In comparison between two myopia subgroups, which agreed with our previous study8, LT was found to be one of the main differences between myopia T1DM children and myopia controls. The myopia T1DM children showed a marked increase in LT accompanied by a significantly ACD decrease than myopia controls; P was further analyzed in this study, which showed a markedly increase in myopia T1DM children compared to the myopia controls. K was still identical within the two subgroups in the present study; AL was longer in myopia controls after adjustment of age and sex. Nevertheless, the lucubrate research now revealed that the differences between T1DM patients and healthy controls in ocular parameters(ACD, LT, AL, and P) only exist in myopia subjects rather than non-myopia subjects. A question arose that why only the myopic eye and how does the ocular biometry change in the myopic eye? We could read from the comparison between two healthy control subgroups, the myopia controls had longer AL than non-myopia controls, but the eye seemed to try to refocus on the retina again as the flattened LT and smaller P appeared to contend with it. When turning to the myopia T1DM children, notwithstanding the elongated AL versus non-myopia T1DM children, the unchanged LT and P made the compensation mechanism more passive. This was the evidence that myopia was suspected to be accelerated in T1DM children. In the Atropine for Treatment of Myopia study11, less myopic progression accompanied by less increase in LT was found in atropine-treated eyes compared with placebo-treated eyes, when atropine was stopped, a marked increase in myopia and increase in LT was found compared with placebo-treated eyes. Although this was suspected to be the pharmacological effect of atropine, we still can see the relevance of LT increase and myopia increase. We have discussed the possible reasons for LT growth in T1DM children in the previous study, lens overhydration and less able to flatten due to ciliary muscles8. It was proved by MRI that the unaccommodated shapes of lenses in people with T1DM mimic the accommodated shape of lenses in people without T1DM12. It probably the very reason that the lens lacks the ability to lose power in myopia T1DM children compared to myopia controls.
When we added LT in the logistic regression model, there was no significant relationship between LT and myopia both in T1DM and Non-DM groups, for LT was correlated with P13, therefore, the association between LT and myopia seemed to be reflected in the impact of P. Despite larger P was associated with a higher risk of having myopia in the T1DM group (OR 1.348, P = 0.033), it was not a risk of having myopia in the non-DM group (OR 1.400, P = 0.187). Owing to better elasticity in non-DM lens, a more relaxed lens avoid lens power to become a risk of myopia. Gao et al. examined the ocular components before and after cycloplegia, a significant decrease in LT and backward movement of the lens appeared after cycloplegia. That means during accommodation, there were both an increase in LT and a forward movement of the lens, while the myopia eyes showed the least change in LT and lens movement compared to hyperopic and emmetropic eyes. Also, before cycloplegia, the myopic eyes held the thinnest LT and the deepest ACD compared to the other groups14. In other words, the thinnest LT with the least thickness increase, meanwhile the deepest ACD with the least lens forward movement, which means that the myopia eyes maintained the thinnest LT and deepest ACD regardless of accommodative or not, which was the same with Li et al6. Deeper ACD and thinner LT also appeared in our study of myopia controls when compared to non-myopia controls, but deeper ACD and unchanged LT were found in our myopia T1DM children compared to non-myopia T1DM children. Namely, the lens in myopia T1DM children held the ability to move backward in relaxed accommodation but cannot flatten. According to the Bennette-Rabbetts formula, ACD is a determinant of P here. However, P was smaller in myopia controls than myopia T1DM children (21.37 ± 0.16 vs 22.60 ± 0.17, p < 0.001), while showed no significant difference in comparation between myopia T1DM children and non-myopia T1DM children after adjustment of age and sex (22.58 ± 0.20 vs 22.98 ± 0.22, p = 0.203). In the logistic regression model, there was no significant relationship between ACD and myopia both in T1DM and non-DM groups. It was obvious that ACD change alone was not profound enough to down regulate P in myopia T1DM children.
The growth curves of ACD and LT were both displayed in a two-phase pattern, with the convex pointing to the opposite direction, U-shape for LT, and invert U-shape for ACD in not only myopia controls but also in myopia T1DM children. Our result agreed with the previous study of SCORM. It was speculated that the first decrease phase of LT was caused by stretching of the elongating eyeball, and the followed increase phase was by the thickness growth of the lens outpaced the stretching. Furthermore, the hyperopia children displayed a flat line and the emmetrope children showed a less concave line than myopia children15. Hence, we deduced the initial decrease of LT also attributed to the more flattened pattern in healthy myopia children. In our study, the myopia controls showed constant thinner LT comparing to the myopia T1DM group since the latter lack the ability to flatten. We inferred the ACD trend consist of several reasons, first, the two-phase pattern of the lens front surface; second, the first increase phase also caused by the backward movement of the lens in myopes, since in the SCROM study, the ACD showed an invert U-shape in children with myopia, while was constant in the children with hyperopia15; third, the increase phase caused by the growth of AL, but the unidirectional increase pattern of AL probably provided the least contribution. In our study, the additional thinning of the lens in myopia controls made the reciprocal ACD convex more profound than the myopia T1DM children.
The influence of LT and ACD would eventually lead to the influence of P. From the previous study, there was no doubt that P decrease with age in children7, 13, 16, 17. Our result agreed with Xiong et al. that in myopia controls, P decreased rapidly before 10 years and slowed down after 10 years. They also analyzed the associations of P with SE or AL in healthy Chinese children and found P showed a positive correlation with SE when SE > -5.00 D meanwhile exhibited a stronger negative association with AL in non-myopes but a weaker negative association in myopes. They made it clear in their article that before the onset of myopia, P reduces in the compensation of AL elongation to maintain emmetropia, which is a co-effect of changes in the flattening and thinning of the lens and the decline in internal power due to the gradient is compacted, but P loss might be limited, while AL growth might have no endpoint, myopia develops when the rate of AL growth outpaces the compensatory loss of P, thus the compensatory ability of P became smaller not only with increasing age for its natural development but also in myopes as AL elongating7. In our study, AL grew with age, while P declined with age only in myopia controls but showed no trend of declination in the myopia T1DM group. The same as the above P compensation theory of myopes and non-myopes, in T1DM, the thickness of the lens and the gradient refractive index of the lens were suffered from hyperglycemia, which made the lens less capable to compensate for AL growth in myopia T1DM. As we can see from two myopia subgroups in our study, with the same level of SE (p = 0.064), longer AL (p = 0.004) and smaller P (p < 0.001) were found in myopia controls than myopia T1DM group, indicated that P was more tolerant to longer AL in healthy myopes.
AL elongation contributed most of the SE progression1, it was not an exception here, the OR of AL was as high as 10.665 (p < 0.001) in the T1DM group and 11.956 (p < 0.001) in the non-DM group, compared to P, AL elongation played a more dominant role in having myopia. However, as for our knowledge, AL change was not caused by T1DM, so, we would not further discuss it in our study.
To conclude, in myopia T1DM children, though ACD deepened to a certain degree, LT failed to get thinner comparing to non-myopia T1DM children, still, the lens lacked the ability to lose power as compensation to AL growth, these were the evidence that myopia would be accelerated in T1DM children.