Start low, step up? What an uncontrolled atropine escalation study can and cannot show

Parents often ask whether a child should start on the lowest concentration of atropine and move up only if the eyes keep changing, or start higher from day one. A prospective study from Beijing, published in Ophthalmology Science in 2026, set out to explore the first approach. It is also a lesson in what an uncontrolled, response-allocated study can and cannot show. This review is based on the full text, although the tables and the funding and disclosure footnotes were not available to us.

What the study did

The Stepwise Increase in Concentrations of Atropine Treatment for Myopia Control study enrolled 150 children aged 8 to 14 years (mean age 12.02 years) with myopia between -1.00 D and -6.00 D at a single hospital. Every child used 0.01% atropine daily in both eyes for 6 months. What happened next depended on each child's own result:

• progression of less than -0.25 D: stay on 0.01% (77 children, 52.74%)
• progression between -0.25 D and -0.375 D: switch to 0.02% (46 children, 31.51%)
• progression over -0.375 D: switch to 0.04% (23 children, 15.75%)

The children were then followed for a further 12 months, with cycloplegic refraction and axial length (measured by optical biometry) at 6, 12 and 18 months. Of the 146 who completed the first phase, 13 did not finish, leaving 133 in the main analysis. There was no placebo group and, importantly, no group of faster progressors who simply stayed on 0.01%.

What it found

In the first 6 months the groups differed by design: refraction changed by -0.03 D, -0.31 D and -0.53 D in the children who went on to 0.01%, 0.02% and 0.04% respectively. Those figures reflect the assignment rule, not a treatment effect.

Over the following 12 months, the mean change in refraction was -0.31 D (0.01%), -0.21 D (0.02%) and -0.18 D (0.04%). After adjusting for baseline refraction, age and sex, the figures were -0.30 D (95% CI -0.39 to -0.20), -0.23 D (95% CI -0.35 to -0.12) and -0.18 D (95% CI -0.36 to 0.01). A 95% confidence interval (CI) is the range within which the true average plausibly lies; narrower is more precise. The interval for the 0.04% group, with only 23 children, is wide, and the three intervals overlap.

So the faster progressors appeared to slow after the concentration went up. The design does not allow us to conclude that the higher concentration did it.

Why sorting children by their own result is a problem

Regression to the mean. A 6-month change in refraction is a noisy number. It mixes real eye growth with measurement variability. If you select the children with the largest change in one period, some of them are there partly by chance, and their next measurement will tend to sit closer to average whatever you do. The same logic runs in reverse. The children kept on 0.01% were selected for having almost no change (-0.03 D), and over the next 12 months they changed by -0.31 D on the same drop. Nobody would say 0.01% atropine stopped working for them.

Natural slowing with age. Myopia progression generally slows as children get older, and this cohort was already fairly old for a myopia control study. The authors note that older age is associated with slower progression regardless of treatment.

The authors estimate, from an indirect analysis, that regression to the mean accounted for less than 20% of the slowing. The paper also states that a non-escalation control group is required to know. We agree with the more cautious statement.

Refraction versus axial length

Axial length is the more objective measure here. In the first 6 months, the eye grew by 0.09 mm, 0.11 mm and 0.16 mm in the 0.01%, 0.02% and 0.04% groups. From 6 to 18 months, growth was 0.15 mm, 0.13 mm and 0.14 mm, with adjusted values of 0.15 mm (95% CI 0.12 to 0.18), 0.13 mm (95% CI 0.09 to 0.17) and 0.14 mm (95% CI 0.08 to 0.20).

After the switch, the three groups grew at very similar rates. That is compatible with escalation bringing faster progressors back into line. It is equally compatible with similar children drifting back towards the same average. Without a comparison group, the data cannot separate those explanations.

Side effects

No serious adverse events were attributed to atropine, and no child had to stop or reduce the concentration. Light sensitivity was reported by 5.2% of the 0.01% group, 6.5% of the 0.02% group and 13.0% of the 0.04% group. Pupil size increased by about 1 mm on 0.02% and by less than 1.5 mm on 0.04%.

Limitations of this evidence

• Uncontrolled. No placebo arm and no non-escalation arm, so the effect of stepping up cannot be isolated.
• Response-allocated, not randomised. Groups were defined by each child's own 6-month result, which builds regression to the mean into the comparison.
• Small higher-concentration group. Only 23 children moved to 0.04%.
• Short follow-up. Twelve months after the switch, with no data on what happens when drops are stopped.
• Completers analysis. The main analysis used the 133 children who finished and required more than 80% compliance, though an analysis of all 150 gave similar figures.
• Single centre, older cohort. Results may not transfer to younger children.
• Not checked by us. The funding and conflict of interest statements were not in the text we retrieved.

The authors describe the study as exploratory and hypothesis-generating only, and call for a randomised trial.

What we do with this in practice

This paper does not tell us whether stepping up works. It does show that side effects at 0.02% and 0.04% were mostly mild. For the question of whether higher concentrations slow eye growth more, randomised evidence is the better guide: see our review of the low-dose atropine trials and this week's dose-response meta-analysis. Stopping matters too: see our post on rebound after myopia treatment.

One fast 6-month interval, on its own, is weak grounds for concluding a treatment has failed, and a slower interval after a change is weak proof that the change worked. Axial length trends over several visits are more informative. Atropine is a prescription medicine, and concentration and suitability are something to discuss with your optometrist. In Hamilton, the optometrist team at Rose Optometry provides myopia control care, and the team offers a free Myopia Chat for families who want to talk through the options. The optometrists at Rose Optometry are Anjali Hira, Emilie Lawson, Jacqueline Rowe, Jagrut Lallu, Jason Shen, Jessica Wood and Stella Wong, and myopia management is part of their clinical practice.

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.

Reference

• Wei S, An W, Sun Y, Liang X, Bai W, Cai Z, Zhao S, Li SM, Wang N. Stepwise Increase in Concentrations of Atropine Treatment for Myopia Control Study. Ophthalmology Science. 2026;6(10):101334. PMID 42666917. https://doi.org/10.1016/j.xops.2026.101334