The findings showed a relatively high prevalence (39.8%) in the Chinese population and demonstrated that fabella presence and parameters are related to MM tears. The fabellar parameters varied among the patient gender, age, articulating groove, and MM tear conditions but not between the sides of the legs. Multivariate analysis showed that both fabellar presence and age were risk factors for MM tears, which indicated that the fabella should not be identified as a simple normal anatomic variant. A prediction model combining age, width, and length/thickness ratio showed good diagnostic performance with an AUC of 0.741 and might aid clinicians in identifying patients at risk for an MM tear and informing patients for their higher MM tear risk.
Fabellar prevalence and parameters varies
Fabellar prevalence ranges among ethnic groups, e.g., from 3.1% to 31.3% in Caucasian populations and from 30.6% to 92% in Asian populations [2,3,4,5,6,7,8,9, 11,12,13,14, 16,17,18,19,20,21,22,23,24,25]. Our study findings were consistent with the relatively high prevalence reported in the Asian population [2, 5, 6, 8, 9, 12, 14, 17, 22]. A previous study suggested that the form of the fabella was genetically controlled; however, related genes and pathways were not identified . The ossification of fabellae was correlated with environmental factors, such as mechanical stimuli , which was supported by an Asian lifestyle, including kneeling, squatting and tailor sitting. All of the above factors lead to persistent pressure on the fabellae [5, 14].
A controversial relationship was observed between age and fabellar prevalence. Most studies have reported that age is not correlated with fabellar prevalence [2, 3, 13, 14, 16, 21, 22]. However, a positive correlation between age and fabellar presence was observed in our study, which was similar to a large sample size study reported by Hou et al. . These controversial results were assumed to be attributed to the small sample size, which might not be powerful enough to detect such relationships, and the radiological methods they used, which could not distinguish cartilaginous fabellae.
For the fabella size, larger dimensions were recorded in males than in females, as reported previously [14, 18]. In our study, 44.5% of fabellae had articular grooves on the lateral femoral condyle, which was less than that in previous studies [7, 14]; however, we agreed with the fact that a fabella with an articular groove was more likely to be larger . These findings supported that larger fabellae might be more efficient in shifting load through surrounding structures [14, 18]. The fabellar length and thickness varied among age groups, and a trend for this pattern was discovered for width; however, no consistent correlation between age and fabellar parameters was found in our study. To assess the fabellar morphology, the fabellar length/thickness ratio, width/thickness ratio and length/width ratio were calculated. We found that the smaller the fabellae, the larger the length/thickness ratio and width/thickness ratio, i.e., the flatter the fabellae, which reflected the biomechanical advantage of larger fabellae [14, 18]. Our results further suggested that larger fabellae could be tougher against persistent pressure to maintain their original oval-shaped morphology Most previous fabella studies only estimated the prevalence of fabella, and studies with fabella measurements are limited. Only two studies [14, 22] measured the length, thickness and width of the fabella. Chew et al.  measured the fabella size of 80 patients who underwent arthroscopy and reported that the average length, thickness and width were 7.06 mm, 4.89 mm and 6.12 mm, respectively. Tabira et al.  measured the fabella size of 51 cadavers and reported that the length, thickness and width were 8.2 mm, 5.4 mm, and 9.9 mm, respectively. The size of the fabella measured by Tabira et al.  was larger than that measured by Chew et al.  and the present study. This may be attributed to the older age of the patients (74.5 vs. 28.4, 41.5 years) in the latter two studies. This result was also supported by our correlation analysis, which indicated that elderly individuals might have larger fabella. The DFI measured by Chew et al.  was 33.2 mm, which was supported by our relatively larger sample size (33.0 mm). The fabellar length/thickness ratio, width/thickness ratio and length/width ratio were only calculated in the present study. We further found that the length/thickness ratio and width/thickness ratio were correlated with MM tears. Considering the potential clinical significance of fabellar morphology, future fabella studies should report these measurements to allow further biodynamic research.
Fabella and MM tear
Previous studies have suggested that fabellae were associated with medical conditions, such as osteoarthritis [8, 26], although none of them considered fabellae as an influencing factor for knee ligaments or meniscus disorders. The present study first revealed that fabellar presence and morphology were associated with MM tears. However, the causal biodynamic mechanism needs to be confirmed by experimental study. Age was a main confounding factor in this study. Although degeneration led to both the presence of fabellae and MM tears [8, 24], we found more fabellae in knees with MM tears among patients older than 20 years. Further logistic regression showed that fabellar presence was an independent risk factor for MM tears, and an ROC analysis demonstrated that fabellar parameters influenced the predictive model, indicating that mechanisms other than age-related degeneration contributed to the association between fabellae and MM tears. A radiographic analysis considering the presence and severity of osteoarthritis might provide a chance to draw a more robust conclusion. In contrast, the OPL was a relatively consistent structure on the posterior aspect of the knee. The OPL originates from the posterior surface of the posteromedial tibia condyle, merges with fibers from the semimembranosus tendon and from the posteromedial part of the capsule, then converges and courses in a diagonal oblique course, and finally attaches to the fabella when that is present [27, 28]. As a structure involved in both posteromedial and posterolateral corners , the OPL plays a role in preventing excessive external rotation and extension of the knee [12, 28]. The fabella is considered a stabilizer during this procedure . For this purpose, forces might shift from the fabella along the OPL and separate into a horizontal force and a vertical force ; then, it could make a persistent stretch upon the fabella. As a result, its length and width are extended, and thus, a flatter morphology is shown. On the other hand, an equal but opposite force chronically influences the posteromedial corner, thereby leading to posteromedial corner disorders, which decrease the dynamic function of the MM and increase the risk of injury . This hypothesis might explain the flatter fabella seen in knees with MM tears. In knees where a fabella is absent, OPL attaches to the tendon of the lateral head of the gastrocnemius . Without forces transferred from the fabella, the OPL might sustain less force as a dynamic knee stabilizer. Further experimental analysis is needed to determine the causal relationship between the fabella and MM tears, especially the effect of OPL and other surrounding structures on the development and degeneration of the fabella.
Meniscus injury is a common disease in arthroscopic practice . Meniscus tears are the most serious type of meniscus injury. Arthroscopic meniscectomy or meniscal repair is the most recommended surgery for patients with degenerative meniscal tears or traumatic meniscal tears. Our study demonstrated that for a patient with knee pain, the patient is more likely to have an MM tear if the patient has a fabella. In clinical settings, radiography is often the first radiologic examination for patients with knee pain. If the X-ray shows that the patient has a fabella, or even a flat fabella, the patient is recommended to undergo further MRI examinations to determine whether they have an MM tear, which potentially needs arthroscopic surgery.
Okano et al.  described a patient who experienced posterolateral knee pain after total knee arthroplasty due to fabella and cured the patient with fabellectomy. Although fabella syndrome could be cured with fabellectomy, as Dekker et al.  presented recently, preoperative planning is encouraged, especially in patients with posterolateral knee pain, which includes acquiring a detailed history, assessing the symptoms, performing specific tests concerning fabellar lesions, and radiological evaluation . Intraoperatively, assessment of fabellar impingement against surrounding structures is needed to avoid postsurgical complications. Patients' complaints about posterolateral knee pain should be taken seriously as a sign of potential biomechanical imbalance other than normal postoperative phenomena, even in a patient who underwent medial meniscectomy. These considerations might help clinicians determine whether the fabella should be treated.
There are several limitations in our study. First, our cross-sectional study design was unable to make causal inferences, and only patients with knee pain were included in our study. Moreover, detailed information, such as body mass index, career, habits and anatomic risk factors, was not available for analysis due to the retrospective nature. An external validation may provide more reliable measurements of the model performance. Second, plain films were not available for all included patients. However, MRI measurements are sensitive enough for the detection of fabella [5, 8], and more adequate for detection of MM tears in this study. Third, our model was built mainly based on MRI measurements without clinical risk factors for MM tears. The model would be more practicable if clinical risk factors were included. Fourth, our study did not provide suggestions on whether patients need to be surgically treated. A cohort study with follow-up may provide insights for clinical decision-making. Last, our study did not provide experimental evidence for a causal relation between fabella and MM tears.