Twice the dose, the same result? What the MARS trial tells us about 0.02% vs 0.04% atropine

By Jagrut Lallu — BOptom (Hons), MSc Specialty Lenses (Hons), FIAOMC. Therapeutic Optometrist & Contact Lens Specialist. Founder, New Zealand Eye Research Centre; Partner, Rose Optometry; WCO Asia-Pacific Myopia Ambassador; International Myopia Institute NZ Ambassador; Clinical Senior Lecturer, Deakin School of Optometry; Honorary Teaching Fellow, University of Auckland.

Low-dose atropine eye drops are among the most studied options for slowing myopia (short-sightedness) progression in children, but which concentration to use remains an open question. A new Czech trial, MARS, compared 0.02% and 0.04% atropine with placebo. Both doses slowed eye growth over the first year. The more tempting question, whether doubling the dose adds anything, is one this trial was not designed to answer, and the reason is worth understanding.

What the researchers did

MARS was an investigator-initiated, phase IIIb trial run across five Czech centres. It was randomised, double-masked and placebo-controlled: compounded atropine and placebo came in identical bottles, and neither families nor assessors knew the allocation. Children aged 6–11 years with cycloplegic refraction between −0.50 and −4.75 D and no previous myopia control were eligible. Importantly, the cohort was enriched for fast progressors: each child needed documented axial elongation of at least 0.10–0.12 mm (depending on age) in the six months before randomisation. All analysed participants were reported as white, and the mean age was 10.13 years.

Of 235 children randomised 2:1:1 (119 to 0.02%, 58 to 0.04%, 58 to placebo), 191 (81.3%) contributed to the primary analysis. Children used one drop in each eye nightly. This report is the prespecified 12-month analysis of a planned 24-month treatment phase followed by a 12-month washout. The primary outcome was the between-group difference in axial length change (eye growth) for 0.02% versus placebo.

What they found

Placebo eyes grew 0.265 mm over 12 months (95% CI 0.219–0.311). Compared with placebo:

• 0.02% atropine: −0.144 mm (95% CI −0.201 to −0.087; P<0.0001)
• 0.04% atropine: −0.157 mm (95% CI −0.222 to −0.092; P<0.0001)

A 95% confidence interval (CI) is the range of values the true effect plausibly lies in. Neither interval crosses zero, so the slowing is unlikely to be chance alone; P<0.0001 means a difference this large would be very unlikely if atropine truly had no effect. The findings held in sensitivity analyses, including one imputing missing data for all 235 children. For refraction, a secondary outcome, placebo eyes shifted −0.501 D; the between-group differences versus placebo were 0.352 D for 0.02% and 0.389 D for 0.04%. Corneal measurements did not differ between groups, consistent with the effect coming from slower eye growth rather than corneal change.

Twice the dose, the same result?

Comparing the two active doses directly, the difference in eye growth was 0.013 mm (95% CI −0.043 to 0.069; P=0.8935). It is tempting to read this as "0.02% works as well as 0.04%". The authors explicitly state that it does not establish equivalence, and the confidence interval shows why. The data are compatible with 0.02% producing anything from 0.043 mm less growth to 0.069 mm more growth than 0.04% over a year. By our arithmetic, 0.069 mm is close to half the size of the 0.02% effect against placebo, which would matter if real. "No significant difference" means the trial could not rule out zero; it does not mean it ruled out a clinically meaningful difference.

Three design features reinforce that caution. First, recruitment ended early (235 randomised against 288 planned) because DIMS/HAL myopia-control spectacles entered routine Czech practice and families declined the chance of placebo. Post hoc power for the primary comparison was still above 90%, but the trial was never powered to compare doses, and only 49 children in the 0.04% arm were analysed. Second, no equivalence margin was prespecified, which is how a trial formally declares two treatments interchangeable. Third, the cohort was selected for fast growth, so effects may differ in slower-progressing children.

This fits the wider picture discussed in our post on the atropine dose–response plateau, and sits alongside approaches that adjust concentration according to response, covered in our review of stepwise atropine escalation. The authors also cite CHAMP-UK (a small effect with 0.01%), MOSAIC (0.05% benefit in year 3) and PEDIG (no benefit with 0.01% in North America) as context.

What about side-effects?

This paper does not report adverse events. Pupil and accommodation data from 127 participants are reported separately (treatment-related pupil dilation, with no significant between-group difference in near accommodation facility), and near vision, comfort, irritation and adverse-event analyses are still to come. This paper alone is therefore not evidence that the two doses are equally well tolerated.

Limitations of this evidence

• Early termination of recruitment, with 18.7% of randomised children excluded from the primary analysis (including 5 who started DIMS spectacles and 33 without evaluable month-12 data).
• Not powered to compare the two active doses, and no prespecified equivalence margin.
• An enriched, fast-progressing, white Czech cohort aged 6–11; how well results transfer to children in Aotearoa New Zealand of other ethnicities is uncertain.
• Twelve months only: durability, rebound after the drops are discontinued, and safety data are still to be reported.
• Compounded hospital-pharmacy drops were used; results may not transfer directly to other preparations.

On the other side of the ledger, the trial was publicly funded (Ministry of Health of the Czech Republic and CZECRIN), with no industry funding and no declared conflicts beyond one author's editorial board role at the journal; she was not involved in its review.

What this means in practice

MARS adds well-controlled evidence that both 0.02% and 0.04% atropine slowed eye growth over 12 months in children who were already progressing quickly. It does not show the two doses are interchangeable. Choosing a concentration still depends on the individual child, including age, rate of axial growth, tolerance of the drops and response over time, and should be monitored with axial length measurement rather than refraction alone. Our optometrists at Rose Optometry in Hamilton, including Jagrut Lallu and Jacqueline Rowe, provide myopia control care, including low-dose atropine. If you would like to talk through options first, you can book a free Myopia Chat with Jagrut.

References

• Hložánek M, Dostálek M, Autrata R, et al. Effectiveness of 0.02% and 0.04% Atropine Therapy for Myopia Control in Czech Children: The MARS Trial. Ophthalmol Ther. 2026. PMID: 42832165. https://doi.org/10.1007/s40123-026-01506-x