PRP for Hair Loss: How Strong Is the Evidence?

Platelet-rich plasma, or PRP, probably improves hair density for some people with androgenetic alopecia, but the evidence does not support a precise, universal estimate of benefit. A 2024 meta-analysis of randomized trials found about 28 additional hairs per square centimeter compared with control. The same analysis found extreme inconsistency among trials, low certainty, and signs that small favorable studies were more visible than unfavorable ones.

The practical conclusion is narrower than either “PRP works” or “PRP is unproven.” There is a reproducible positive signal for pattern hair loss, but preparation methods, injection schedules, comparators, measurement techniques, and follow-up differ so much that the procedure you are offered cannot be assumed to reproduce a pooled result.

Define the condition before judging the treatment#

Hair loss is not one diagnosis. Androgenetic alopecia, also called male or female pattern hair loss, involves progressive follicular miniaturization in a characteristic distribution. Alopecia areata is immune mediated, telogen effluvium reflects a shift in the hair cycle, and scarring alopecias can destroy follicles and require timely diagnosis. Nutritional, endocrine, medication-related, infectious, and mechanical causes add further possibilities.

Most favorable PRP evidence concerns androgenetic alopecia. A study mixing different conditions cannot answer the same question as a trial restricted to pattern hair loss. Before you interpret a response rate, check the diagnostic criteria, the sex distribution, the baseline severity, the concurrent therapies, and whether scarring disease was excluded.

This distinction also affects endpoints. A temporary reduction in shedding, regrowth after telogen effluvium, and reversal of follicular miniaturization are different biological outcomes. Pooling them under “hair improvement” creates a clinically vague result.

What PRP actually is#

PRP is made from a person's blood by centrifugation to concentrate platelets in a plasma fraction; platelets release growth factors and signaling proteins that may influence follicle cells, local blood vessels, and the hair cycle. The preparation is injected into the scalp.

That description sounds standardized, but it is not. Systems differ in the amount of blood processed, number and speed of centrifugation steps, final platelet concentration, red-cell contamination, leukocyte content, activation method, injected volume, depth, spacing, and storage time. Baseline platelet count and processing efficiency add biological variability. Two studies can both call their intervention PRP while delivering materially different products, so a meta-analysis estimates an average across a family of procedures rather than the effect of one reproducible formulation.

What the 2024 randomized-trial meta-analysis found#

Kieling and colleagues identified 14 randomized clinical trials involving 431 participants. Thirteen contributed to the pooled hair-density analysis. PRP increased hair density by a mean of 27.55 hairs per square centimeter versus control, with a 95% confidence interval from 14.31 to 40.79.

That estimate is statistically compatible with a positive average effect, but its apparent precision is misleading if read without heterogeneity. The I-squared statistic was about 96%. This means differences between study results were much larger than expected from sampling error alone. The pooled mean may not describe the effect of any particular protocol.

The analysis did not establish a clear improvement in hair diameter. Density and diameter are separate outcomes, and photographic appearance can also be affected by hair length, styling, color contrast, and image technique. A treatment can change a counting endpoint without producing a cosmetic difference you would notice in the mirror.

The authors rated the evidence as low certainty and detected probable publication bias. With small trials, a few missing neutral studies can materially shift the pooled estimate. Funnel-plot methods also have limited reliability when study numbers are small, so publication bias is a warning rather than a precisely measured correction.

What the larger 2025 review adds#

A 2025 systematic review searched through July 2025 and included 43 randomized trials with 1,877 participants across alopecia conditions and PRP comparisons; the broader evidence base continued to support improvements in hair-density outcomes and generally mild short-term adverse effects.

More participants improve the map, but they do not erase heterogeneity. The review combined trials with different diagnoses, protocols, comparators, and outcome scales. Some included studies had commercial relationships or product-specific interests. Broad pooling can answer whether the intervention family has a signal, yet it remains less useful for choosing a specific preparation or schedule.

The two reviews should therefore be read together. The narrow 2024 analysis gives a focused randomized estimate for androgenetic alopecia and makes uncertainty visible, and the larger 2025 review shows that the positive direction persists across a wider literature, while also confirming that standardization remains unresolved.

Comparator choice changes the claim#

PRP has been compared with saline injections, no procedure, topical therapies, other active treatments, and combinations. These comparisons answer different questions.

A placebo-controlled trial estimates the biological and procedural effect beyond the control injection, assuming masking works. A no-treatment comparison also includes attention and procedure effects. A PRP-plus-standard-therapy group compared with standard therapy alone tests added value, not whether PRP should replace established treatment. A noninferiority comparison needs a prespecified margin and adequate sample size; similar P values do not prove equivalence. Split-scalp designs, where each participant receives PRP on one side and control on the other, reduce between-person variability and can be efficient, but investigators still have to consider whether effects cross between sides and whether assessors remain masked when local reactions differ.

Outcome measurement can create apparent certainty#

Hair counts should use a defined target area, standardized photography or trichoscopy, reproducible clipping and positioning, and blinded assessment. Counts from slightly different scalp locations can vary. A change from baseline within the PRP group is not enough because hair density can fluctuate and controls can also improve.

Patient-reported improvement matters because cosmetic benefit is the aim for many people, but satisfaction is affected by expectations and knowledge of treatment. Trials should report both objective measures and validated patient assessments, with the timing specified.

Short follow-up is a major constraint. Hair cycles take months, and androgenetic alopecia is chronic. An improvement shortly after a series of injections does not establish that benefit persists, how often maintenance is needed, or what happens after stopping.

Safety evidence is reassuring but incomplete#

Reported adverse effects are usually local and short lived, including pain, tenderness, swelling, bruising, bleeding, headache, or itching. Because the material is autologous, immune reaction to the plasma itself is less likely than with a foreign product.

Autologous does not mean risk free. Any scalp injection can introduce infection, injure local structures, or worsen symptoms in susceptible skin. Anticoagulant use, platelet disorders, active scalp disease, and other medical factors may affect procedure risk or product quality. Small trials cannot reliably detect uncommon complications, and adverse-event collection is not uniform. So safety reporting should give you denominators, severity, duration, relationship to each session, withdrawals, and follow-up. “No serious events observed” in a small sample means no serious event was seen, not that one is impossible.

Commercial translation is the weak link#

A clinic may cite a favorable meta-analysis without showing that its product resembles those studied. Useful procedural reporting includes the preparation device, centrifugation protocol, whole-blood and final volumes, platelet and leukocyte counts, activation, injection pattern, number and interval of sessions, co-interventions, and maintenance plan.

If those details are absent, neither you nor your clinician can connect a quoted effect size to the procedure on offer. Brand claims and before-and-after photographs do not substitute for randomized comparative data with standardized imaging.

Cost and treatment burden also matter. Repeated injections cost you time, discomfort, and continuing expense. Most trials do not provide strong cost-effectiveness evidence or identify which baseline features predict a clinically meaningful response.

A useful evidence checklist#

1. Was androgenetic alopecia diagnosed clearly, and were other causes addressed?
2. Was treatment randomized, and was the comparator appropriate to the claim?
3. Was the PRP product characterized beyond its name?
4. Were hair counts measured in a fixed target area by masked assessors?
5. Were absolute changes, between-group differences, and confidence intervals reported?
6. Were concurrent hair-loss therapies balanced or controlled?
7. Was follow-up long enough to assess persistence?
8. Were adverse events collected systematically after every session?
9. Are commercial ties and protocol ownership disclosed?
10. Does the offered procedure match the studied procedure closely enough for the evidence to travel?