This study describes, for the first time to our knowledge, the prevalence of aniso-TA using both non-vectorial aniso-TA and vectorial aniso-TA in Chinese preschool children. The prevalence of vectorial aniso-TA was twice as common as non-vectorial aniso-TA, which did not vary with sex and age. The prevalence of aniso-TA was much lower than that of TA.
Prevalence of aniso-TA from previous studies on similar age population was shown in Table 4. These studies reported different prevalence rate, which might be due to different ethnicity, age, and whether vectorial analysis was used. We compared the prevalence of non-vectorial aniso-TA with previous studies defined as ≥ 1.0DC. The prevalence of non-vectorial aniso-TA in the present study was lower than that found in the Tohono O’odham Native American children, in the Northern Ireland Childhood Errors of Refraction (NICER) study and in the rural area of southwestern Japan [23–25]. However, it was higher than that in the Sydney Myopia Study, and similar to that in the Sydney Paediatric Eye Disease Study [14, 26]. Among these studies, the prevalence of non-vectorial aniso-TA found in the Tohono O’odham Native American children was the highest, in accordance with the population’s high TA prevalence [23, 27]. When compared with the Shandong Children Eye Study, which was also carried out among Chinese children, the prevalence in this study was lower [15]. Our previous study also showed the TA prevalence was lower than that from The Shandong Children Eye Study [20, 28]. Few studies revealed the prevalence of vectorial aniso-TA. The prevalence of vectorial aniso-TA in the present study was lower than that found in the Multi-Ethnic Pediatric Eye Disease Study (MEPEDS) [16]. In their study, vectorial aniso-TA was twice as common as non-vectorial aniso-TA,similar with our results. Children in MEDPEDS were African American and Hispanic, who also showed higher TA prevalence. The difference between non-vectorial aniso-TA prevalence and vectorial aniso-TA prevalence was reasonable as the later one took astigmatic axis into consideration
Table 4
Studies of aniso-stigmatism among young children.
| Author | Year | Location | Age | Sample size | Definition | Prevalence |
| The Sydney Myopia Study | 2006 | Sydney, Australia | 6 years | 1765 | ≥ 1.0DC | 1.60% |
| Dobson et al. | 2008 | Tohono O’odham, American | 4–13 years | 1041 | ≥ 1.0DC | 15% |
| The Northern Ireland Childhood Errors of Refraction Study | 2013 | Northern Ireland, England | 6–7 years | 661 | ≥ 1.0DC | 7.70% |
| 12–13 years | 389 | 5.60% |
| The Sydney Paediatric Eye Disease Study | 2013 | Sydney, Australia | 6–72 months | 2090 | ≥ 1.0DC | Overall 3% |
| European-Caucasian 1.9% |
| East-Asian 5.2% |
| South-Asian 3.6% |
| Middle-Eastern 3.3% |
| The Multi-Ethnic Pediatric Eye Disease Study | 2010 | California, America | 6–72 months | Hispanic American, 3030 | ≥ 1.0DC | 5.60% |
| ≥ 0.50 in JO/J45 | 10.40% |
| African American, 2994 | ≥ 1.0DC | 4.50% |
| ≥ 0.50 in JO/J45 | 11.90% |
| The Shandong Children Eye Study | 2015 | Shandong, China | 4–18 years | 6025 | ≥ 1.0DC | 3.70% |
| Yamashita et al. | 1997 | Rural area of southwestern Japan | 6 years | 350 | ≥ 1.0DC | 2.6% |
| 7 years | 2.3% |
| 8 years | 2.0% |
| 9 years | 3.4% |
| 10 years | 3.7% |
| 11 years | 4.3% |
Whatever definition was used, aniso-TA was associated not only with increased interocular differences in CR, but also with AL, possibly due to the relationship among aniso-TA and anisometropia. A similar correlation was reported by Huynh et al [14], O'Donoghue et al [24], Singh et al [29], and Hameshi et al [30]. These studies showed non-vectorial aniso-CA was associated with non-vectorial aniso-TA. This finding is in agreement with our knowledge that most aniso-TA of the eyes is due to corneal issues. Interestingly, we found interocular differences in ACD were associated with vectorial aniso-TA. The finding is in accordance with Hameshi et al, but contradicts with the results of the NICER Study [24, 30]. Vectorial aniso-CA and vectorial aniso-RA can only be explained by interocular differences in CR.
Genetic factor plays an important role in the development of astigmatism. Accordingly, this study revealed that parental astigmatism was a risk factor for vectorial aniso-TA. Previous genetic studies on astigmatism provided contradicting results on the genetic contribution to astigmatism. Wixson concluded that both parents seemed to play roles in determining the corneal power characteristics of the child. Early twin studies showed that the correlations between monozygotic twins and those between dizygotic twins for astigmatism were not significantly different, which indicated low genetic contribution to astigmatism. However, some other twin studies drew different conclusions, with the estimated heritability ranging from 30–60%. A meta-analysis of five Asian cohorts identifies PDGFRA as a susceptibility locus for CA [21]. Similarly, previous studies have obtained contradicting results on genetic contribution to aniso-astigmatism. Recently, a population-based twin study showed that the correlation between monozygotic twins for aniso-CA were significantly different from dizygotic twins [31]. A study in Korea found that intraclass correlation coefficients for spherical equivalent and ocular biometrics were significantly higher in monozygotic twins compared with singleton, with greater consistency and conformity [32]. However, another study did not find any significantly difference between children being twin or siblings in refractive error, corneal curvature, ACD and CCT [33].
Our study found that being twin or triple was a risk factor for vectorial aniso-CA. Silventoinen et al. found that differences between singletons and twins can persist into adult life with twins being shorter, lighter. In addition, twins demonstrate lower muscle strength than singletons [34]. Another study showed that twins had higher prevalence of prematurity or low birth weight and presented difference in gross motor, fine motor, language and social development [35]. A systematic review showed that twins tend to have worse academic outcomes and lower ratings in arithmetic, language, and reading than singleton children [36]. These studies all suggested being twin had a side effect on neurodevelopment, cognitive development and whole body development. As for ocular development, a study using optical coherence tomography showed twins had thicker RNFL [37]. Further investigations are needed to clarify the ocualr developmental difference between twin and singleton.
Our study demonstrated that children with a 5-min Apgar score < 7 had a higher likelihood of developing aniso-TA at 5- to 6- years compared to those with an Apgar score of 7–10 (within the normal range), while pre-term or post-term delivery were risk factors for vectorial aniso-CA. A previous study found asymmetrical growth restriction in perterm-born children [38]. Dubois reported structural asymmetries of perisylvian regions in the preterm newborn [39]. Additionally, several studies have revealed abnormal nervous system function in preterm born children. Michalczuk suggested that Apgar score seemed to be a predicting factor for developmental rate of brain function in children with history of prematurity [40]. Teli found that low 5-minute Apgar score in very perterm infants hindered corpus callosum microstructural development [41]. Moreover, eye growth is parellel to neurodevelopment. White matter changes were found in children with anisometropic amblyopia [42]. It has also been reported that low 5-minute Apgar score increased the risk of reduced vision in children [43]. The Sydney Myopia Study found that paternal age > 35 years was accociated with non-vectorial aniso-TA in unadjusted analyses. After multivariable adjustment, breast feeding had a significant protective association (P = 0.02) with non-vectorial aniso-TA. In our study, neither paternal age > 35 years or breast feeding was a risk factor for non-vectorial aniso-TA. To sum up, intrauterine hypoplasia and poor birth condition may be associated with asymmeric whole body development, neurodevelopment, and asymmeric visual and refractive development such as aniso-astigmatism. Further work is required to clarify the developmental mechanism behind these associations.
The strengths of this study include its population-based design, large sample size, and standardized examination protocols performed by an expert team, risk factors during pregnancy and early childhood. Our analyses are different from most previous studies by considering vectorial features of aniso-astigmatism. The limitation of this study is that some eligible children were not included into the anaysis due to missing data in questionnaire or refractive error measures. In addition, the risk factor data collected through questionnaire may be subjective and biased.