Older-Adult Age from Sternal Rib Ends and Cranial Sutures

M. Yasar Iscan and Susan Loth began their systematic study of age-related changes at the sternal rib end in 1984, motivated by the observation that the costal cartilage end of the rib undergoes predictable progressive changes that had been noted in the pathology literature but never systematically scored for forensic purposes. Their foundational 1984 paper in the Journal of Forensic Sciences described the method and reference data for white males; the 1985 companion paper in the same journal provided reference data for white females; and subsequent papers extended the method to black males and black females, with some population-specific differences in phase timing.

The method focuses on the sternal (anterior) end of the right fourth rib. The selection of the right fourth rib was based on Iscan and Loth's observation that this rib showed the most consistent and progressive age-related changes and had the least individual variation compared to adjacent ribs. The right side is preferred over the left because left-sided ribs have been reported to show slightly different change rates linked to cardiac loading, though in practice the left fourth rib is used when the right is absent or damaged.

The sternal end is examined on the fresh, dry, or macerated rib. The key morphological features observed are the shape and depth of the pit at the sternal end (the invagination of the costochondral articulation surface), the texture of the pit walls (smooth versus porous versus irregular), the thickness of the pit walls, the presence and character of any peripheral bony projections or spurs, and the overall structural integrity of the sternal end. These observations are combined to assign one of nine phases (Phases 0 through 8, with Phase 0 representing the youngest morphology).

The method depends on trained visual judgment, which introduces inter-observer variability. Studies assessing inter-rater reliability for the Iscan-Loth method have reported kappa values ranging from 0.6 to 0.8 for experienced observers, indicating moderate to good but not perfect agreement. Training against the original cast sets (which were produced and distributed by Iscan and Loth from the Florida Atlantic collection) substantially improves reliability.

Phase 0 describes the sternal rib end of a young individual, typically in the late teens or early twenties, showing a flat or slightly billowing surface with dense, smooth bone and no pit formation. The surface has a uniform texture and no peripheral projections.

Phase 1 shows the beginning of pit formation: a shallow central indentation surrounded by still-dense, relatively smooth bone. The pit walls are flat or slightly irregular, the bony margins remain well-defined, and no projections are present. This phase corresponds to the approximately 20 to 23 year range in males.

Phase 2 shows a deepening pit with smooth or finely granular walls and a rounded rim. The bony structure remains largely intact but the pit has become more clearly defined. In males, approximately 22 to 28 years.

Phase 3 shows a moderate pit depth with the walls beginning to show porosity. The rim is becoming irregular. Small peripheral projections may appear at the superior or inferior margin. In males, approximately 25 to 37 years.

Phase 4 shows a deeper pit with clearly porous walls, an irregular rim with multiple peripheral projections, and overall moderate structural change. In males, approximately 28 to 45 years.

Phase 5 shows a moderately deep to deep pit with porous, coarsely irregular walls, a clearly irregular rim with multiple bony projections, and evidence of structural thinning. In males, approximately 35 to 55 years.

Phase 6 shows a deep, irregular pit with thin, porous walls and a markedly irregular rim. Peripheral projections are prominent. Overall structural integrity is beginning to compromise. In males, approximately 40 to 65 years.

Phase 7 shows a very deep, irregular pit with extremely thin and porous walls. The bony structure is fragile and extensively remodeled. In males, approximately 50 to 70 years.

Phase 8 is the terminal phase, showing complete breakdown of the pit architecture: the sternal end is open, irregular, and extensively porous, with thin wall remnants and extensive peripheral projections. In males, this phase is consistent with ages above approximately 55 years with no useful upper bound. In females, the phase timing parallels the male sequence but each phase tends to appear somewhat earlier on average, with Phase 6 and beyond beginning to appear from approximately 40 years.

The Iscan-Loth sternal rib end reference data were derived from white American males and females (the original 1984 and 1985 papers) from the Maricopa County Medical Examiner's Office (Arizona) and the Palm Beach County Medical Examiner's Office (Florida). The subsequent papers on black American males and females (1985 and 1986) found somewhat different phase timing: black males tended to show Phase 5 and above changes at somewhat earlier ages than white males, on average by approximately two to five years per phase in the middle adult range. This population difference is modest but not trivial for forensic age estimation purposes.

Applying the white male or white female Iscan-Loth tables to individuals of other ancestral backgrounds introduces unknown systematic bias. For South Asian, East Asian, or African populations, validation studies are sparse. A South African study by Oettlé and Steyn (2000, Journal of Forensic Sciences, vol. 45, no. 5, pp. 1071-1079) found that the Iscan-Loth method performed acceptably for South African black females but showed larger age underestimation in some phase categories. Indian forensic literature on sternal rib end age estimation is limited to small-sample studies from individual AIIMS departments, with insufficient sample sizes to support population-specific correction factors.

In the United Kingdom, the method is used as one component of a multi-method assessment. The Forensic Science Regulator's Codes of Practice do not mandate a specific method but require that the practitioner disclose the population basis of the reference data and the known error rates. Several UK forensic anthropologists have noted in published testimony summaries that they apply the Iscan-Loth phases but cite the full published range and add a caveat when the individual's ancestral background may differ from the reference sample.

In the United States, SWGANTH (Scientific Working Group for Forensic Anthropology) and the OSAC Anthropology Subcommittee have noted in published standards that population-specific validation data are needed for the sternal rib end method, and that the method should not be applied without acknowledging its reference-population limitations.

Richard Meindl and C. Owen Lovejoy published the cranial suture closure scoring system in the American Journal of Physical Anthropology in 1985, derived from the Hamann-Todd Collection (the documented autopsy collection at the Cleveland Museum of Natural History, with approximately 3,000 individuals whose ages at death range from birth to over 90 years). The Hamann-Todd Collection, assembled primarily by T. Wingate Todd between 1912 and 1938, is one of the most important documented reference skeletal collections in North American forensic anthropology.

The Meindl-Lovejoy method scores the closure status of ten ectocranial suture sites (observable from the outer surface of the skull) in two composite systems. The vault system scores five sites: midlambdoid (lambda), lambda, obelion, anterior sagittal, and bregma. The lateral-anterior system scores five additional sites: midcoronal, pterion, sphenofrontal, inferior sphenotemporal, and superior sphenotemporal. Each site is scored on a four-point scale: 0 (open, no closure visible), 1 (minimal closure, less than 50 per cent), 2 (significant closure, 50 per cent or more), 3 (complete closure). The scores for the five sites within each system are summed to give a composite score.

The composite score for each system (vault: 0 to 15; lateral-anterior: 0 to 15) is converted to a mean age and range using published regression tables. Importantly, Meindl and Lovejoy recommend using the vault and lateral-anterior systems independently and combining the two estimates, rather than treating either alone as definitive.

A fundamental limitation of the cranial suture method, acknowledged by Meindl and Lovejoy and confirmed by multiple subsequent validation studies, is its large error rate. The coefficient of determination (R-squared) for suture closure score versus chronological age is approximately 0.3 to 0.5, meaning that suture closure score explains only 30 to 50 per cent of the variance in age. The standard deviation around the mean age estimate for any given composite score is typically 10 to 15 years, producing 95 per cent confidence intervals that span 25 to 35 years. This accuracy is substantially worse than the pubic symphysis or sternal rib end for the same age range.

A consistent finding in the suture closure literature is that endocranial (inner surface) suture closure proceeds slightly earlier than ectocranial (outer surface) closure, and that endocranial scores show somewhat less inter-individual variability at equivalent ages. The practical challenge is that endocranial scoring requires physical access to the inner cranial surface, which either means a macerated dry skull opened at the calvarium or a CT cross-section through the suture. For a dry forensic skeletal case where the calvarium is intact, this is straightforward. For a recent-death case or a case involving a living person, it requires CT imaging.

The reliability hierarchy for suture sites is site-specific, not just endo-versus-ecto. The sagittal suture (particularly the obelion point) and the coronal suture sites show more progressive and predictable closure than the lambdoid suture, which is notorious for its high inter-individual variability, including a subset of individuals (approximately 10 to 15 per cent of populations) who show complete lambdoid suture fusion before age 40 and another subset who retain partially open lambdoid sutures past age 70. The lambdoid site has the lowest weight in the Meindl-Lovejoy scoring and should be interpreted with caution when it is discordant with the other sites.

A specific and confounding phenomenon is craniosynostosis: the premature pathological fusion of cranial sutures in infants and young children, which occurs in approximately 1 in 2,500 live births and involves various sutures depending on the syndrome. Forensic anthropologists must exclude craniosynostosis (which leaves characteristic asymmetries and compensatory skull shape changes) before interpreting premature sutural closure in a sub-adult or young adult skull as an age indicator.

Inca bones (accessory ossicles in the lambdoid suture, also called Wormian bones) can interfere with suture scoring by interrupting the suture's closure pattern. Their presence should be noted and the affected site scored separately or excluded from the composite if the interruption is extensive.

The ADBOU program (Anthropological Data Base Odense University, developed by Jesper Boldsen, George Milner, Lyle Konigsberg, and colleagues at the University of Southern Denmark and Penn State University) implements a Bayesian transition-analysis framework for skeletal age estimation. The first major publication appeared in the American Journal of Physical Anthropology in 2002. The software has been updated through multiple versions; the most widely cited versions used in forensic casework as of 2025 are ADBOU 2.0 and later releases. The same Bayesian framework is applied in war-crimes mass-grave identification work where posterior age distributions are integrated with DNA and documentary evidence.

The transition-analysis framework treats each morphological feature (a Suchey-Brooks phase, a Buckberry-Chamberlain composite score, a Meindl-Lovejoy vault score, an Iscan-Loth rib phase) not as a direct predictor of age but as an observable that has a certain probability of occurring at each age. The transition parameters (the ages at which 50 per cent and 95 per cent of the reference population have made each transition from one phase to the next) are estimated from the reference data by maximum likelihood. Given the observed feature values for a case individual, the program computes the likelihood of observing those features at each possible age, multiplies by the prior probability distribution over age (which the analyst can set based on contextual information about the population from which the case is drawn), and normalises to produce a posterior probability distribution.

The output is a probability density over age from which the analyst can read off the mode (most probable age), the credible interval (the range within which 90 per cent of the posterior probability falls), and any desired quantile. The Bayesian framework explicitly incorporates the analyst's prior beliefs about the case: a mass-grave case where documentary context suggests the victims were adults aged 20 to 50 can use an informative prior that differs from the prior used for an unidentified single burial where age is genuinely unknown.

ADBOU offers three practical advantages over informal phase reconciliation. First, it enables formal combination of multiple methods without requiring the analyst to informally "average" or "reconcile" conflicting phase assessments; the likelihood contributions of each indicator are multiplied and the combined posterior automatically resolves any apparent conflicts. Second, it provides uncertainty quantification that is calibrated against reference data rather than relying on the analyst's informal judgment about how wide a range to report. Third, it allows the prior to be stated explicitly and varied in sensitivity analyses, which is appropriate for transparent court testimony.

The multi-method composite approach begins with an inventory of what age indicators are available and in what condition. A case with an intact pelvis (pubic symphysis plus auricular surface), an intact fourth rib bilaterally, and an intact calvarium has four independent age indicators. A case where only the cranium survives has at best the cranial suture closure method and possibly the dental wear rate as supplementary information.

The composite reasoning proceeds as follows. Each indicator is scored and converted to its age distribution using published tables or ADBOU software. The overlapping region of all indicator-specific age distributions defines the initial composite estimate. Indicators that are clearly discordant with all others (for example, a sternal rib end suggesting Phase 2, consistent with approximately 22 to 28 years, while the pubic symphysis shows Phase 5 consistent with 28 to 78 years) are examined for pathological explanations: localized trauma, occupational or recreational loading, inflammatory arthritis, parity effects, or other conditions that can accelerate or retard a specific indicator's degeneration.

The final reported estimate is a biologically defensible age range, accompanied by a "most probable age" or modal estimate, with an explicit statement of the methods used, their reference populations, and their documented accuracy limitations. A representative multi-method statement for an older adult skeleton might read: "Assessment of the right pubic symphysis (Suchey-Brooks Phase 5), the right auricular surface (Buckberry-Chamberlain composite score 11), the right fourth rib sternal end (Iscan-Loth Phase 5), and the ectocranial vault sutures (Meindl-Lovejoy vault composite 8) yields consistent findings. Combining these four indicators by transition analysis (ADBOU 2.0 with a uniform adult prior), the posterior credible interval at the 90 per cent level is 42 to 67 years, with a modal estimate of approximately 53 years."

1. Inventory available indicators
   Document which skeletal elements are present and assessable: pubic symphysis (Suchey-Brooks), auricular surface (Buckberry-Chamberlain or Lovejoy), fourth rib sternal end (Iscan-Loth), cranial vault sutures (Meindl-Lovejoy vault and lateral-anterior), and any supplementary indicators such as dental wear, osteon density (histological), or vertebral osteophyte scoring.
2. Score each indicator independently
   Score each available indicator according to its published criteria, ideally without looking at the estimates from other indicators to avoid anchoring bias. Record the phase or composite score for each method separately.
3. Convert each to an age distribution
   Apply the published phase-to-age tables (or ADBOU transition parameters) to each scored indicator. Record the mean, standard deviation, and full range. Note any population differences between the reference sample and the case individual's likely ancestry.
4. Check for discordance
   If two or more indicators give substantially non-overlapping age distributions, investigate pathological or taphonomic explanations before forcing a composite. Document any discordance and its likely explanation in the case file.
5. Combine using ADBOU or formal overlap
   If ADBOU is available, enter all scored indicators and run the transition analysis with an appropriate prior. If ADBOU is not available, identify the overlapping age range across all indicator-specific distributions and report that range as the composite estimate.
6. State the composite estimate with full uncertainty
   Report the composite age range and the modal (most probable) age, citing each method and its reference. Explicitly state the known population limitations of each reference and the over-50 ceiling problem if the estimate extends into the older adult range.

The admissibility of forensic skeletal age opinions depends on the jurisdiction's standards for expert evidence. Three frameworks are relevant to the work of forensic anthropologists operating internationally.

In England and Wales, the Criminal Procedure Rules (CPR 19) require expert witnesses to help the court on matters within their expertise, to state the substance of all material facts and matters that affect the opinions, and to state when their opinion is a range of views rather than a single answer. The Forensic Science Regulator's Codes of Practice and Conduct (2022 edition) additionally require that the method's validated accuracy and population basis be documented. The BSA 2023 (Forensic Science International guidance on age assessment published in 2023, referenced by the UK Crown Prosecution Service) specifically notes that age estimates from skeletal indicators must be presented with their full uncertainty, that the over-50 ceiling must be disclosed, and that the examiner must disclose any population mismatch between the reference data and the case individual.

In the United States, expert testimony in federal court is governed by Rule 702 of the Federal Rules of Evidence and the Daubert standard (Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 1993). Under Daubert, the trial court acts as gatekeeper and assesses whether the methodology is scientifically valid, whether it has been tested, and whether the error rate is known and acceptable. For Suchey-Brooks, Iscan-Loth, and Meindl-Lovejoy, the methods are published in peer-reviewed journals, have been tested in multiple validation studies, and have known (if wide) error rates; they generally pass Daubert scrutiny when properly presented. An expert who compresses the age range without justification may fail Daubert on the grounds that the stated error rate does not reflect the method's actual performance.

At the International Criminal Tribunal for the former Yugoslavia (ICTY) and the International Criminal Court (ICC), expert evidence is assessed under rules specific to each tribunal, but in practice the standards require scientific validity, transparent methodology, and honest disclosure of uncertainty. Forensic anthropologists including Karen Burns, Clyde Snow, and William Haglund have testified at these tribunals and have consistently presented wide age ranges rather than false precision. The Srebrenica and other ICTY case testimony records, publicly available in the tribunal archives, provide concrete examples of how skeletal age estimates are presented and cross-examined in an international criminal law context.

In India, the Bharatiya Sakshya Adhiniyam 2023 (BSA 2023), which replaced the Indian Evidence Act 1872, governs expert evidence including forensic anthropological opinions. Section 39 of the BSA 2023 allows courts to receive expert opinions on matters of science. The practical standard applied by Indian courts for forensic age estimation is generally more flexible than Daubert or CPR 19, with considerable weight given to the medical board's composite assessment under the Juvenile Justice Act 2015 framework for living persons and to medico-legal officer reports for skeletal cases. However, as forensic anthropology develops as a recognised discipline in India (represented institutionally by AIIMS forensic medicine and by the Forensic Science Laboratory networks), the expectation of method documentation and uncertainty disclosure is increasing.

Age estimation data from these methods feeds directly into personal identification by comparative radiography and frontal sinus comparison when an antemortem record exists that can be matched. Skeletal histomorphometry uses the microscopic structure of cortical bone as an age indicator. Bone undergoes continuous remodeling throughout adult life, with old osteons being resorbed and replaced by new secondary osteons. The cumulative number of secondary osteons per unit area of cortical bone cross-section, and the proportion of fragmentary osteons (osteons incompletely replaced), increases progressively with age. The foundational work by D.H. Thompson (1979) and subsequently by Sam Stout, Anna Agnew, and others established regression equations relating osteon density to age in various skeletal elements, with the femoral midshaft and the anterior iliac crest being the most commonly sampled sites.

Histomorphometry requires destructive sampling (a section must be cut from the bone), which is a significant disadvantage in active criminal investigations where exhibit integrity may be contested. The accuracy of histological methods is similar to or slightly better than the sternal rib method in the middle adult range, with 95 per cent prediction intervals of approximately 10 to 20 years; it does not substantially resolve the over-50 ceiling problem. The method is used in UK and US casework as a supplementary approach when standard phase methods give conflicting or inconclusive results.

Dental cementum annulation counting, developed notably by Wittwer-Backofen et al. (2004) and others building on earlier zoological work, uses the annual bands laid down in dental cementum to count age in years from a known tooth formation date. The method has reported accuracies of plus or minus two to four years in some validation studies, which would represent a substantial improvement over skeletal phase methods for the older adult range. However, replication studies have shown more variable results (plus or minus five to ten years in some hands), and the technique requires thin-section preparation and microscopy that is technically demanding. The method is used in selected specialised laboratories in Germany (particularly the University of Freiburg group) and the United Kingdom, and has been applied in ICTY and German forensic cases.

Epigenetic clock methods, particularly DNA methylation-based age estimation from blood, saliva, or soft tissue samples, have shown promise in the biological literature with reported accuracy of plus or minus three to five years in some studies (Horvath 2013 methylation clock; Hannum et al. 2013 blood methylation). These methods are not yet routinely applicable to skeletal remains because DNA degradation reduces the CpG methylation signal in aged bone and tooth samples, though extraction from dental pulp in well-preserved teeth has produced usable results in some published cases. As of 2025, epigenetic clock methods are a research tool rather than a routine forensic method for skeletal age estimation, but several European forensic genetics groups (including the Netherlands Forensic Institute and the Erasmus University Rotterdam group) are actively validating the approach for casework use.