How to Read a Skincare Study (Without a Science Degree)

Skincare marketing loves to cite “studies” and “clinical trials.” But not all studies are created equal, and knowing how to evaluate research claims can help you distinguish genuine evidence from cherry-picked marketing.

You don’t need a PhD to critically assess skincare research. Here’s a practical guide to reading studies like a skeptic.

The Hierarchy of Evidence

Not all research carries equal weight. From weakest to strongest:

1. In Vitro Studies (Weakest)

“In vitro” means “in glass” — experiments on cells in petri dishes or test tubes.

What they show: How ingredients behave in controlled laboratory conditions.

Limitations: Cells in a dish don’t behave like cells in living skin. The ingredient may not penetrate in real use. Concentrations used may be far higher than in products.

Marketing use: “Shown to increase collagen production” might mean cells in a dish produced more collagen when bathed in concentrated ingredient. This doesn’t prove the same happens when you apply a cream.

2. Animal Studies

Research conducted on animals (often mice, pigs, or rabbits).

What they show: How ingredients behave in living organisms.

Limitations: Animal skin differs from human skin. Ethical concerns aside, results may not translate to humans.

Marketing use: Rarely cited directly but sometimes referenced as “preclinical research.”

3. Uncontrolled Human Studies

Studies where people use a product, but without a comparison group or placebo.

What they show: What happened to participants who used the product.

Limitations: No way to know if results were due to the product, placebo effect, natural variation, or other factors. If 80% reported improvement, was that because of the product or because people expect improvement from products they’re testing?

Marketing use: “In a study, 87% of participants reported improved skin texture.” This sounds impressive but means little without a control group.

4. Controlled Human Studies

Participants are divided into groups: one uses the product, another uses a placebo or comparator. Groups are compared.

What they show: Whether the product produces better results than the comparison.

Strengths: Can actually demonstrate that the product does something beyond placebo effect.

Marketing use: “Clinically proven” — though look at what it’s compared to and how results were measured.

5. Randomised Controlled Trials (RCTs)

Participants are randomly assigned to groups. Ideally “double-blind” — neither participants nor researchers know who’s getting the real product.

What they show: Cause-and-effect relationships with minimal bias.

Strengths: Random assignment prevents selection bias. Blinding prevents expectation effects.

Limitations: Expensive and complex, so skincare studies are often smaller and shorter than ideal.

6. Systematic Reviews and Meta-Analyses (Strongest)

Combine results from multiple studies to draw overall conclusions.

What they show: The weight of evidence across multiple investigations.

Strengths: More reliable than any single study.

Limitations: Rare in cosmetic research because there often aren’t enough quality studies to combine.

Key Questions to Ask

When encountering a study claim, ask:

1. Was there a control group?

Without comparison to placebo or alternative treatment, improvements might be due to the placebo effect, natural variation, or simply using any moisturiser.

“50% improvement in wrinkle depth” is meaningless if the control group showed 45% improvement.

2. How many people participated?

Small studies are common in skincare research. While they can suggest effects, they’re less reliable than larger studies.

Rules of thumb:

Under 20 people: Preliminary, treat claims cautiously
20-50 people: Moderate evidence, still limited
50-100 people: Reasonable for initial claims
100+ people: Stronger evidence
500+ people: Rare in cosmetics, but robust if conducted properly

3. How were results measured?

Objective measurements (instrumentation measuring skin hydration, thickness, wrinkle depth):

More reliable
Can detect changes participants can’t perceive
May show changes that aren’t visible to the naked eye

Subjective assessments (participants rating their own improvement):

Influenced by expectation
Subject to placebo effect
Can be gamed through question phrasing

Expert grading (dermatologists rating before/after photos):

Better than self-assessment
Still somewhat subjective
Blinding of assessors matters

A study reporting “instrument-measured 15% reduction in wrinkle depth” is more reliable than “87% of participants felt their wrinkles improved.”

4. How long was the study?

Skin changes take time. Studies running 4-6 weeks may not capture meaningful long-term effects.

Duration matters because:

The stratum corneum renews in 2-4 weeks
Collagen changes take months to become apparent
Short studies favour fast-acting cosmetic effects over genuine structural improvement

A 12-week study is more informative than a 4-week study for anti-ageing claims.

5. Who funded it?

Manufacturer-funded research is not automatically invalid, but it carries higher bias risk.

Funding effects:

Tendency to test products under favourable conditions
Publication bias (positive results published, negative results buried)
Methodological choices that favour the product

Independent research (academic institutions, government agencies) carries less bias risk.

6. Was it published in a peer-reviewed journal?

Peer review means other scientists evaluated the methodology before publication.

Peer-reviewed sources:

Journal of the American Academy of Dermatology
British Journal of Dermatology
International Journal of Cosmetic Science
Journal of Cosmetic Dermatology

Not peer-reviewed:

Company websites
Press releases
“Internal studies”
Conference presentations (limited peer review)

7. Can you find the actual study?

If a brand cites “clinical studies,” can you access them?

Red flags:

“Studies show…” with no citation
References to “internal research”
Claims based on single study by the brand’s affiliated laboratory

Better transparency:

Links to published papers
Citation details (journal, authors, year)
Willingness to share study details on request

Common Statistical Tricks

The Percentage Game

“300% improvement” sounds dramatic. But 300% improvement from a baseline of 1 to 4 on a 100-point scale is trivial.

Always ask: percentage of what? On what scale? Compared to what baseline?

Statistical Significance vs. Clinical Significance

“Statistically significant” means the result is unlikely due to chance. It doesn’t mean the effect is large or visible.

A study might find a statistically significant 5% improvement in hydration — real, but would you notice it? Probably not.

Cherry-Picked Endpoints

Studies often measure multiple outcomes. If most show no effect but one shows improvement, guess which one appears in marketing?

Look for “primary endpoints” — what the study was mainly designed to measure — not secondary outcomes discovered after the fact.

The Subgroup Shuffle

If the overall group doesn’t show improvement, maybe a subgroup does. “Particularly effective for women over 50 with combination skin” might mean the product didn’t work for most participants.

Reading Study Abstracts

Most people won’t read full papers, but abstracts are accessible and informative.

Key sections to look for:

Methods:

How many participants?
Was there a control group?
How long did it run?
What was measured?

Results:

What were the actual numbers?
What was the control group’s result?
Were results statistically significant?

Conclusions:

Do conclusions match the results?
Are limitations acknowledged?

Be wary of abstracts where impressive conclusions don’t match modest results.

Where to Find Studies

Free access:

PubMed (pubmed.ncbi.nlm.nih.gov): Abstracts for most medical/scientific literature
Google Scholar: Broad search including some full-text access
PubMed Central: Full-text articles in open access

What to search:

Ingredient name + “clinical trial”
Ingredient name + “human study”
Ingredient name + “randomized controlled trial”

When Evidence Is Lacking

For many skincare ingredients, high-quality evidence simply doesn’t exist. This doesn’t mean they don’t work — just that we don’t have proof.

Reasonable positions when evidence is limited:

Ingredients with in vitro evidence and plausible mechanisms: Worth trying, lower confidence
Ingredients with small human studies: Moderate confidence, especially if multiple studies agree
Ingredients with large, replicated RCTs: High confidence

Don’t reject everything without large trials, but calibrate your expectations to the evidence available.

The Practical Takeaway

When a product claims “clinically proven”:

Ask for the study details
Assess study quality (size, control, measurements, duration)
Consider who funded it
Look at what was actually measured
Compare the claim to the evidence

Most marketing claims fall apart under scrutiny — not because companies lie, but because “clinically tested” can mean almost anything.

Becoming a critical reader of research protects you from overpaying for under-proven products. It’s not about rejecting science; it’s about holding marketing claims to the standards science actually requires.