Why Do We Have Fingerprints? | The Everyday Biologist

Fingerprints get treated as a solved problem. Everyone assumes they exist so we can grip things and so police departments have a way to identify us. Both of those things are true in a loose sense, but neither one holds up as an actual explanation once you look at the biology closely. The friction story turns out to be more complicated than tire tread logic, and the patterns themselves are laid down by a process that has almost nothing to do with helping you hold a coffee mug. Recent work in genetics, developmental biology, and biomechanics has pieced together a much stranger and more interesting picture, so let’s go through what’s actually going on.

The three basic shapes, and how they’re built

Every fingerprint on every finger falls into one of three broad categories: arch, loop, or whorl [1,2]. Arches are the simplest, with ridges running roughly in one direction and no real center point. Loops curve back on themselves, entering and exiting from the same side of the finger. Whorls form a full circular or spiral pattern with a defined core. These types are classified by counting triradii, the Y-shaped points where three ridge directions meet, and cores, the center points ridges wrap around [2].

The pattern isn’t painted onto fully formed skin. It’s built during fetal development, starting at around week 13 of gestation on the raised volar pads at the fingertips, then spreading outward across the rest of the palm and digits over the following weeks [2]. By week 17 the primary ridges are complete, and by week 16 sweat glands have already started budding off from the deepest points of those ridges [2]. Whatever pattern you end up with is locked in before birth and stays fixed for the rest of your life, aside from stretching as the finger grows.

Fingerprints are almost hair follicles that gave up halfway

This is probably the most unexpected finding from recent work. A 2023 study using single-cell sequencing on developing human skin found that fingerprint ridges start out looking almost identical, at the molecular level, to hair follicles [2]. Both structures begin as small clusters of cells expressing the same trio of genes: EDAR, FGF20, and BMP2 [2]. Both intensify WNT signaling in the same way. For the first stretch of their development, a ridge cell and a hair follicle cell would be very hard to tell apart.

Then they diverge. Hair follicles go on to recruit a cluster of mesenchymal cells underneath them called a dermal condensate, which drives the follicle to grow down deep into the skin and eventually sprout a hair shaft under the control of sonic hedgehog (SHH) signaling [2]. Fingerprint ridges never do this. The mesenchyme under volar skin doesn’t respond to the signals that would normally trigger condensate formation, in part because it expresses high levels of a WNT inhibitor called DKK2 [2]. So the ridge stalls out as a shallow epithelial structure with no hair-producing machinery, expressing a growth factor called TGFa instead of SHH, which drives sweat gland formation rather than hair growth [2].

The actual patterning of where ridges go and how far apart they sit follows what’s called a Turing reaction-diffusion system, the same general kind of process that produces stripes on a zebrafish or spots on a leopard [2]. WNT and EDAR act as activators, BMP acts as an inhibitor, and the two signals interact across the tissue to break symmetry and produce a periodic pattern of ridges and gaps. Researchers confirmed this by manipulating the signals directly in mice: suppressing EDAR turns stripes into spots, boosting it makes the ridges thicker and further apart, and blocking BMP widens the ridges substantially [2]. This is a strong signature of a genuine Turing system rather than just a convenient metaphor.

What actually decides whether a given finger ends up an arch, loop, or whorl comes down to where these patterning waves start. Ridges initiate independently at three anatomical landmarks: the center of the volar pad, the tip of the digit near the nail, and the crease line closest to the fingertip [2]. These waves spread out and eventually collide, and the geometry of that collision determines the pattern type. Where three waves meet, you get a triradius. Simulations built around this collision model reproduce arches, loops, whorls, and even the rarer, oddball patterns seen occasionally in real fingerprints [2].

The genes involved are limb genes, not skin genes

A large genome-wide association study spanning more than 23,000 people from Han Chinese and European cohorts identified 43 genomic regions associated with fingerprint pattern type [3]. The striking part isn’t the number of hits, it’s what those genes actually do elsewhere in the body. When researchers ran enrichment analysis on the associated genes, the strongest hits weren’t for skin development at all, they were for limb development, embryonic limb morphogenesis, and abnormalities of digit and limb bone structure [3]. Skin-related pathways barely cleared the significance threshold.

The single strongest signal sat near a gene called EVI1, which the researchers followed up on directly in mouse models. Mice carrying an EVI1 mutation showed altered ridge patterns on their digits, and EVI1 expression tracked closely with limb bud growth and digit elongation rather than with the skin cells actually forming the ridges [3]. In human fetal tissue, EVI1 is heavily expressed in the mesenchyme under the volar pads early on, but by the time ridges are actually emerging at 16 weeks, expression has dropped off in that exact spot [3]. In other words, EVI1 isn’t shaping the ridge itself, it’s shaping the volar pad the ridge will later form on top of.

This lines up with something dermatoglyph researchers have suspected for decades: the size and shape of the fetal volar pad strongly predicts whether a whorl, loop, or arch forms there, with large domed pads tending toward whorls and flat or small pads tending toward arches [2,3]. The GWAS also found genetic correlations between fingerprint type and hand proportions, particularly the length of the little finger relative to the whole hand [3]. So genes that were never “trying” to build a fingerprint end up determining one anyway, purely as a side effect of building the underlying limb correctly.

What happens when the system fails completely

There’s a rare condition called adermatoglyphia, sometimes nicknamed “immigration delay disease” because affected people run into problems at any border checkpoint that relies on fingerprint scanning [4]. A 2011 study traced the cause in one large Swiss family to a mutation in a specific short isoform of a gene called SMARCAD1, one that’s normally expressed almost exclusively in skin [4]. The mutation disrupts a splice site, and the researchers confirmed with a minigene assay that this produces abnormal, unstable RNA transcripts that mostly get degraded before they can do anything [4].

What’s notable is how limited the effect is. People with this mutation have no fingerprints and reduced sweat gland density on their hands, but they’re otherwise completely unaffected [4]. That’s only possible because the mutation hits a skin-specific isoform of the gene rather than the main version, which is expressed everywhere in the body and is essential for normal development. Mice missing the full-length version of the same gene die around birth with severe skeletal defects [4]. This is a nice illustration of how tightly the fingerprint-forming machinery is walled off from everything else the same genes are doing elsewhere.

So what are they actually for?

The traditional explanation, that raised ridges increase friction the way tire tread does on a wet road, has been tested directly and doesn’t hold up well. A 2009 study using rubber models of primate skin found that fingerprint ridges reduce the contact area with a flat surface, which should decrease friction rather than increase it, since skin friction scales with contact area rather than surface roughness the way rubber does [5, referenced in 1].

A 2020 study using terahertz spectroscopy and infrared imaging offers a more complete answer. The ridges and the furrows between them work as a moisture-regulation system [1]. When a finger is too dry, sweat pores in the ridges release moisture, and the furrows act like microfluidic channels, wicking water to the edges where it evaporates quickly thanks to the sharp corners formed where ridge meets surface [1]. When a finger is too wet, pressing against something impermeable eventually flattens the ridges enough to physically block the pores, cutting off further sweat and letting excess moisture evaporate off through the furrows instead [1]. Either way, the system converges on a fairly narrow moisture range that happens to be exactly where friction is maximized [1]. Push past that point in either direction, too dry or too saturated with a thin film of water, and grip gets noticeably worse.

That also explains something odd about where fingerprint sweat glands sit in the body’s priority list. Unlike sweat glands elsewhere, which mostly manage temperature, the glands in fingerprint ridges respond to stress and emotional state rather than heat [1]. They’re a “fight or flight” system tuned for grip performance in a tense moment, not for cooling you down.

Putting it together

None of these four lines of research were looking for the same thing, and none of them lands on a single tidy answer. Fingerprints are shaped by genes whose real job is building limbs, laid down by a patterning system borrowed from hair follicle development and then cut short before it finishes the job, and useful in daily life mostly through fine moisture control rather than raw texture. The identification use case, which is the one thing most people associate fingerprints with, seems to be a coincidental byproduct of all this complexity rather than anything evolution was optimizing for. The actual reason we have them is closer to a limb-building side effect that happened to come with a built-in grip-assist function, which is a much stranger story than the one usually told.

Common Questions

Do fingerprints actually improve grip?

Not by increasing friction through texture the way tire treads do. What they seem to do instead is regulate moisture in the skin, keeping the surface hydrated enough for good friction without letting sweat pool into a slick film.

Are fingerprints determined by genetics?

Partly. Twin studies put heritability of pattern type around 30 to 80 percent. But the genes involved mostly control limb and digit development rather than skin biology directly, so they shape fingerprints indirectly by shaping the pads they form on.

Can someone be born without fingerprints?

Yes, though it's extremely rare. A handful of families have been documented with a condition called adermatoglyphia, caused by mutations affecting skin-specific gene isoforms. It's sometimes called immigration delay disease because it complicates travel through fingerprint-based border checks.

Why does everyone have a different fingerprint?

Fingerprint ridges form through a self-organizing patterning process similar to the one that lays out hair follicles, and this process is sensitive to small, essentially random variations as it unfolds. Combined with each person's slightly different hand geometry, no two patterns turn out the same, even in identical twins.

References