Which Retinoid Works Best for Photoaging? New Rankings

The retinoid you choose may depend on what you want to treat. A 2025 network meta-analysis ranked topical isotretinoin highest for fine wrinkles and tazarotene highest for coarse wrinkles. Tretinoin, the prescription standby with the deepest direct clinical record, ranked as a strong all-purpose option rather than first for a single photoaging outcome. The catch is substantial: the 23 randomized trials behind those rankings mostly enrolled lightly pigmented participants, so they cannot settle which retinoid works best for the global majority of people with photoaging.

That makes the usual advice—just use tretinoin—reasonable, but incomplete. Tretinoin remains the best-characterized starting point for many patients. Still, the newer comparison suggests that the dominant problem matters: superficial lines, deeper structural wrinkles, roughness, or uneven pigment.

A ranking, not a final answer

Lin et al.'s 2025 network meta-analysis combined 23 randomized controlled trials of isotretinoin, tazarotene, tretinoin, retinol, and glycolic acid for facial photoaging. It assessed fine wrinkles, coarse wrinkles, roughness, hyperpigmentation, and tolerability. The authors used a Bayesian model with four chains, 250,000 iterations, and a 50,000-iteration burn-in period, and excluded treatment arms with fewer than 30 participants.

Unlike a conventional trial, a network meta-analysis can compare treatments that were never tested against each other in the same study. If one trial compares tretinoin with vehicle and another compares tazarotene with vehicle, the analysis can estimate a tazarotene-versus-tretinoin comparison through that shared control. This is useful, but it rests on an assumption: the linked trials must be similar enough in their participants, formulations, treatment schedules, and follow-up periods for the indirect comparison to hold.

The analysis used ranking probabilities, including surface-under-the-cumulative-ranking-curve measures, rather than a direct measurement of superiority in one large head-to-head trial. The publication's reported rankings place isotretinoin first for fine wrinkles and tazarotene first for coarse wrinkles. Tretinoin did not rank first in an individual outcome category, while tretinoin and retinol performed particularly well for hyperpigmentation. The available publication record does not provide the numerical ranking values here, so the rankings should not be mistaken for a known percentage advantage or a guaranteed clinical difference.

Why different retinoids could produce different results

The result is biologically plausible. Tretinoin is all-trans retinoic acid, while retinol must move through enzymatic conversion steps before it becomes retinoic acid in skin. That helps explain why retinol can improve mild photoaging but generally has lower potency than prescription retinoic-acid derivatives at equivalent concentrations. A 2022 clinical review by Milosheska et al. describes tretinoin's direct trial evidence for increasing collagen I and III synthesis, reducing matrix metalloproteinase activity, and improving epidermal thinning.

Tazarotene is a prodrug converted to tazarotenic acid. It has more selective retinoic-acid-receptor activity than tretinoin and may penetrate tissue differently. Mambwe et al.'s 2025 review describes these distinctions among retinoid compounds and their receptor activity. They offer a possible explanation for different wrinkle rankings, but they do not prove that receptor selectivity caused tazarotene's coarse-wrinkle result. The trials measured skin outcomes, not a causal chain from receptor binding to a specific crease beside the nose.

Isotretinoin is a 13-cis retinoic acid isomer with effects on sebum and collagen-related pathways that differ from all-trans retinoic acid. Those differences could help explain its fine-wrinkle ranking. They do not turn the network result into direct proof that topical isotretinoin will outperform tretinoin for every person with fine lines.

Glycolic acid belongs in a different mechanistic category. It works as a chemical exfoliant rather than through retinoic-acid receptors. That distinction matters when interpreting a ranking across all five compounds: the analysis compares treatments with different expected benefits and side-effect patterns, not five interchangeable versions of the same ingredient.

Why tretinoin still earns its place

Tretinoin's position in clinical practice comes from more than habit. It has extensive direct randomized-trial support and a long treatment history in photoaging. Sadick et al.'s 2023 review identifies cumulative ultraviolet exposure as the leading contributor to facial aging and reports that topical retinoids can improve wrinkling, texture, elasticity, and skin tone through effects on collagen and skin-cell biology.

So the apparent tension between tretinoin's clinical dominance and the 2025 rankings is not a contradiction. The older clinical literature answers, "Which topical retinoid has the most established direct evidence?" Tretinoin is the answer. The network meta-analysis asks, "Which treatment has the highest estimated rank for each outcome across a connected body of trials?" Those are related questions, not identical ones.

For someone seeking one well-studied prescription option for several signs of photoaging, tretinoin remains a sensible default. For someone whose main concern is a specific wrinkle type, the rankings create a useful question for a dermatologist—not a reason to discard tretinoin on the strength of indirect evidence.

The darker-skin evidence gap changes the advice

The most important limitation is who appears in the studies. Photoaging research has concentrated on lightly pigmented people, while Fitzpatrick III–VI skin is inadequately represented in the trials underlying these rankings. That matters because photoaging does not always look the same across skin types. Dyschromia and texture may dominate the clinical picture more often in darker phototypes, whereas fine-line patterns feature more prominently in Fitzpatrick I–II skin.

Ultraviolet exposure still drives photoaging across phototypes. But a trial program that prioritizes wrinkle scales in lightly pigmented participants may not capture the outcome a darker-skinned patient most wants to change: uneven pigment. The NMA's finding that tretinoin and retinol performed well for hyperpigmentation may therefore matter more clinically for many patients than the fine- or coarse-wrinkle leaders.

Irritation adds another constraint. Tazarotene has a higher irritation profile than tretinoin at standard concentrations, according to Mambwe et al. In skin prone to post-inflammatory hyperpigmentation, inflammation can leave pigment changes behind. The existing trials do not establish how the NMA's tolerability ranking maps onto post-inflammatory hyperpigmentation risk in Fitzpatrick III–VI populations. A treatment cannot claim a practical advantage if irritation prevents continued use or worsens the complaint being treated.

How much confidence should you place in the rankings?

Network meta-analysis can organize a scattered literature, but rankings can look more precise than the underlying trials. Differences in vehicle, concentration, application frequency, outcome scoring, and follow-up can distort an indirect comparison. Small studies remain vulnerable to random variation even after an analysis excludes arms with fewer than 30 participants. A 2024 network meta-analysis of acne-scar treatments illustrates the same broader problem in aesthetic dermatology: multiple interventions, variable outcome measures, and relatively small trials produce rankings that need clinical context.

The 2025 photoaging analysis also does not replace direct head-to-head trials of isotretinoin, tazarotene, and tretinoin. Its rankings are probabilistic inferences, not evidence that one product beat another under identical conditions. Industry sponsorship could also favor a formulation, and the available abstracts do not permit a firm assessment of that risk.

A practical way to use this evidence

• Fine wrinkles: Topical isotretinoin ranked highest in the network analysis. Discuss availability and suitability with a prescriber rather than assuming it is automatically better for you.
• Coarse wrinkles: Tazarotene ranked highest, but irritation may limit its usefulness. Escalate gradually and reassess if persistent inflammation develops.
• Hyperpigmentation or broad photoaging goals: Tretinoin remains the best-established prescription standard, and retinol offers an over-the-counter option with meaningful evidence but lower expected potency.
• Fitzpatrick III–VI skin: Treat the wrinkle rankings as unvalidated for your demographic. A conservative start, slow titration, sun protection, and early attention to irritation are more defensible than chasing the top-ranked ingredient.

The useful lesson is not that tretinoin suddenly became second-best. It is that "best" depends on the outcome, the skin's tolerance, and whether the trial population resembles the patient in front of you.

The study that would materially change practice is straightforward: a large direct comparison of topical isotretinoin, tazarotene, and tretinoin in diverse Fitzpatrick III–VI populations, with dyschromia and hyperpigmentation measured alongside wrinkles and tolerability. Until then, the rankings are an informed starting point for individualized treatment—not a universal prescription.