Human biological variation in sesamoid bone prevalence: the curious case of the fabella

Abstract

The fabella is a sesamoid bone located in the gastrocnemius behind the lateral femoral condyle. In humans, fabellae are 3.5 times more common today than they were 100 years ago, with prevalence rates varying between and within populations. In particular, fabellae have been assumed to be more common in Asians than non-Asians, equally common in men and women, potentially more common in older individuals, and bilateral cases (one per knee) appear to be more common than unilateral ones. The roles of genetic and environmental factors in this phenotypic variation have been hypothesized, but not rigorously investigated. Given its clinical and evolutionary significance (i.e. being associated with several knee ailments, causing medical issues on its own, interfering with medical devices, and being less common in humans than in other mammals), it is important comprehensively to understand prevalence rate variation, and the roles of genetics and environmental factors in that variation. To address these questions, we performed a meta-analysis on data from studies published from 1875 to 2018 to investigate possible variation in sexual dimorphic (n = 22 studies, 7911 knees), ontogenetic (n = 10 studies, 4391 knees), and global (n = 65 studies, 21 626 knees) fabella prevalence rates. In addition, we investigated what proportion of cases are bilateral (n = 37 studies, 900 individuals), and among unilateral cases (n = 20 studies, 204 individuals), if fabellae are more common in the left or right knee. Our results show that, today, fabellae are 2.47-2.60% more common in men than women, and prevalence rates increase ontogenetically in old age (i.e. 70 years old), implying that fabellae can ossify early (i.e. 12 years old) or late in life. Approximately 72.94% of cases are bilateral, and among unilateral ones, fabellae are equally common in right and left knees. There is marked regional variation in fabella prevalence rates, with rates being highest in Asia, followed by Oceania, South America, Europe, Middle East, and North America, and lowest in Africa. Worldwide, an average of 36.80% of knees has ossified fabellae detectable by dissection. These results imply that, while the ability to form a fabella may be genetically controlled, the mechanisms that trigger fabella ossification may be environmentally controlled. What these environmental factors are, can only be speculated.

Keywords: bilateral/unilateral; fabella; ontogeny; prevalence; regional/global variation; sexual dimorphism.

Figure 1

Sexual dimorphic effects of prevalence rates (n = 22 studies, 8066 knees). Prevalence rates increase with time, and are higher for dissection‐based studies than for X‐ray‐based ones, possibly because the dissection‐based studies include some cartilaginous/less dense, ossified fabellae. Raw data are scattered on top of the violin plot, which was constructed by resampling the posterior distribution and creating a hypothetical dataset of prevalence rates.

Figure 2

Ontogenetic effects of prevalence rates (n = 10 studies, n = 4391 knees). Prevalence rates increase with age, and are higher for dissection‐based studies than for X‐ray‐based ones, possibly because the dissection‐based datasets studies include some cartilaginous/less dense ossified fabellae. The median prevalence rate is represented by the black solid line, and is framed by the 50%, 75%, 95%, and 99% confidence intervals. Raw data are plotted as a scatter. Sample sizes are given under the age range in terms of number of knees per age range.

Figure 3

Frequency of unilateral and bilateral cases (n = 37 studies, 900 individuals). Raw data are scattered on top of the violin plot, which was constructed by resampling the posterior distribution and creating a hypothetical dataset of prevalence rates. The median frequency of unilateral and bilateral cases is 27.0% and 72.94%, respectively.

Figure 4

Within unilateral cases (n = 20 studies, 204 individuals), fabellae are equally likely to be found in right and left knees (P = 0.9992). Raw data are scattered on top of the violin plot, which was constructed by resampling the posterior distribution and creating a hypothetical dataset of prevalence rates.

Figure 5

Fabella prevalence rates for 1875 and 2018. Black and grey lines indicate prevalence rates using X‐rays and dissections, respectively. Prevalence rates are given for each region and worldwide. Worldwide estimates were calculated using average effect of region. Datapoints are medians and error bars are 95% confidence intervals. Sample sizes are given in number of studies (s) and knees (k).

Figure 6

Flow chart depicting how fabellae form. Ability to form a fabella refers to the genetic component of fabella growth and development.