Ethical Clearance was obtained with the approval and consideration of Komisi Etik Penelitian Kesehatan (KEPK) of the Faculty of Medicine, Diponegoro University and RSUP. Dr. Kariadi Semarang (No.16188/EC/KEPK-RSDK/2024). All data obtained by researchers was kept confidential and used for research purposes. Research subjects who chose not to continue the study did not receive any consequences. All patients or parents signed the informed consent forms.

A total of 85 patients (43 males, 42 females) with OI from eight different centres across Indonesia were investigated. There were 58 patients from the centre in Semarang, seven from Surabaya, seven from Makassar, five from Surakarta, four patients from Palembang, two from Aceh, one from Riau, and one from Papua. Table 1 presents data on patient’s gender, age distribution and maternal pregnancy history (see Table 1). This study showed that the most prevalent age group was 0-5 years, followed by 6-10 years and 11-15 years. Low birth weight (<2500 g) was detected in 25 patients (31.25%). Family history is seen in 25 patients (30.9%). Of those, 5 patients (6.1%) were from consanguineous marriages. Among the maternal risk factors associated with OI, preterm pregnancies are associated with the highest prevalence, with an incidence rate of 19 (23.75%). This was followed by the history of miscarriages, which has an incidence rate of 13 (15.6%), and advanced maternal age (over 35 years), with an incidence rate of ten (12.5%).

Among the various clinical manifestations of OI, the most frequently observed was fractures at various sites, affecting 88.75% of patients. Fracture timing was recorded for 80 participants. Most fractures occurred between the ages of 1 and 5 years (36.25%), followed by the first year of life (33.75%), in-utero (12.5%), and after age 5 (6.25%). Meanwhile, 11.25% reported never experiencing a fracture. The femur was the most common fracture site (33.3%), followed by the cruris (24.4%), humerus (17.9%), antebrachii (3.8%), and multiple locations (8.9%). Lower extremity deformities were the most prevalent (83.1%), followed by upper extremity deformities (53.0%), sternal deformities (23.8%), and vertebral deformities (20.2%). Ambulation status showed 27.7% of participants could walk normally, while others required ambulatory aids (21.7%), used a wheelchair (14.5%), were limited to sitting (22.9%), or were confined to lying (13.2%) (see Table 2).

As for the extraskeletal clinical manifestations, blue sclera is seen in 84.7% of patients, followed by dentinogenesis imperfecta at 35.4%, myopia at 3.8%, and hearing loss at 3.8% (see Table 2).

From 85 patients, we were only able to obtain radiological data from 34 patients. However, based on available data, bowing of long bones was found in 97% of patients, followed by fractures in 93.5% of patients. Osteopenia was present in 82.1% of patients, wormian bones were observed in 45.8%, and cortical thinning was noted in 44.4% (see Table 3). Of the 27 patients (14 males and 13 females) who underwent evaluation for cortical thinning, 12 were found to have cortical thinning. Among these 12 patients, 58.3% were male and 41.6% were female.

Fractures in our cohort predominantly involved the femur and cruris, consistent with findings from other populations [18,21]. This contrasts with fracture patterns commonly seen in healthy children from an epidemiological study in the UK, where the upper limbs are more commonly affected [27,28]. This difference reflects the underlying collagen defect in OI, which leads to impaired bone strength and structural integrity, particularly affecting weight-bearing bones [3,29]. Additionally, a study from Spain suggested that fracture patterns may vary according to the disease severity, with femoral fractures becoming more prominent in moderate to severe forms of OI [30].

The prevalence of spinal and chest wall deformities in our cohort was lower than that reported in other studies [18,21,24]. This may indicate younger patient population, as these deformities tend to progress over time [30,31]. Upper and lower limb deformities were significantly less frequent in Italy than in our study. In addition, ambulation rates in our study were also lower compared to other populations. These findings suggest differences in disease severity and access to early intervention. Bisphosphonate therapy, which is widely used in children with OI, has been shown to improve bone mineral density and reduce fracture risk, potentially limiting the progression of skeletal deformities [32]. In our setting, bisphosphonate therapy is part of the standard management of OI and widely administered. However, its effect on long-term functional outcomes remains variable, likely because it primarily targets bone resorption and does not address the underlying defects in collagen synthesis and bone matrix integrity [33]. Furthermore, differences in timing of initiation may influence clinical outcomes. However, detailed data on treatment patterns were not systematically collected in this study, which limits further interpretation.

Blue sclera remained one of the most consistent clinical manifestations of OI across populations (see Table 4) with other local studies from Jakarta reporting similar trends [10,11], supporting its role as a reliable phenotypic marker despite regional variation. In contrast, dentinogenesis imperfecta showed variable prevalence across populations (see Table 4) [10], suggesting possible differences in phenotypic expression. The low prevalence of hearing loss is likely because hearing loss typically manifests after the age of 30, whereas our cohort included participants mostly under the age of 15 [33].

Radiographic data were available for 34 out of 85 patients, as imaging was not consistently performed in all cases due to variations in accessibility and completeness of medical records in each centers. Based on the available data from the medical database, the most prominent radiographic features observed were bone bowing and fractures, followed by osteopenia in our population. These findings align with the key characteristics of OI and are pathognomonic when seen in combination [34].

Wormian bones were observed in nearly half of our patients, consistent with a study in Montreal reporting 58% of their patients had Wormian bones. Notably, their study also indicated that the presence of Wormian bones is strongly associated with more severe phenotypes [35].

Cortical thinning was observed in a substantial proportion of patients, with higher prevalence in males. This contrasts with a previous study, which reported that male mice exhibit significantly greater cortical bone thickness than their female counterparts, attributed to differences in growth hormone levels and sex steroids. This discrepancy may be due to the fact that the patient sample in this study consists mainly of younger patients, where puberty tends to start earlier in girls, at around 8-13 years old, than in boys, at around 9-14 years old and at puberty, GH tends to increase [36,37].

A key limitation of this study is the potential for recall bias, as a large proportion of data were collected through questionnaire-based evaluation. Participants may have difficulty accurately recalling past medical events. Additionally, patients included in the study may have been selected from centres with the availability of pediatric endocrinologists, potentially excluding those with undiagnosed OI. Survival bias may have also influenced the findings as individuals with a more severe OI may not survive to be included in the study. Another challenge in this study is the difficulty in classifying OI accurately, as many patients did not attend regular clinical follow-ups. Given the genetic heterogeneity of OI, reliance on clinical and radiological manifestations alone may introduce diagnostic uncertainty. Incomplete radiographic data may introduce selection bias, as patients with available imaging may not be fully representative of the overall cohort. The relatively small sample size in comparison to the broader Indonesian population [8] presents another limitation, as it remains uncertain whether the findings are fully representative of the population at large. In addition, the descriptive design of this study limited the ability to perform comparative statistical analyses between subgroups, which may be explored in future studies with larger sample sizes. Despite these limitations, it is important to note that this is the largest study reporting OI conducted in the region, providing valuable insights to OI within the country.