Sacral dimples are commonly observed in 1.8%–7.2% of neonates [1]. However, they may be associated with occult spinal dysraphism, which can be accompanied by neurological abnormalities, and are evaluated using ultrasonography or magnetic resonance imaging (MRI) [2,3]. Most sacral dimples have a benign course, but certain associated features increase the risk of spinal dysraphism and warrants further evaluation [4]. These features include the presence of hair at the dimple site, associated subcutaneous lesions such as hemangiomas or lipomas, dimples larger than 5 mm or deep in appearance, multiple dimples, location more than 2.5 cm above the anus, or an abnormal gluteal cleft such as bifurcation or deviation.

In Korea, a national infant health-screening program is conducted at 1 month of age [5], and the number of infants undergoing early detailed evaluation for suspected sacral dimples has been increasing. Ultrasonography is commonly used as the first-line imaging modality because it is safe, noninvasive, and does not require sedation. On ultrasonography, a tethered cord may be suspected when the conus medullaris is located below the L2–3 level or when the thickness of the filum terminale is ≥2 mm [4]. When spinal dysraphism such as tethered cord is present, it may lead to complications including chronic urinary tract infections, neurogenic bladder, lower extremity paralysis, and gait disturbances [4]. However, most simple sacral dimples in asymptomatic, healthy neonates are not clinically significant. Therefore, concerns have been raised about the routine use of ultrasonography for neonates with sacral dimples [6].

A study conducted in the United States reported that screening ultrasonography for 3,991 healthy infants with simple sacral dimples identified abnormal findings in 133 (3.4%), among whom only five (0.13%) had suspected tethered cord and subsequently underwent surgical treatment [7]. In a Korean study, only one of 304 infants who underwent ultrasonography for sacral dimples showed abnormal findings [1]. Therefore, this study aimed to determine the findings in infants who underwent ultrasonography for sacral dimples at a single institution and evaluate the clinical usefulness of screening ultrasonography.

In this study, spinal ultrasonography was performed for 588 infants with sacral dimples. Clinically significant abnormal findings were identified in only two cases (0.3%), both of which showed thickening of the filum terminale with a diameter of ≥2 mm. However, the position of the conus medullaris and movement of the nerve roots were within normal limits, and no cases were diagnosed with a tethered cord. These findings suggest that the clinical utility of ultrasonography is limited for infants with simple sacral dimples.

Most of the 20 cases with abnormal findings were filar cysts or hydromyelia, which are benign variants commonly observed in neonates [8,9]. Filar cysts are found in approximately 12% of normal neonates and have limited clinical significance, whereas hydromyelia represents transient dilatation of the central canal during the neonatal period and usually resolves spontaneously within several weeks [9].

The purpose of ultrasonography in infants with sacral dimples was to evaluate the risk of spinal dysraphism, including that of the tethered cord. Tethered cord is a form of spinal dysraphism caused by midline embryologic defects, for which the conus medullaris is abnormally attached to the sacrum by an intraspinal lipoma or thickened fibrous band. This results in a lower-than-normal position and subsequent neurological impairment [4]. Clinical manifestations may include lower extremity weakness, muscle atrophy, sensory abnormalities, and bladder dysfunction and may be accompanied by spinal anomalies such as spina bifida and foot deformities.

Reported risk factors for spinal dysraphism include folate deficiency, iron deficiency, young maternal age, exposure to teratogenic drugs, obesity, multiple pregnancies, and the use of antiepileptic drugs [10]. The global incidence of spinal dysraphism has declined with widespread folic acid supplementation, but the introduction of the 1-month infant health-screening program in Korea in 2021 increased awareness and concern about sacral dimples among caregivers and healthcare providers. Consequently, unnecessary imaging may be performed even for simple dimples. In this study, 307 infants (52.2%) were referred from pediatric clinics for suspected sacral dimples, and 73.3% were born outside hospitals. Even without formal referrals, many infants likely visited our hospital due to parental concerns or physician recommendations.

More than 50% of spinal dysraphism cases are diagnosed based on cutaneous findings such as sacral dimples [4]. However, not all sacral dimples are associated with spinal dysraphism, and differentiation based on the location of the lesion and associated skin abnormalities is necessary. Cutaneous findings suggestive of spinal dysraphism are typically located along the midline and include hypertrichosis, capillary hemangioma, subcutaneous masses (e.g., lipoma), meningocele, and caudal appendages. In addition, abnormal gluteal clefts, such as those with asymmetry or duplication, require careful evaluation. Imaging studies such as ultrasonography are recommended for cases with multiple dimples, dimples larger than 5 mm, location more than 2.5 cm above the anus, or cutaneous stigmata described above [4]. In contrast, a single dimple located within the gluteal crease is considered a normal variant, and it is not regarded as abnormal in the absence of associated findings even if it is slightly off the midline.

Gomi et al. [11] proposed a classification system dividing sacral dimples into three types based on their location and evaluated their association with spinal lesions. Type 1 dimples are located within the gluteal crease and are rarely associated with clinically significant spinal abnormalities. Type 2 dimples are located at the upper edge of the gluteal crease and are associated with a curved or deformed appearance of the crease, often described as a swanneck deformity. Type 3 dimples are situated above the gluteal crease and may be associated with duplication (Y-shaped morphology) or pigmentation. Types 2 and 3 are associated with a significantly higher incidence of underlying spinal abnormalities, including spinal lipoma, thickened filum terminale, and tethered cord and require further imaging evaluation such as MRI. This classification, based on the dimple location, provides a clearer way to assess risk than the usual simple versus complex classification.

In a Korean study, the incidence of spinal abnormalities was higher for cases with combined cutaneous lesions than for those with simple dimples alone (1.7% vs. 13.4%) [12]. Another Korean study reported that spinal abnormalities were detected in 3.4% of cases when ultrasonography was performed in infants with hair or abnormal gluteal clefts, such as asymmetry or deviation [13]. Similarly, a French study of 144 infants reported that simple dimples were associated with normal ultrasonographic findings (relative risk [RR]=0.32; P=0.0029), whereas abnormal gluteal clefts were associated with abnormal findings (RR=3.09; P=0.0029) [14].

In this study, detailed descriptions of the dimple morphology were available for only 196 cases (32.8%) due to its retrospective design. Only one of the 94 infants (1.1%) with simple dimples had fatty filum terminale, whereas three of the 99 infants with complex dimples had positive findings. These results are consistent with reports of previous studies suggesting that the clinical usefulness of ultrasonography is limited when evaluating simple dimples. However, the lack of detailed morphological descriptions for many cases limited the statistical analysis.

When spinal dysraphism is suspected, spinal ultrasonography or MRI is performed for diagnosis. Ultrasonography is widely used as a first-line screening tool because it is safe, noninvasive, and does not require sedation. However, the resolution of ultrasonography decreases as ossification of the posterior spinal elements progresses, and examination is recommended before 3–4 months of age [3,15]. In this study, ultrasonography was performed within 3 months of birth in 87.7% of the infants, whereas the 10 patients for which the conus medullaris could not be clearly identified had a mean examination age of 5.5 months.

Ultrasonography evaluates the position of the conus medullaris, thickness of the filum terminale, and position and movement of the spinal cord and nerve roots. A tethered cord is suspected when the conus medullaris is located below the L2 level, the filum terminale thickness is ≥2 mm, or movement is decreased or the spinal cord is positioned posteriorly. For such cases, further evaluation with MRI is required [4]. Some studies suggest a normal range of 1.0–1.5 mm for filum terminale thickness [16], but most studies define <2 mm as norma [l4,17].

A previous Korean study analyzing the data of 230 infants with sacral dimples reported that five infants (2.2%) had a filum terminale thickness of 2 mm, while only one infant required surgical treatment for a suspected tethered cord [1]. Despite the low prevalence of clinically significant findings, the authors recommended ultrasonography as a screening tool due to its accessibility, safety, and non-invasiveness. However, most studies suggest that ultrasonography is not useful for simple sacral dimples [6,7,12,14,18]. In a large study by Kucera et al. [7], abnormal findings were identified in 133 infants (3.4%) of 3,991 neonates with simple sacral dimples; however, none required surgical treatment. Of these, the dimples of 52 were later found to be normal on follow-up imaging. Ultimately, only 76 infants (2.1%) had abnormal findings. Among them, 71 were clinically normal on neurosurgical evaluation and did not undergo surgery, and only five infants (0.13%) required surgical intervention. These findings were supported by a literature review including 3,027 infants, of which only five (0.17%) underwent surgical treatment [6,7]. Furthermore, a systematic review of 5,166 infants with sacral dimples reported abnormal ultrasonographic findings for 3.4% of the cases; however, simple dimples alone were not predictive of clinically significant spinal abnormalities, suggesting that routine ultrasonography may be unnecessary [6].

Taken together, these findings indicate that the likelihood of diagnosing clinically significant spinal dysraphism in infants with simple sacral dimples is extremely low, raising questions regarding the cost-effectiveness of ultrasonographic screening [18]. The low diagnostic yield of 0.3% observed in our study further supports this conclusion.

This study had several limitations. First, it was a retrospective, single-center study. Second, detailed descriptions of dimple location, size, and associated findings were insufficient due to limitations of the medical records. This prevented subgroup analysis of simple versus high-risk dimples. Future studies should classify dimples based on morphological characteristics and compare ultrasonographic findings and clinical outcomes between simple- and high-risk groups. In addition, MRI was not performed at our institution for patients referred for neurosurgical consultation, and long-term follow-up was limited because of loss to follow-up or transfer to a tertiary center.

In conclusion, most infants who underwent ultrasonography for sacral dimples showed no significant abnormalities, and none were diagnosed with a tethered cord. Routine ultrasonographic screening of all neonates with sacral dimples was inefficient. Ultrasonography is recommended for high-risk cases, especially those with abnormal gluteal clefts or combined skin lesions (Gomi’s types 2 and 3). Careful physical examination is essential to identify high-risk skin signs, and unnecessary imaging is not recommended for simple dimples. Further large-scale prospective studies are required to establish clear, risk-based guidelines for ultrasound screening.