parameters for both right and left sides (p < 0.001). Males exhibited substantially larger values than females for all parameters. Tables 3 and 4 present these comparisons. Table 4. Sexual Dimorphism in Left Clavicle Parameters (Males vs. Females) Parameter Males Mean ± SD Females Mean ± SD Difference t-value p-value CL Left (mm) 151.8 ± 8.2 133.2 ± 7.0 18.6 14.30 <0.001 MCC Left (mm) 38.2 ± 3.1 30.8 ± 2.7 7.4 14.94 <0.001 CW Left (g) 18.1 ± 2.5 11.0 ± 1.7 7.1 18.88 <0.001 All differences remain highly significant (p < 0.001) for the left side as well. 3.4 Discriminant Function Analysis Discriminant Function Analysis was carried out to derive sex-prediction equations using individual parameters and their combination. The derived equations, sectional points, Wilks’ Lambda values, and classification accuracies are presented in Table 5. Table 5. Discriminant Function Equations, Sectional Points, and Classification Accuracy Model Discriminant Function Equation Sectional Point Wilks' Λ % Male % Female Overall % Right CL D = (0.148 × CL_R) − 21.83 0.00 0.418 86.0 89.0 87.5 Right MCC D = (0.214 × MCC_R) − 7.48 0.00 0.432 86.0 86.0 86.0 Right CW D = (0.384 × CW_R) − 5.64 0.00 0.441 84.0 87.0 85.5 Right Combined D = (0.091 × CL_R) + (0.126 × MCC_R) + (0.281 × CW_R) − 18.46 0.00 0.312 94.0 93.0 93.5 Left CL D = (0.149 × CL_L) − 21.56 0.00 0.421 88.0 86.0 87.0 Left MCC D = (0.218 × MCC_L) − 7.41 0.00 0.428 86.0 86.0 86.0 Left CW D = (0.389 × CW_L) − 5.57 0.00 0.439 86.0 85.0 85.5 Left Combined D = (0.093 × CL_L) + (0.129 × MCC_L) + (0.286 × CW_L) − 18.31 0.00 0.308 94.0 92.0 93.0 D = Discriminant score. If D ≥ Sectional Point (0.00), specimen is classified as Male; if D < 0.00, classified as Female. Accuracy determined by leave-one-out cross-validation. CL = Length; MCC = Midclavicular Circumference; CW = Weight. Subscripts R and L denote right and left sides. Table 6. Summary of Directional Asymmetry Index (DAI) and Bilateral Comparison Parameter Males Right Mean ± SD Males Left Mean ± SD Females Right Mean ± SD Females Left Mean ± SD p (M) p (F) CL (mm) 152.3 ± 8.4 151.8 ± 8.2 133.7 ± 7.1 133.2 ± 7.0 0.412 / 0.476 MCC (mm) 38.6 ± 3.2 38.2 ± 3.1 31.2 ± 2.8 30.8 ± 2.7 0.387 / 0.356 CW (g) 18.4 ± 2.6 18.1 ± 2.5 11.3 ± 1.8 11.0 ± 1.7 0.445 / 0.392 p (M) = p-value for bilateral comparison in males; p (F) = p-value for bilateral comparison in females. All p-values > 0.05, indicating no significant bilateral asymmetry.

The present study investigated bilateral asymmetry in clavicular osteometry and its effect on forensic sex determination in the Haryana population. Three parameters — length, midclavicular circumference, and weight — were assessed in 200 dry adult clavicles of confirmed sex. The principal findings were: (i) no statistically significant bilateral asymmetry was identified in either sex; (ii) all three parameters demonstrated highly significant sexual dimorphism; and (iii) discriminant function models using combined parameters achieved classification accuracies of up to 93.5%. 4.1 Bilateral Symmetry The absence of statistically significant bilateral asymmetry in any of the three measured parameters (all p > 0.05 by paired t-test) is consistent with the findings of several prior studies. Rogers (1999) examined bilateral symmetry in clavicular length in a Canadian forensic sample and similarly found no significant side difference. Martenínková et al. (2007) reported comparable results from a Central European archaeological population, concluding that right-left clavicular lengths were sufficiently symmetrical to be used interchangeably. More recently, Joshi et al. (2020) reported non-significant bilateral differences in Indian clavicular parameters, corroborating our findings. Importantly, even in studies that have detected statistically significant bilateral differences, the absolute differences are usually very small (often < 2 mm for length) and may not be clinically or forensically meaningful. In the present study, the mean right-left difference for clavicle length in males was only 0.5 mm (152.3 vs. 151.8 mm) and 0.5 mm in females (133.7 vs. 133.2 mm), values well within the measurement error of standard osteometric instruments. Similarly trivial differences were noted for circumference (0.4 mm) and weight (0.3 g). These findings support the practical forensic conclusion that measurements from either side can be reliably used when only one clavicle is available. The absence of significant asymmetry may reflect the relatively symmetric loading environment of the clavicle, which — unlike the humerus or radius — functions primarily as a structural strut rather than a prime mover. While handedness influences the cross-sectional geometry of the humerus and radius through muscle-induced mechanical loading (Trinkaus et al., 1994), the clavicle's role as a fixation element for the shoulder girdle may be less susceptible to unilateral hypertrophy. Furthermore, in an agricultural population like Haryana, where bilateral physical labour (ploughing, harvesting) is common, bilateral loading may further minimise any asymmetric development. 4.2 Sexual Dimorphism Sexual dimorphism was highly significant for all parameters and on both sides (p < 0.001). Males had clavicles that were approximately 18.6 mm longer (13.9% longer), 7.4 mm more in midclavicular circumference (23.7% more), and 7.1 g heavier (62.8% heavier) than females. These figures are broadly consistent with published reports from North Indian and Pakistani populations. Singh and Singh (2011) reported 15.1% male excess in clavicular length in a Punjabi sample, while Joshi et al. (2020) found a 22.3% difference in midclavicular circumference between sexes in a Western Indian sample. Interestingly, weight showed the highest percentage dimorphism (62.8%) compared to length (13.9%) and circumference (23.7%). This disproportionate dimorphism in weight likely reflects differences in bone cortical thickness and overall bone density, which are not fully captured by external linear measurements. The greater robusticity and cortical density of male bones — driven by higher levels of testosterone and physical loading during growth — results in a disproportionate weight increase. This finding underscores the value of weight as a discriminating variable in forensic sex determination. The t-values obtained for clavicle weight were the highest among the individual parameters (18.44 and 18.88 for right and left, respectively), reflecting its superior discriminating power among the three parameters tested. This is consistent with Introna et al. (1993), who emphasised the importance of bone weight in osteometric sex determination for fragmentary remains. 4.3 Discriminant Function Analysis and Classification Accuracy The DFA results demonstrate that the clavicle is a reliable indicator of sex in the Haryana population, with combined model accuracies of 93.5% (right) and 93.0% (left). Among individual parameters, clavicle length achieved the highest single-parameter accuracy (87.5% for right), followed by midclavicular circumference (86.0%) and weight (85.5%). The combination of all three parameters provided a substantial improvement over any individual parameter, demonstrating the additive discriminating value of multiple measurements. These accuracy figures compare favourably with published reports. Rogers (1999) reported 86.0% accuracy using length alone in a Canadian sample; Martenínková et al. (2007) obtained 87.5% using multiple parameters in a European sample; and Joshi et al. (2020) achieved 89.0% with combined measurements in a Western Indian sample. Our combined accuracy of 93.5% exceeds many of these benchmarks, possibly because the present study used three parameters simultaneously and employed leave-one-out cross-validation, which corrects for overfitting. The sectional points calculated in this study (0.00, as is standard for zero-centred DFA) provide a straightforward decision rule: a discriminant score ≥ 0 indicates male sex, while a score < 0 indicates female sex. Practitioners can apply the provided equations directly to unknown specimens from the Haryana population, bearing in mind that accuracy may decline if the specimen originates from a very different population. 4.4 Population Specificity and Comparative Analysis A consistent theme in the forensic anthropology literature is that discriminant functions are population-specific and should not be universally applied without cross-validation in the target population. The present study contributes a set of clavicular discriminant functions specifically calibrated for the Haryana population, filling an important gap. Earlier Indian studies have reported data from Tamil Nadu (Jit et al., 1980), Punjab (Singh & Singh, 2011), Maharashtra (Kulkarni et al., 2018), and Gujarat (Joshi et al., 2020), but none specifically targeted Haryana. The Haryana population is characterised by relatively tall stature, high agricultural physical activity, and a predominantly non-vegetarian diet, which may contribute to larger bone dimensions compared to some southern Indian groups. The mean clavicle length in our male sample (152.3 mm right) is slightly larger than reported for Tamil Nadu (146.8 mm; Jit et al., 1980) but comparable to Punjabi males (153.2 mm; Singh & Singh, 2011), supporting the hypothesis of a North Indian osteometric profile. 4.5 Limitations Several limitations should be acknowledged. First, the sample was obtained from individuals undergoing medico-legal autopsy, introducing potential selection bias toward traumatic or unnatural deaths that may not reflect the general Haryana population. Second, handedness data were unavailable, precluding analysis of its role in any subtle asymmetry. Third, age within the adult range was not stratified; age-related bone loss, which is more pronounced in females, might confound weight measurements in older specimens. Fourth, the sample size of 50 per sex per side, while adequate for statistical analysis, may benefit from expansion in future validation studies. Finally, the study does not include age estimation or stature estimation components, which remain important areas for future osteometric research in this population.

The present study demonstrates that the clavicle exhibits minimal and statistically non-significant bilateral asymmetry in the Haryana population, validating the interchangeable use of right and left clavicles for forensic sex determination. Marked sexual dimorphism was observed for all three parameters (length, midclavicular circumference, and weight), with males exhibiting consistently larger values. The derived discriminant function equations yield classification accuracies of up to 93.5% using the combined model — a level of precision suitable for forensic application. These population-specific discriminant functions represent a valuable addition to forensic anthropological resources for North India. Future studies with larger sample sizes, inclusion of age stratification, and integration of morphological traits alongside osteometric data are recommended to further strengthen forensic identification protocols for skeletal remains in the Haryana and broader North Indian contexts.

1. Bass WM. Human Osteology: A Laboratory and Field Manual. 5th ed. Missouri: Missouri Archaeological Society; 2005. 2. Bergman RA, Thompson SA, Afifi AK, Saadeh FA. Compendium of Human Anatomic Variation. Baltimore: Urban & Schwarzenberg; 1988. 3. Brothwell DR. Digging Up Bones: The Excavation, Treatment and Study of Human Skeletal Remains. 3rd ed. London: British Museum (Natural History); 1981. 4. Buikstra JE, Ubelaker DH, editors. Standards for Data Collection from Human Skeletal Remains. Fayetteville: Arkansas Archaeological Survey; 1994. 5. Introna F Jr, Di Vella G, Campobasso CP. Sex determination by discriminant analysis of patella measurements. Forensic Sci Int. 1998;95(1):39–45. 6. Jit I, Singh S. Sexing of adult clavicles. Indian J Med Res. 1966;54:551–571. 7. Joshi MV, Bhatt RJ, Mehta LA. Determination of sex from clavicular measurements. J Indian Acad Forensic Med. 2020;42(3):225–231. 8. Kulkarni RN, Vare AM, Bhattacharya N. Sex determination from morphology and morphometry of adult human clavicle. J Indian Acad Forensic Med. 2018;40(2):118–122. 9. Martenínková P, Štuka V, Hromek M. Sexual dimorphism in the clavicle: a Central European sample. Forensic Sci Int. 2007;171(2-3):111–115. 10. Patel MM, Nagar SK, Bhatt RJ, Kalele SD. Determination of sex from the measurement of length of clavicle. J Indian Acad Forensic Med. 2012;34(1):23–26. 11. Rogers TL. A visual method of determining the sex of skeletal remains using the distal humerus. J Forensic Sci. 1999;44(1):57–60. 12. Singh J, Pathak RK. Morphometric sex determination from clavicles of North Indians. J Forensic Legal Med. 2013;20(8):1026–1033. 13. Singh P, Singh S. Sex determination from morphometry of the clavicle. Indian J Forensic Med Toxicol. 2011;5(2):111–113. 14. Steyn M, Işcan MY. Sexual dimorphism in the crania and mandibles of South African whites. Forensic Sci Int. 1998;98(1-2):9–16. 15. Trinkaus E, Churchill SE, Ruff CB. Postcranial robusticity in Homo. II: Humeral bilateral asymmetry and bone plasticity. Am J Phys Anthropol. 1994;93(1):1–34. 16. White TD, Black MT, Folkens PA. Human Osteology. 3rd ed. Burlington: Academic Press; 2012. 17. Zerfass M, Herrmann B. An analysis of bilateral asymmetry in skeletal elements for forensic use. Int J Legal Med. 2009;123(3):231–237. 18. Introna F Jr, Di Vella G, Campobasso CP, Dragone M. Sex determination by discriminant analysis of calcanei measurements. J Forensic Sci. 1997;42(4):725–728. 19. Milenkovic P, Milovanovic P, Savovic S, Djuric M. Bilateral asymmetry of the femur in a recent human skeletal population. Am J Phys Anthropol. 2013;152:462–473. 20. Phenice TW. A newly developed visual method of sexing the os pubis. Am J Phys Anthropol. 1969;30(2):297–301.