Part III. The human olfactory epithelium and bulb

3.1 The olfactory epithelium: numbers that have never been remeasured

Area. Textbooks give about 5 cm² per nasal cavity, 10 cm² in total. Tracing this back shows that the only direct measurement is that of von Brunn in 1892, on two cadavers: 307 and 238 mm², probably for one side of the nose Human. The collection strands found no modern measurement. For comparison, one review states that the olfactory epithelium covers about 3% of the nasal cavity surface in humans against about 50% in rats (cited via a review, not checked against the original).

Location. Sources disagree. Biopsies of 97 samples from 33 people found olfactory epithelium on the middle turbinate as well, with the proportion of samples containing olfactory epithelium ranging 30–76% depending on site; electro-olfactogram recordings also found responses in the region anterior to the middle turbinate in 7/12 people. But a 2021 study concluded that the epithelium lies only on the anterior two thirds of the superior turbinate and the opposing septum. Measurements on 13 specimens aged 18–45 placed the lower border 12–16 mm below the cribriform plate. The older the person, the lower the probability of obtaining olfactory epithelium at biopsy (r = −0.73), as respiratory epithelium gradually intrudes.

Total neuron count. The often-cited figure of “6–10 million” or “6–20 million” sensory neurons has never been counted stereologically in humans. In mice, each side of the epithelium has about 5 million neurons; in rats the number rises from about 0.5 million at birth to about 21 million at 8 weeks of age Animal.

Cellular structure. Each neuron has about 10–30 olfactory cilia on its olfactory knob according to measurements in 8 people; another source gives 10–20. As for stem cells, older histological studies could not separate the two basal cell types in humans, but newer molecular data confirm that humans have two progenitor populations similar to those of rodents.

3.2 Sensory neuron regeneration and single-cell data

The olfactory epithelium is one of the few places in the human nervous system that generates new neurons throughout life. But we have not found a measurement of the turnover rate in humans — there has been no carbon-14 dating of olfactory sensory neurons as has been done for the olfactory bulb. The only reference figure is a half-life of about 90 days in mice Animal.

The largest single-cell sequencing study of the human epithelium (4 patients, 28,726 cells) captured only 222 mature neurons, and described the expression of 140 receptors. The figure 140 does not mean that humans use only 140 receptors; it reflects a sampling limit — with 222 neurons it is impossible to encounter all of the roughly 350–400 intact receptor genes.

3.3 The olfactory bulb: volume, glomeruli and mitral cells

Glomerulus number — correcting a textbook figure. The figure of “1,100–1,200 glomeruli per bulb in humans” that circulates through many popular sources is not based on any count in humans. Actual histological counts give:

• 5,568 ± 830 (standard error) glomeruli per bulb (range 2,975–9,325), in about 7 bulbs aged 39–89, with immunolabelling and statistical correction. The under-50 group had a mean of 6,960 (n = 2), the over-50 group 5,012 (n = 5) — a difference that was not statistically significant (p = 0.33) Human.
• About 8,000 glomeruli in young people by Nissl staining, falling by about 10% per decade to under 30% by age 90–100 Human.

These two figures are compatible once specimen age and identification criteria are taken into account. The usable conclusion: about 5,000–8,000 glomeruli per bulb, with large inter-individual variation, declining with age. Notably, the figure of 1,100–1,200 coincides with the number of functional receptor genes in mice — possibly the origin of the confusion, although no source has confirmed this.

An important theoretical consequence: in mice, each receptor gene corresponds to about 2.5 glomeruli per bulb — close to the principle that “one receptor type converges onto about two glomeruli”. In humans, the ratio is about 16 glomeruli per receptor. The classic convergence principle derived from rodents does not apply unchanged to humans, and the reason is still unknown.

Mitral cells. In young people about 40,000–51,000 cells per bulb depending on the source, falling to about 33,000 at age 60 and about 14,500 at age 95 in an all-female sample Human.

Volume measured by MRI — comparable only within the same protocol.

There is a difference of about 2.5-fold between laboratories for the same structure. Volume falls by about 0.32 mm³ per year. The relationship with function in healthy people is weak: r about 0.2 in 125 people (conference abstract), and only β = 0.09 standard deviations in a sample of 5,605 people. An olfactory sulcus depth of ≤ 8 mm distinguishes congenital anosmia with 100% specificity but 72% sensitivity.

3.4 The adult human olfactory bulb does not replace its neurons

In rodents, the olfactory bulb continuously receives new neurons from the subventricular zone, with up to 50% of neurons replaced each year. Whether humans do the same was once a major debate. The strongest quantitative evidence currently available answers no:

• Carbon-14 dating of olfactory bulb neurons using the twentieth-century nuclear weapons tests (15 neuronal DNA samples) gave a turnover rate of 0.008 ± 0.08% per year (confidence interval includes 0) — under 1% after 100 years. Non-neuronal cells turn over at 2–3.4% per year Human. Limitation: the method measures the whole neuronal population, so it cannot exclude a small subpopulation with rapid turnover, such as granule cells.
• The stream of cells migrating to the olfactory bulb is strong during the first 18 months of life and then almost shuts down Human.
• One study found no neuroblasts at all in the adult human olfactory bulb Human.
• On the other side is a 2007 study describing a migratory stream in adults, but it has not been replicated quantitatively and has been directly rebutted.

The consequence: rodent mechanisms of olfactory learning based on new olfactory bulb neurons — such as the role of granule cells born at a specific time in childhood odour memory, presented in Volume 2 — transfer poorly to humans. In humans, olfactory bulb plasticity most likely relies on other mechanisms.

3.5 The olfactory peduncle, olfactory tract and anterior olfactory nucleus

A non-peer-reviewed 2024 preprint on 6 people measured the olfactory peduncle at 2.8 cm long, the olfactory bulb at 1.0 cm, and the cross-sectional area of the peduncle at about 2.1 mm² Human. But the number of axons in the human olfactory tract has never been counted, and the volume or neuron count of the anterior olfactory nucleus has never been measured stereologically.

3.6 Early pathology in the olfactory bulb

The olfactory bulb is one of the first sites to accumulate pathological proteins, and this is far more common than generally imagined — even in people without symptoms.

Two samples on α-synuclein in “normal” older adults gave 7% and 20% from the same brain bank; the discrepancy may be mainly due to how the control group was defined, together with differences in staining method and sample size.

A monograph compiled from 259 sourced facts collected under a common protocol. Practical-implication sections are the compilers’ inferences. Diagrams are newly drawn from published data. Scientific reference material, not medical advice.