Myopia has become a global public health problem, and its incidence is growing year by year, seriously affecting the vision and visual quality of teenagers (1). It is well recognized that genetics plays an essential role in the development of myopia, but genes cannot be blamed for all myopia occurrences (2, 3). With the prevalence of electronic products and the emphasis on education, there has been a significant increase in close work time for adolescents, which may be one of the risk factors for myopia (4). In the meantime, less outdoor activity can also contribute to the development of myopia (5, 6). Another factor associated with myopia is consumption of refined carbohydrates, where prolonged high sugar intake has a significant effect on the body’s serum glucose and insulin levels (7, 8).

As the human economy improves, high-carbohydrate diets and high sugar intake become more common. The eye axis of adolescents in the developmental stage is also growing, and insulin is associated with eye axis growth in animal studies (9). It has also been suggested that a high glycemic load may increase the incidence of myopia, but the outcomes of the current researches are not entirely consistent. For this reason, we conducted a cross-sectional study with a larger sample based on the entire U.S. population to determine the relationship between insulin levels, plasma fasting glucose, and adolescent refractive status. Simultaneously, we conducted an in-depth examination of the disparities across different genders and ethnicities, with the goal of devising customized testing protocols suitable for distinct population cohorts.

The basic characteristics of the included participants were displayed in Table 1. There were 2,234 male participants, representing 52% of the total. The participants’ refractive status ranged from myopia of −20.25 D to hyperopia of +9.5 D, and the median [interquartile range (IQR)] spherical equivalent was −0.250 (−1.375, 0.125) D. The median insulin levels [interquartile range (IQR)] were 56.340 (39.480, 81.720) pmol/L for males and 64.380 (46.740, 89.880) pmol/L for females. Median fasting serum glucose concentrations [interquartile range (IQR)] were 5.218 (4.940–5.495) mmol/L and 4.994 (4.752–5.273) mmol/L for male and female participants, respectively.

In comparison to female participants, men had markedly lower levels of serum insulin, and had higher levels of plasma fasting glucose, glycohemoglobin, and total calcium. No significant differences were seen in age, race, education level, and income poverty ratio.

Table 2 demonstrated the relationship between serum insulin concentration and refractive status in the different multivariable linear regression models and stratified analyses. It was shown that higher serum insulin levels were associated with more negative refractive status. This negative correlation was more significant in men (β = −0.021, 95% CI: −0.035 to −0.007) with each increase of 10 pmol/L. This situation was also shown in Mexican American (β = −0.018, 95% CI: −0.033 to −0.002). There was no significant negative correlation between females, non-Hispanic white, non-Hispanic Black, other Hispanic and other races. In different quartiles of insulin, this negative correlation was found in Q3 (β = −0.234, 95% CI: −0.398 to −0.070) and Q4 (β = −0.284, 95% CI: −0.459 to −0.109). P for trend was less than 0.05.

To minimize the deviations, we divided the data from 1999 to 2002 into one group and the data from 2003 to 2008 into another group due to differences in methods of insulin determination. The results were presented in Table 3 and Table 4, respectively. After adjusting for covariates between serum insulin and spherical equivalent, each 10 pmol/L insulin level increase was associated with β −0.016, 95% confidence interval (CI) −0.031 to −0.002 in 1999–2002 and β −0.013, 95% confidence interval (CI) −0.026 to 0.001 in 2003–2008. The results continued to suggest a negative correlation between insulin levels and spherical equivalent. In the male subgroup, the negative correlation trend was still significant, with β values of −0.027 (95% CI: −0.049 to −0.005) and −0.018 (95% CI: −0.036 to −0.000) respectively. In Table 3, we observed a consistent negative correlation in quartiles Q2 (β = −0.404, 95% CI: −0.629 to −0.179), Q3 (β = −0.403, 95% CI: −0.635 to −0.170), and Q4 (β = −0.256, 95% CI: −0.506 to −0.006) compared to Q1 when insulin levels were stratified. In Table 4, a similar pattern was observed in Q4 (β = −0.308, 95% CI: −0.552 to −0.064). This trend remained robust and statistically significant (p for trend <0.05).

In different multivariable linear regression models, there was no significant correlation between plasma fasting glucose and refractive status (Table 5). Meanwhile, this correlation remained insignificant in the stratified analysis by gender, race, and different concentrations. We also used smoothed curve fitting to reveal the nonlinear relationship between the variables, and the results remained stable (Figures 1–3).

The primary objective of this study was to investigate the potential relationship between serum insulin and plasma fasting glucose levels with refractive status. In the present study, we used a sample of American adolescents aged 12–19 years. The outcome showed a negative association between serum insulin and spherical equivalent in males, and Mexican American, but no significant correlation in female adolescents, non-Hispanic white, non-Hispanic Black, other Hispanic and other races. Subgroup analysis was performed according to insulin measurement and found that the trend of myopia deepening with increasing insulin levels remained robust, especially in male individuals, but there were some differences in the effect between races. Also, among the adolescents included in the analysis demonstrated that fasting glucose levels were not significantly linked to spherical equivalent.

Recent studies have shown that insulin levels may have an effect on refractive status (12, 13). Feldkaemper’s study on chicks discovered that insulin promotes cell proliferation and accelerates the growth of the eye axis (9). As people’s quality of life improves, prolonged high glucose diets lead to more common hyperinsulinemia (14, 15). The chronic plasma hyperinsulin state suppressed hepatic synthesis of insulin–like growth factor binding protein-1 (IGFBP-1), thereby increasing free insulin-like growth factor-1 (IGF-1) and promoting cell and tissue growth (15, 16). Zhuang et al. found that IGF-1 polymorphism was associated with the high myopia population in the Chinese population (17). In addition, reduction of IGFBP-3 could stimulate further tissue growth through its effect on the retinoic acid signaling pathway (18, 19). The results of our study revealed different effects of insulin levels in different races and genders of participants. Individualized monitoring for adolescents may be more beneficial for myopia control and less wasteful of medical resources.

The effect of gender on refractive status is currently more controversial, with some studies finding a higher prevalence of myopia in males (20, 21), and individual studies showing opposite results (22, 23). The results of this study showed a negative correlation between spherical equivalent and insulin levels, and this performance was more significant especially in the male population. Plasma fasting glucose and glycohemoglobin were significantly higher in male individuals than in females in the study population, which may be one of the reasons why men are more susceptible (24). Also, the secretion of sex hormones can lead to differences in the response to insulin between men and women. Some studies have shown that insulin resistance predisposes to the development of myopia (25). Prior to menopause, insulin resistance occurs less frequently in women than in men of the same age. However, this protective effect disappears after menopause, suggesting that estrogen may have some protective effect (26, 27). The above study can partially explain the negative correlation between refractive status and insulin levels in the male population. Because women are less prone to insulin resistance due to the potential protective effect of estrogen, the tendency for myopia is relatively lower.

Chronic high sugar intake can directly affect the body’s plasma glucose, insulin, and glucagon levels (28, 29). Some studies suggest that high plasma glucose levels may be a risk factor for myopia (30, 31). Our multivariable linear regression results, however, showed that elevated fasting glucose levels were not significantly associated with adolescent myopia.

The present research also had certain limitations. To begin with, this research was a cross-sectional study and lacked long-term follow-up data. Second, the measurement of refractive status in adolescents was obtained using a non-ciliary muscle paralysis method, which might have some error. Methodological errors were minimized by fogging the vision and using the median of the three measured refractive values. Due to the limitations of cross-sectional studies, we could only establish correlations and not effectively establish causal relationships. Hence, we adjusted for potential confounding factors in an attempt to minimize this limitation as much as possible.

Our study found that higher insulin levels were associated with myopia, especially in male adolescents. Refractive status in Mexican Americans showed a negative correlation with rising insulin levels, but this status disappeared after subgroup analysis.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: the datasets for this study can be found at www.cdc.gov/nchs/nhanes/.

All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published. The NCHS Ethics Review Board granted approval for the conduct of NHANES and written informed consents were obtained from all participants. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin.

PH: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. JL: Conceptualization, Data curation, Writing – original draft. MH: Writing – original draft. YF: Writing – original draft. MW: Conceptualization, Writing – review & editing.