The word "collagen" comes from the Greek kólla, meaning glue. That's not a figure of speech: right up until the 19th century, the glue used for wood, leather, and paper was literally made by boiling down animal bones, skins, and tendons until that same protein was extracted. The gelatin in Jell-O, or in gummy candy, is exactly the same molecule, just cooked. The collagen in your capsules, the glue an 18th-century carpenter kept on his bench, and the wobble in your dessert are, chemically, the same family of proteins.
It's also the single most abundant protein in the human body, making up close to 30% of all the protein you have, more than any other. It's the cement holding together skin, tendons, cartilage, bone, and blood vessels. Nearly thirty distinct types have been catalogued across the body's tissues, but in skin, two of them dominate by far: type I and type III.
This isn't a question for wrinkles at 80
Here's a common misconception: collagen gets filed under wrinkles and old age, as if the subject only applied to the later decades of life. The reality is different. A study covering a large number of people measured skin collagen across adulthood and found a continuous, year-on-year decline starting early in adult life, not a sudden collapse after 70 (Shuster et al., Br J Dermatol, 1975).
That decline often speeds up around the forties, exactly when sex hormones begin to shift: skin, like other tissues, responds directly to those changes, and collagen production is part of what responds (Sator et al., Exp Dermatol, 2004). That's precisely why the collagen question starts making sense in your late thirties, not at some age where the topic suddenly becomes relevant overnight.
To put the full scale of the process in perspective: comparing skin from 18-to-29-year-olds with skin from people over 80, type I collagen production turned out to be about 32% lower in the older group, with fibroblasts — the cells that make collagen — running noticeably less active (Varani et al., Am J Pathol, 2006). That figure doesn't mean nothing happens before then. It shows the full arc of a process that starts much earlier.
You can't just swallow collagen as it is
Here's a detail people often miss: collagen is a very large protein, far too large to cross the gut wall in one piece. To be absorbed, it first has to be cut down into small fragments, a process called hydrolysis. That's why it's called "hydrolyzed collagen" or "collagen peptides", never whole collagen.
Did you know? Researchers had volunteers drink a collagen preparation, then drew their blood hour by hour. The result: a specific small fragment called Pro-Hyp shows up in the blood, peaks after 1 to 2 hours, then tapers off. This fragment is so stable it survives digestion and circulates intact (Iwai et al., J Agric Food Chem, 2005). A follow-up study found that how much of this fragment reaches the blood depends on where the collagen comes from — fish skin, fish scale, and pig skin don't get absorbed the same way (Ohara et al., J Agric Food Chem, 2007). In other words, an "instant" collagen, already broken down into small soluble fragments, isn't a marketing flourish. It's the condition for the thing to do anything at all.
The dose that's actually held up
Here's the most useful question, and the one least often answered honestly: does more mean better?
One study directly compared two daily doses, 2.5g and 5g, in 69 women aged 35 to 55 — precisely the age bracket at issue — over 8 weeks, against a placebo group. The result: both doses significantly improved skin elasticity compared with placebo, and neither outperformed the other (Proksch et al., Skin Pharmacol Physiol, 2014). Doubling the dose, in other words, changed nothing. It's one of the few studies to test this question head-on, and its conclusion is clear: 2.5g a day is a dose that works, not a bare-minimum dose.
A broader synthesis of 19 studies and 1,125 people (95% women, aged 20 to 70) confirms improvements in hydration, elasticity, and wrinkles, placing the effect at around 90 days of regular intake (de Miranda et al., Int J Dermatol, 2021). A more recent review of 26 trials and 1,721 people points the same way on hydration and elasticity, while honestly noting that larger trials are still needed to confirm the exact size of the effect (Pu et al., Nutrients, 2023).
What to take from this: consistency matters more than quantity. A moderate dose, taken daily for several weeks, showed results just as good as a dose twice as high. Patience matters too: the effects build up over weeks, not days — a pattern that shows up in almost everything biology teaches us about aging.
Type I and III: not a coincidence
Why these two types specifically, out of the roughly thirty on record? Because they're overwhelmingly the two types of collagen that make up skin itself. Type I forms the bulk of the dermis's structural framework; type III, more supple, works alongside it and dominates in younger skin, gradually giving ground to type I with age. This isn't an arbitrary formulation choice. It's a choice that follows the biology of the tissue it's meant to support, fragment by fragment.
Cells running out of steam, not just a slowing factory
Why do fibroblasts make less collagen as we age? Part of the answer touches on something bigger than skin alone: cells that, over time, stop dividing and functioning normally, rather like exhausted workers still at their post but no longer really producing. Varani's team explicitly describes "cellular fibroblast aging" to explain what they observed (Varani et al., Am J Pathol, 2006). It's the same cellular mechanism, at work in skin as elsewhere in the body, that we cover in more depth in another Journal article on these aging cells.
Marine, bovine, porcine: does the source actually matter?
One question comes up constantly: is marine collagen better than bovine or porcine? The same study that tracked the Pro-Hyp fragment through the blood compared three sources on the same volunteers, under the same conditions: fish skin, fish scale, and pig skin. The result: total fragment levels found in the blood were significantly higher with fish scale than with pig skin (Ohara et al., J Agric Food Chem, 2007). A real, measured difference on the same protocol, but on its own it can't prove that one source produces more visible skin results — that still needs trials comparing clinical outcomes directly, not just blood levels. What can be said plainly is that fragment size and solubility matter at least as much as which animal it came from.
Beyond skin, a surprising lead
Collagen doesn't stop at skin — logically enough, since it also makes up tendons, cartilage, and bone. A meta-analysis of people with osteoarthritis found a significant improvement in joint symptoms with collagen supplementation, particularly on stiffness (García-Coronado et al., Int Orthop, 2019). In athletes with knee pain, 5g of collagen peptides a day for 12 weeks significantly reduced pain during activity compared with placebo (Zdzieblik et al., Appl Physiol Nutr Metab, 2017). These studies use different doses and target a different part of the body — mainly, they show collagen is a molecule with a far wider reach than most people assume, not just a cosmetic trend born on social media.
What to take from this
Collagen isn't just a word on a label, and it isn't a subject reserved for the later decades of life. The question starts to matter once the decline begins speeding up, often around the forties, and it connects to something bigger than skin alone: cells that age and simply do less, not just a factory running short on raw material. What matters, in order: that it's hydrolyzed into absorbable fragments, that it matches the types I and III skin actually uses, and that it's taken at a consistent dose — 2.5g a day has proven itself against double that amount, not worse, just as good. Here as elsewhere, the right question isn't "how much can I take", but "what precise dose matches what the body can actually use." It's the least dramatic version of the collagen story. It's also the most accurate.
This article is published for informational and educational purposes. It does not constitute medical advice and does not replace consultation with a healthcare professional.
