What Bones Tell — One Person's Diary, Split Between Teeth and Bone

A Skeleton from a Grave

You have probably seen the bones of an ancient person lying inside a glass display case at a museum. The label next to them usually has a few lines: "Woman in her 30s, height around 150 cm, poor nutritional status in childhood." That grave would have had no name and no letters, but where did sentences like this come from?

The answer is the bone itself. The human body writes things down little by little on its bones and teeth while it lives. Therefore, a skeleton is like a diary without letters. Different pages of this diary record different periods. Teeth record childhood, and bones, which are continually rewritten, record more of the later years.

This article reads a person's diary from beginning to end. Let us say a skeleton was found in grave number 7 of an ancient village communal cemetery. This person, and the values reported to come from those bones, are all fictionalized by this article to aid understanding. Real research comes after that picture is grasped.

First, we look at what can be read from the diary, and at the end, we look at what this diary leaves out. That is because the diary of certain people is not left in the grave at all.

From the Cover — Estimating Age and Sex by Shape

When we lay out the bones from Tomb No. 7, we read the cover first. Who were they, and how old were they? Let us assume that judging by the shape and size of the bones, they were likely female, and their age came out to be between 35 and 50. Age does not come out as a single number; it comes out as a range. Here, the span is 15 years.

There is a field that does this work. Bioarchaeology is the field that reads ancient human bones and teeth, alongside the places where they were buried, to retrace how those people lived. For a long time, bones and teeth were pushed aside to the status of mere appendices in excavation reports. Then, an overview from 2002 notes that archaeologists and historians began to recognize them as the most direct means to gauge the bodies of past people (DOI: 10.1017/cbo9780511549953.003).

The basis for reading the cover is shape and size. The shape and size of bones differ slightly depending on sex and age. There are even methods to estimate height, sex, age, and ancestral lineage from just a single upper part of a thigh bone (PMID 35504140).

Estimation comes with errors. A child's age is estimated by how much the bones and teeth have grown, and because the rate of growth varies by sex and group, that much margin of error must be attached (PMID 3312468). Adults are harder. Being unable to precisely pin down the age of adult bones is a problem that remains persistent in this field (PMID 29539545). One review article that gathered and organized multiple studies notes that adults over 50 in particular are underestimated in published studies (PMID 36787786).

So when old people seem scarce in an ancient graveyard, one of the reasons is the method of reading age itself. Even the "35–50 years" for the person in Tomb No. 7 is merely the most plausible range, and we must keep open the possibility that they were actually older.

The age and sex told by bones are estimates and must always be read as a range.

Childhood Pages — Stature and Horizontal Lines on Teeth

Suppose we measure the thigh bone length of the person from Tomb No. 7 and find their height to be around 155 cm. We estimate their height in life from the lengths of the long bones in their arms and legs (PMID 35389750). Adult height is the accumulated result of conditions during growth, so it is used as a proxy indicator to gauge the health of a population. Numerous studies have reported that the greater the reliance on farming, the more adult height decreased (PMID 21507735).

Height also mixes in inherited genes. If a height measured from bones comes out small, it is difficult to separate based on bones alone whether it is due to growing up undernourished or because they inherited a short stature to begin with. Therefore, a study that read together the bones and genes of 167 ancient Europeans placed two heights side by side for each person. One is the "height they should have reached" estimated from that person's genes, and the other is the actual height measured from bones. This study reads the amount by which actual height fell short of genetic height as the degree of growth stunting during childhood.

For Neolithic people in the early farming era, this shortfall was an average of 3.82cm greater than for people in the preceding hunter-gatherer era. This means they fell short of the height expected from their genes by a larger margin. However, Europe's earliest farmers shared ancestral lineages that differed substantially from the hunter-gatherers who came before them. When the portion that could stem from this difference in ancestral lineage was subtracted through calculation (adjusted), that difference shrank to 2.82cm. This is a level hard to distinguish from a chance difference, in other words, a level difficult to call statistically distinct. While the authors view this as suggesting that the health of early farmers was relatively poor, they also note this limitation (PMID 35389750). Height points to health, but it is hard to read apart from genes.

Now let us look at the incisor of the person from Tomb No. 7. Suppose that when illuminated by oblique light, we see two horizontal grooves. Enamel is the hardest layer covering the outside of a tooth. It is laid down layer by layer at a steady rhythm during childhood while the tooth is forming inside the gums (DOI: 10.1002/ajpa.1330330506).

Enamel hypoplasia is a mark left as a horizontal line on the tooth surface because enamel is laid down thinner when the body suffers a major illness during the childhood years when teeth were forming. Once formed, enamel is not repaired and used again. Therefore, these marks are not erased and remain in the order the teeth grew, showing in chronological order when the body underwent severe stress (DOI: 10.1002/ajpa.1330330506). To put it in terms of a diary, the tooth is the childhood page, and a page that, once written on, is never rewritten.

The age when those lines formed is counted backward by the height at which the lines sit on the tooth. For lines formed around age two or three, estimation methods generally yield similar values, but for lines formed very early or late, the values diverge depending on the method (PMID 34081785). Let us assume the two lines on the person from Tomb No. 7 are from around ages two and three. This is the range where the methods match best.

There is also something these lines do not say. A review published in 1990 noted that while horizontal lines are sensitive markers of malnutrition and infection, they do not point to a specific disease, and it is not yet known precisely what creates the lines (DOI: 10.1002/ajpa.1330330506). Alongside malnutrition and illness, metabolic or genetic abnormalities and physical shocks to the body are also cited as candidate causes (PMID 34081785). Therefore, what these lines tell us goes only as far as "the body suffered severe stress around age two," and whether that stress was starvation must be inquired separately through other evidence.

The lines have one more meaning. While the body gets through a crisis, it redirects energy to places essential for survival, and during that time the growth of teeth and bones slows down. Therefore, such a mark is also evidence that the person got through that crisis (PMID 30561095). The person from Tomb No. 7 suffered severe bodily stress twice, around ages two and three, and survived both times.

Horizontal lines on teeth record indelibly around what age the body suffered severe stress, but they do not record what caused it.

Later Pages — Diseases Left on Bone

Let us assume that the left arm bone of the person in Grave No. 7 shows a healed break. Paleopathology is the field that reads the traces left on the bones and teeth of ancient people to reconstruct what illnesses they suffered from while alive.

How healed fractures are distributed across a community points to whether they lived on rugged land, did dangerous work, or how frequently fighting and violence occurred (DOI: 10.1017/cbo9780511549953.003). That is a story told when you gather the fractures of many individuals. A single fracture in one person tells us only this: "They were injured, and they lived until the bone healed."

Unlike teeth, bones are continuously remodeled throughout life (DOI: 10.1520/jfs14829j). That is why bone is a page in the diary that keeps getting rewritten. Some marks, such as a healed break, remain for a long time, but the material making up the bone is replaced by that of the final few years. This story of bone materials returns in the next section.

Disease marks left on bone rarely state a disease's name. Tuberculosis is a good example. The way tuberculosis collapses the spine is considered a signature unique to tuberculosis, but other marks, such as changes left on the ribs, are not signs pointing solely to tuberculosis (PMID 25802030). Moreover, only a few percent of people who had tuberculosis show bone changes (PMID 38012917).

This page of the diary records only long-lasting illnesses and those that reached the bone. If the spine of the person in Grave No. 7 is clean, that could mean this person never had tuberculosis, or that they had it but it did not reach the bone.

The diseases written on bones are only a fraction, and a blank page is not evidence of health.

Elements in Bone — What We Ate, Where We Grew Up

What the person in Grave 7 ate is not written in the shape of their bones. What was eaten is written in the building blocks that make up bones and teeth.

Isotopes are versions of the same element that differ only slightly in weight. Carbon, nitrogen, and strontium all have mixtures of different weights, and the proportions of those mixtures vary by food and by land. These differences in proportion across the food chain are stamped onto a person's bones and teeth (DOI: 10.1515/anre-2017-0017). However, the ratios of carbon and nitrogen can also change slightly after they enter the body and become bone (DOI: 10.1002/ajhb.22398). The method of measuring carbon isotopes that decrease over time to assign an age is covered in "Dating with Carbon — Radiocarbon Dates, Read with an Hourglass and Tree Rings". What we are looking at here is the ratio left behind by what was eaten.

Piecing together what an ancient person ate using the ratio of carbon and nitrogen is a direct and widely used method (DOI: 10.1002/ajhb.22398). Carbon values tell us apart what kinds of plants were eaten. For Late Bronze Age people from the Great Hungarian Plain, carbon values typical of crops like millet appear, raising the possibility that they ate millet (PMID 29746569). Which rung of the food chain the food came from is also revealed. The diets of weaned children were generally similar to those of adults, but in societies that did not practice hunter-gathering, children appear to have eaten more food from a lower rung of the food chain than adults did—namely, plants (PMID 28786488).

Where someone grew up is told by different elements. Strontium isotopes are strontium atoms with different weights, and their mixed proportions vary slightly from land to land; the ratio is embedded as-is in the teeth of a person who grew up eating things produced on that land. This ratio does not change even after passing through the food chain (DOI: 10.1520/jfs14829j).

Which body part it is recorded in matters. A tooth's value is the average of what was eaten during childhood when the tooth formed, and does not change thereafter. Because bone is constantly being remodeled, a bone's value is an average of the final few years, and a methodological paper from 2000 sets that period at approximately the last 10 years (DOI: 10.1520/jfs14829j).

"The values of this village's land" must be found out separately. Establishing a baseline using the bones of animals or close family members that lived in the same place and ate the same things is best; if that is not possible, it is estimated from the rocks and geology of that land (DOI: 10.1520/jfs14829j).

Let us suppose that when compared against the baseline values of the village established in this way, the teeth of the person in Grave 7 differed from those values. Then this person very likely spent their childhood in a different land. Let us suppose that the carbon and nitrogen values of the bone pointed to a diet similar to the people of this village. This means that during their final few years, they lived by eating like the people of this village. When you place both sides side by side, it tells the story of a person who grew up in a different land and moved to this village to live.

A real-life case shows both sides. The bones of a man who died about 5,000 years ago emerged from the bog of Vittrup in Denmark. He was beaten severely before being placed in the bog, raising the possibility that he was offered as a sacrifice. Around that time, Denmark was a land of farming people. Yet the strontium value remaining in his teeth pointed to him being born outside of Denmark, in a place that could be Norway or Sweden. The oxygen value of the teeth pointed to him having spent his childhood in a colder place (PMID 38354111).

His diet also differed between the two periods. The childhood diet remaining in his teeth belonged to a person who fished and hunted, while the later diet remaining in his bone belonged to a farming person. When measured along the direction in which a single tooth grew, the time when he moved from one society to another was around the end of his teens. His genome was also close to ancient hunter-gatherers of Norway and Sweden (PMID 38354111). What genes have to say is covered in "What Genes Can Tell Us — An Old Book Copied Down the Generations".

Limits have also been revealed through real-life cases. Regarding a girl from the Bronze Age around 3,400 years ago from Egtved in Denmark, a study from 2015 concluded that the girl came from outside present-day Denmark (with the exception of Bornholm Island) and traveled long distances during the last few months of her life (DOI: 10.1038/srep10431).

Another study from 2019 took issue with the baseline values. It argued that agricultural lime spread on fields today disturbs the strontium values of modern soil measured to establish baseline values. When compared again using values unaffected by lime, the interpretation emerged that it is most plausible the Egtved girl was born close to where she was buried rather than in a distant foreign region. This study notes that strontium migration studies conducted in low-lime soils need to be re-examined (DOI: 10.1126/sciadv.aav8083).

Depending on what was taken as the baseline for the same girl, she was split into "a person from afar" and "a person from around here".

Teeth record the land and diet of childhood, and bones record the land and diet of the final few years, and those values acquire meaning only when compared against baseline values.

The Trap — A Grave is Not the Whole Village

In a village in the Guatemalan highlands, the skulls of people who died in 1982 remain as forensic data. When children suffer from anemia, small pitted marks are left on the skull bones. The degree of anemia suffered by rural children in that region today matched well with the amount of marks left on those adult skulls. Yet those marks were very rare compared to the skulls of the ancient Maya. These marks, common in ancient Maya skulls, had been cited as evidence that the ancient Maya people back then were extremely undernourished (DOI: 10.1525/aa.1998.100.4.924).

The authors note that small changes in diet and infection may have also contributed, pointing to mortality as a more plausible explanation. If a child who suffered from anemia died before reaching adulthood, there is no adult skull with marks left behind. Therefore, the authors put forward the hypothesis that in the past, more children who suffered from anemia lived to adulthood than do today (DOI: 10.1525/aa.1998.100.4.924).

This problem has a name given by a research team in 1992 (DOI: 10.1002/ajpa.20233). The osteological paradox refers to the problem that a person with signs of disease on their bones may not be the person who was sicker, but rather the person who survived that disease for some time. From this perspective, bones with disease marks may actually be the healthy core of that group, and bones without marks may belong to people whose immunity was weak and who died before they could acquire the marks (DOI: 10.1002/ajpa.20233).

Put in terms of a diary, it is like this. The diaries gathered in a grave do not belong to everyone who lived in that village, but to the people who died in that village. And a disease is written in the diary only as much as one endured that disease.

Even so, simply inverting it to read "more marks meant healthier" can also be wrong. What the paradox says is that the mere fact that there are many marks does not determine which of the two readings is correct. The side with more marks may have been sicker, or they may have endured the disease longer. Which one it is must be examined for each sample.

A study comparing two cemeteries in medieval Denmark is an example that examined this with the paradox in mind. One side is a leprosarium cemetery, likely a group that had things harder. The other side is a cemetery of an ordinary community that had it better. Children on the leprosarium side were shorter in the older age group, had less bone mineralization, and had more stress markers on average. The authors viewed this as the result of overlapping poor health and social disadvantage, and concluded that this result supports the more traditional interpretation of reading the side with more marks as the sicker side (DOI: 10.1002/ajpa.20233). In short, even after accounting for the paradox in this case, the interpretation reading the side with more marks as the sicker side was supported. The paradox does not mean throwing away old interpretations, but rather checking them instead of assuming them in advance.

Since the direction is not determined solely by the abundance or scarcity of marks, each sample must be asked again. A recent synthesis concludes that while this paradox has sometimes been regarded as an insurmountable wall, it is closer to a stepping stone that has pushed the field forward (PMID 42538595).

A grave is a collection of diaries of the people who died in that village, and disease marks are written down only as much as one endured that disease.

When You Read a Museum Label

Let's say you encounter a museum label like this: "Bronze Age woman, 35–50 years old. Experienced starvation in childhood, and came to this village from a distant place." This text is a made-up label. Based on what we read earlier, we can ask four questions.

"35–50 years old" is a range. The age of adult bones is hard to pin down precisely, and older people tend to be underestimated. This leaves open the possibility that this person was even older.

If the basis for "childhood starvation" is horizontal lines on the teeth, what those lines tell us is only that the body underwent severe stress around a certain age. Whether it was starvation cannot be known from the lines alone. And those lines are also a record that this person got through that crisis.

Two things are asked of "came from a distant place." Is it the value of the teeth, or the value of the bones? If it is the value of the teeth, it speaks of the land of childhood. And what was it compared against? If the baseline value wavers, the conclusion can change, just like the Egtved Girl.

Lastly, this person is just one of the people who died in that village. Whether there are marks on this person's bones or not, that does not speak on behalf of the health of the entire village.

The reason the story of the Bottendorf Man was solid is because the strontium and oxygen in the teeth, the diet left separately in the teeth and bones, and the genome all matched up. What a single set of bones tells us becomes clearer the more it is placed alongside other evidence.

Where to Look Further

How to distinguish whether a mark left on a person's bone is a sign of weakness or of endurance is still an ongoing problem. A recent synthesis places its hopes on a method of creating virtual humans inside a computer and testing them (PMID 42538595).

How the transverse lines, bottle marks, and osteological paradox seen in this article are used in actual debates is covered by "Did Farming Make Us Sicker? Reading the Marks Left on Bone". It re-examines the conventional wisdom that agriculture worsened health using bone evidence that turned out differently from region to region.

Other articles in "How Do We Know the Past?" cover the evidence placed alongside bones one by one. Written records are read by "Reading Historical Records — Grandmother's Diary and That Day's Newspaper", the ages of bones and charcoal by "Dating with Carbon — Radiocarbon Dates, Read with an Hourglass and Tree Rings", and genomes extracted from bone by "What Genes Can Tell Us — An Old Book Copied Down the Generations".

Source Papers

• "Reconstructing Health Profiles from Skeletal Remains" (2002) — DOI: 10.1017/cbo9780511549953.003 — An overview of nine bone and tooth lesions used in health indices. The patterns of healed fractures indicate rough terrain, hazardous work, or the degree of violence. Human remains are the most direct means of gauging past bodies.
• "Proximal femur in biological profile estimation - Current knowledge and future directions." (2022) — PMID 35504140 — Stature, sex, age, and ancestry can be quantitatively estimated using only the upper part of the thigh bone (a forensic review).
• "Estimating age at death from immature human skeletons: an overview." (1987) — PMID 3312468 — Estimating the age at death of juvenile remains must reflect sex differences in the growth process and errors arising from population variation (a review).
• "Cementochronology, to cut or not to cut?" (2016) — PMID 29539545 — The inability to accurately estimate the age of adult skeletons is an ongoing problem in paleopathology and paleodemography, and accurate ages are also needed to avoid part of the osteological paradox (a review).
• "Paleodemography: From archaeology and skeletal age estimation to life in the past." (2022) — PMID 36787786 — Estimating the age at death of adult skeletons is particularly problematic, and adults over 50 are underreported in published studies (a review).
• "An integrative skeletal and paleogenomic analysis of stature variation suggests relatively reduced health for early European farmers." (2022) — PMID 35389750 — Comparison of long-bone stature and genetically predicted stature in 167 Europeans: Neolithic people fell short of predictions by an average of 3.82 cm more than previous eras, but the difference weakened to 2.82 cm (P=0.120) after ancestral correction.
• "Stature and robusticity during the agricultural transition: evidence from the bioarchaeological record." (2011) — PMID 21507735 — In the majority of studies since 1984, adult stature decreased as agricultural dependence increased, and the trend of declining health after the agricultural transition remains valid (a comprehensive review).
• "Assessment of systemic physiological perturbations from dental enamel hypoplasias and associated histological structures" (1990) — DOI: 10.1002/ajpa.1330330506 — Enamel is not remodeled and grows regularly like rings, making enamel hypoplasia an indelible chronicle of stress; it is a nonspecific indicator, and the specific causes are not yet known (a review).
• "Estimation of age at onset of linear enamel hypoplasia. New calculation tool, description and comparison of current methods." (2021) — PMID 34081785 — Candidate causes for transverse lines include nutritional disorders, mechanical trauma, disease, metabolic disorders, and genetic abnormalities. Methods for estimating the age of formation generally agree around ages 2 to 3, but diverge for very early or very late lines.
• "Bioarchaeological evidence for adaptive plasticity and constraint: Exploring life-history trade-offs in the human past." (2019) — PMID 30561095 — Early stress marks in teeth and bones are evidence of survival achieved by diverting energy to essential functions to get through crises, at the cost of suppressed enamel and bone growth (a review).
• "Old World tuberculosis: Evidence from human remains with a review of current research and future prospects." (2015) — PMID 25802030 — Spinal tuberculosis (Pott's disease) is specific to tuberculosis, while rib lesions and others are nonspecific. Evidence of Old World tuberculosis is concentrated in the Northern Hemisphere, particularly in late medieval Europe.
• "Probable cases of tuberculosis in Early Medieval pastoralists of Eastern Europe." (2023) — PMID 38012917 — Only a few percent of tuberculosis patients show bone changes, which limits direct comparisons based on morphology alone, and definitive diagnosis requires ancient DNA analysis.
• "Role of stable isotope analyses in reconstructing past life-histories and the provenancing human skeletal remains: a review" (2017) — DOI: 10.1515/anre-2017-0017 — Stable isotope signals from food webs are imprinted in bones and teeth, tracking diet and weaning with carbon and nitrogen, and childhood geography and migration with strontium and oxygen (a review).
• "Beyond diet reconstruction: Stable isotope applications to human physiology, health, and nutrition" (2013) — DOI: 10.1002/ajhb.22398 — Carbon and nitrogen stable isotope analysis is a direct and widely used technique for modeling diets, and is expanding into physiological and disease research (a review).
• "5000 years of dietary variations of prehistoric farmers in the Great Hungarian Plain." (2018) — PMID 29746569 — Carbon and nitrogen values of 52 individuals: mostly reliance on C3 plants, with C4-plant-type carbon values in the Late Bronze Age that may suggest millet consumption.
• "Post-weaning diet in archaeological human populations: A meta-analysis of carbon and nitrogen stable isotope ratios of child skeletons." (2017) — PMID 28786488 — The diets of weaned children and adults were generally similar, but children in non-hunter-gatherer societies appear to have eaten significantly more food from lower trophic levels (terrestrial C3 plants) than adults (a meta-analysis).
• "Strontium Isotope Composition of Skeletal Material Can Determine the Birth Place and Geographic Mobility of Humans and Animals" (2000) — DOI: 10.1520/jfs14829j — Strontium ratios vary by region and do not change even after passing through the food web. Teeth reflect childhood averages, while bones reflect the average of the last roughly 10 years. Baseline values are obtained from animals with the same diet or close relatives, or failing that, established using geological principles.
• "Vittrup Man-The life-history of a genetic foreigner in Neolithic Denmark." (2024) — PMID 38354111 — A man from a Danish bog dating back about 5,000 years: strontium indicates birth in Norway or Sweden, oxygen indicates a cold childhood, teeth indicate a fishing and hunting diet, bones indicate an agricultural diet, and migration occurred in his late teens. His genome is also close to Scandinavian Mesolithic people.
• "Tracing the dynamic life story of a Bronze Age Female" (2015) — DOI: 10.1038/srep10431 — Concludes that the Egtved Girl, from about 3,400 years ago, originated outside present-day Denmark (excluding Bornholm) and traveled long distances during the last few months of her life.
• "Agricultural lime disturbs natural strontium isotope variations: Implications for provenance and migration studies" (2019) — DOI: 10.1126/sciadv.aav8083 — Agricultural lime disturbs natural strontium values. When re-evaluated using unaffected values, it is most plausible that the Egtved and Skrydstrup women were born near where they were buried. Studies in low-calcareous soils need to be reassessed.
• "Porotic Hyperostosis and Paleoepidemiology: A Forensic Perspective on Anemia among the Ancient Maya" (1998) — DOI: 10.1525/aa.1998.100.4.924 — Anemia lesions in 1982 Guatemalan forensic remains match modern child anemia levels, but are much rarer than in archaeological samples. This is likely due to high mortality rates, supporting the hypothesis that more children who suffered from anemia in the past survived to adulthood.
• "Comparison of child morbidity in two contrasting medieval cemeteries from Denmark" (2005) — DOI: 10.1002/ajpa.20233 — Addressing the osteological paradox (Wood et al. 1992) by comparing children from a leper hospital cemetery and an ordinary community cemetery: the hospital group was smaller, had lower mineralization, and showed more stress indicators, supporting the more traditional interpretation.
• "From Optimism to Opportunity: New Perspectives on the Osteological Paradox." (2026) — PMID 42538595 — Although the osteological paradox was sometimes viewed as an insurmountable obstacle, it is closer to a stepping stone that has propelled the field forward. Hopes are pinned on agent-based modeling (a synthesis).