Hair, skin, and nails formulas: what biotin and collagen trials actually show

Hair-skin-nails formulas are among the largest segments of the supplement industry, with retail sales over $1.5 billion annually in the U.S. alone. The two ingredients that drive most of the marketing are biotin and collagen peptides. The trial evidence for each is more limited than the bottles imply: biotin only works when biotin deficiency is the actual cause of the symptom, and collagen has small but real skin effects with weaker hair and nail data.

The biotin myth

Biotin (vitamin B7) is required for several carboxylase enzymes and for healthy keratin synthesis. True biotin deficiency, rare in adults, can cause alopecia, brittle nails, and dermatitis, and supplementation reverses these manifestations. The leap that biotin supplementation in adults without deficiency improves hair or nail outcomes is not supported by trial evidence. A 2017 systematic review of 18 studies on biotin for hair and nail growth found that all reported benefit was in patients with underlying pathology causing deficiency (genetic biotinidase deficiency, long-term anticonvulsant therapy, alcoholism); no trial in adults with normal biotin status showed improvement [1]. The dermatology consensus is that biotin supplementation does not improve hair or nail growth in non-deficient adults [2].

The biotin lab-interference problem

Biotin at supraphysiologic doses (5,000 mcg and above, common in OTC products) interferes with immunoassay-based laboratory tests including thyroid function tests, troponin, vitamin D, and pregnancy hormones. The FDA issued a 2017 safety communication after a fatal misdiagnosis attributed to falsely depressed troponin in a patient taking biotin [3]. Patients on high-dose biotin should stop the supplement 48-72 hours before any laboratory testing.

The collagen peptide skin evidence

Oral hydrolyzed collagen does survive digestion to a meaningful extent, with small bioactive di- and tripeptides including prolyl-hydroxyproline appearing in plasma after ingestion [4]. The most consistent trial data are for skin: a 2021 systematic review of 19 RCTs in 1,125 participants reported small but statistically significant improvements in skin hydration (mean difference 0.28 standardized units) and elasticity (0.31 standardized units) after 8-12 weeks of 2.5-10 g/day collagen peptides [5]. The effect size is roughly comparable to topical moisturizer use, and the durability of effect after stopping supplementation is not well characterized.

Collagen for hair and nails: weaker evidence

The hair-growth trials with collagen are smaller, often manufacturer-funded, and inconsistent. A 2024 pooled analysis of 6 trials on collagen for hair found a small increase in hair density with substantial heterogeneity and high risk of bias [6]. For nail growth, a single open-label 24-week trial of 25 women with brittle nails reported a 12% increase in nail growth rate with 2.5 g/day specific bioactive collagen peptides [7]; a separate placebo-controlled trial in 100 women found no significant nail-growth difference at 16 weeks.

What's actually in a typical formula

A typical hair-skin-nails product contains biotin (often 5,000-10,000 mcg, far above the AI of 30 mcg), 50-200 mg vitamin C, 25-100 mg zinc (above the UL in some products), and a proprietary blend of MSM, silica, and herbal extracts at undisclosed doses. The combination has not been tested as such in randomized trials; the marketing relies on transferring evidence from single-ingredient studies to the multi-ingredient blend. Many products' total active ingredient mass is below the doses tested in positive single-ingredient trials.

Bottom line

For adults without biotin deficiency, biotin supplementation does not improve hair or nails and can interfere with diagnostic blood tests. Hydrolyzed collagen at 2.5-10 g/day for 8-12 weeks produces small improvements in skin hydration and elasticity. The combined hair-skin-nails formula category is sold on transfer of single-ingredient evidence to multi-ingredient blends that have not themselves been tested.