The Smell of Human Bodies

Human body odor is not a single thing. It varies by individual, by moment, by what someone has eaten, by the microorganisms living on their skin, and by their genetic makeup. Some of that variation is random. Some of it follows patterns that are reproducible and meaningful — patterns connected to ancestry, diet, and biology in ways that have been studied with enough rigor to make confident claims. This is not a controversial position in the scientific literature. It's straightforwardly documented.

The interesting question isn't whether the variation exists. It's what produces it.

The architecture of human body odor

Body odor isn't sweat. Eccrine sweat — the kind that covers the body's surface for thermoregulation — is mostly water, salt, and trace metabolites. It's relatively odorless at source. The characteristic smell of a human body comes from a different system: the apocrine glands, which are concentrated in the axillary region, the groin, and around the nipples, and which secrete a thicker, lipid-rich fluid containing a variety of odor precursor molecules.

The precursors themselves are also largely odorless. The smell is produced by a second step: bacterial metabolism. The skin surface is colonized by communities of Corynebacterium and Staphylococcus species that enzymatically break down the precursor molecules into volatile compounds. The primary odorants identified by researchers include 3-methyl-2-hexenoic acid (3M2H), described as goat-like; 3-hydroxy-3-methylhexanoic acid (HMHA), described as cumin-like; and 3-methyl-3-sulfanylhexan-1-ol (3M3SH), which produces the sulfurous, onion-like character of armpit odor. The steroids androstenone and androstenol, which have musky and faintly urinary characters respectively, contribute additional dimensions that vary with hormonal profile.

The overall character of a person's body odor is therefore a function of at least three interacting systems: what their apocrine glands secrete (genetics-dependent), what bacteria live on their skin (microbiome-dependent), and what those bacteria have to work with (diet- and health-dependent). None of these systems operates independently, and the final smell is the product of all three.

The genetic layer: ABCC11

The most dramatic source of body odor variation between populations is a single gene: ABCC11 (ATP-binding cassette transporter C11), which encodes an efflux pump in apocrine gland cells. This pump's job is to actively transport odor precursor molecules — specifically S-glutathione conjugates and amino acid conjugates — from inside the gland cells onto the skin surface where bacteria can reach them.

A single-nucleotide polymorphism at position rs17822931 (538G→A) produces a non-functional version of this transporter. Individuals with the homozygous AA genotype have a largely inactive pump: the precursor molecules are not transported to the skin surface, the bacteria have little to metabolize, and the characteristic axillary odor is nearly absent.

The population genetics of this variant are striking. The functional G allele is ancestral — the original human genotype. The A allele arose as a mutation and swept through East Asian populations to high frequency through what researchers believe was a selective sweep during prehistoric migration. Among East Asians (Chinese, Japanese, Korean, and closely related populations), 80–95% carry the AA genotype. Among Europeans, approximately 2% carry it. Among Africans, it's rarer still. The wet earwax phenotype — which is determined by the same gene, since the same ABCC11 transporter affects ceruminous glands in the ear — follows the same distribution: dry, flaky, grayish earwax in East Asians; wet, sticky, brownish earwax in Europeans and Africans.

The practical olfactory consequence is large. The overwhelming majority of Han Chinese, Japanese, and Korean individuals have minimal axillary odor at the source level — the precursors aren't on the skin surface to be metabolized. The smell that's present tends to be dominated by eccrine sweat metabolites, skin surface sebum, hair, and the diet-derived volatile compounds that come through the skin more uniformly. This produces a qualitatively different odor profile from the typical European one, not merely a less intense version of it. The odorants are different, not just quieter.

Researchers studying axillary odor across ethnic groups have confirmed this directly: both 3M2H and HMHA are significantly more abundant in African-American samples than in Caucasian samples, and more abundant in Caucasian samples than in East Asian ones — following the ABCC11 allele frequency distribution closely. The cumin-like and goat-like character of typical Western axillary odor is in large part an ABCC11-mediated phenomenon.

The microbiome layer

On top of the genetic baseline, the skin microbiome introduces variation. Even within populations where the G allele is common, individuals have different bacterial communities in their axillary region, and those communities differ in which enzymes they express and which compounds they produce.

Staphylococcus hominis is the primary producer of the sulfurous thioalcohol (3M3SH) that contributes the onion-like character. Corynebacterium species primarily produce the fatty acid components (3M2H, HMHA). The balance between these bacterial populations — which varies by individual, by hygiene practice, by environment, and by diet — shifts the dominant character of the odor: more sulfurous-onion from a Staphylococcus-heavy microbiome, more fatty-cumin from a Corynebacterium-heavy one.

Research has found that East Asians tend to have higher levels of S. hominis and lower levels of S. epidermidis in their axillary microbiome than Caucasians or Hispanics. This microbiome difference is likely itself partly genetically mediated (the ABCC11 substrate affects what grows) and partly environmental, and it adds a second layer of population-level variation on top of the precursor availability.

Diet affects the microbiome, and therefore affects the odor. A high-meat diet increases the availability of amino acids that bacterial pathways metabolize into sulfurous and nitrogenous compounds. Garlic and onion introduce organosulfur compounds (allyl methyl sulfide, diallyl disulfide) that are absorbed through the gut, circulate in blood, and are excreted through both breath and skin. Cumin and coriander consumed in food produce metabolites with similar character to the axillary compounds — the olfactory overlap between dietary spices and body chemistry is not coincidental; they share structural chemistry. Cruciferous vegetables produce isothiocyanates. Red meat, particularly in large amounts, is consistently associated with stronger body odor in sensory studies.

This creates a partial feedback loop between cuisine and body smell: populations with culinary traditions heavy in specific spices, fermented foods, or animal proteins will, through diet, modulate their baseline genetic body odor profile in specific directions. The smell of a person reflects what they eat in the same way the smell of a kitchen does.

What you actually smell

The olfactory reality of human body odor, at close range and beyond the level that deodorant removes, tends to fall into a few recognizable clusters that correspond reasonably well to the chemistry.

The warm, slightly fatty-rancid dimension comes from volatile fatty acids — butyric, isovaleric, propionic — which intensify with animal protein consumption and certain skin microbiome profiles. This reads as ranging from "warm skin" at low concentration to "cheese" or "rancid butter" at high concentration. It tends to intensify with age, as sebum composition changes and bacterial communities shift.

The animalic-musky dimension comes primarily from androstenone and androstenol, the steroid compounds. Androstenone has a urine-adjacent, slightly sandalwood character; androstenol is fresher, described as "fresh sweat" or faintly floral. Approximately one third of people have significant anosmia to androstenone and don't detect it at all, which produces genuine individual variation in how masculine body odor registers — some people find the steroid layer almost absent; others find it dominant.

The sulfurous-onion dimension comes from 3M3SH via Staphylococcus metabolism, and tends to vary significantly by individual and by microbiome composition.

The indolic dimension — warmer, darker, more animal in the civet sense — comes from skatole and indole produced during protein digestion in the gut. These are excreted through the skin and in breath and contribute the deeper, heavier animalic character that becomes more prominent with certain diets, with digestive disturbances, and at higher body temperatures.

The practical observation that people who cook and live in specific culinary traditions tend to carry traces of those flavors in their skin chemistry is real and documented — not a stereotype but a biochemical consequence of what the skin excretes during digestion of aromatic compounds. The observation that populations with specific genetic profiles tend to have qualitatively different axillary odor profiles is equally real and equally documented, grounded in one of the best-characterized single-gene effects on human odor biology.

Both of these effects are neutral facts about human biology. They describe variation without hierarchy, the same way genetic variation in earwax type or lactase persistence describes variation. The smell of a person is the sum of their genetics, their microbiome, what they eat, their hormonal profile, their age, and their individual variation within all of those systems. It's one of the most complex sensory signatures a body produces, and it's entirely specific to the person producing it.

Which is, when you consider it in the context of everything else in this series, exactly what we should expect. The ingredients are variable. The blending is complex. The result is unrepeatable.