In the research article, we firstly found that GDF 5 rs143383 polymorphisms affect the risk of knee osteoarthritis in Caucasian but not in Asian, including detailed data from 16 studies in 7997 cases and 12684 controls. The final results are new observations in previous studies. When carrying on the funnel plot to analyze some bias, the results still are very stable and reliable.
GDF5 is a member of bone morphogenetic proteins (BMP) family, which is located on chromosome 20q11.2 and spans 21.43 kb, from 34042573 to 34021146 and can form the earliest markers of joint morphology. BMP is an indispensable signal pathway molecule or protein in most kinds of bones[29-37]. There are many cells expressing the GDF5 gene, including articular cartilage, articular capsule, and ligament. GDF5 gene mutation may result in the down-regulation of the transcription activity of articular chondrocytes[38-40].Decreased GDF5 levels in fully formed adult knees may also influence OA risk by impairing homeostasis in healthy joints or by accelerating degeneration due to injury[41]. The growth differentiation factor 5 gene GDF5 was one of the first reported OA susceptibility signals that showed consistent association to OA, with the transcript single nucleotide polymorphism (SNP) rs143383 demonstrating association in Asians and Europeans[42].
On the one hand, previous literatures have said that GDF 5 rs143383 polymorphisms C allele is a protective factor for the susceptibility of knee osteoarthritis among Caucasian populations (OR=0.74, P<0.001) and Asian populations (OR=0.87, P=0.004) [9]. On the other hand, Sujie Zhang et al holds that interaction GDF 5 rs143383 polymorphisms T allele increases the risk of knee osteoarthritis among Asian (OR=1.62, P<0.001) [5]. But now our research shows that GDF 5 rs143383 polymorphisms are only relate to knee osteoarthritis among Caucasian populations by subgroup analysis, not Asian populations.
Although previous meta-analysis literatures think that high-expressed GDF 5 rs143383(C/T) can reduce the risk of knee osteoarthritis. Their limitations still remain. Firstly, Feng Pan’s meta-analysis has some shortcomings, including unmatching HWE studies in their meta-analysis[9], which results are unreliable. Secondly, Xin Huang’s meta-analysis explores the association between all kinds of osteoarthritis and GDF 5 rs143383 polymorphisms, which can’t precisely describe the relationship between GDF 5 rs143383 polymorphisms and the susceptibility of knee osteoarthritis[12]. Although genetic variants can have different effect sizes in different populations, as is shown for GDF5 for African populations vs Eurasian populations[44, 45], it is highly unlikely that this is the case, based on the data presented by the authors. These studies focused on apes, so there is still a gap between apes and humans. In the discussion, authors state that future studies should include more samples and examine more genetic variants. We not used freely available data from large population association studies (GWAS) which have been published [46-49], because data of the sample size of the five models cannot be extracted. The meta-analysis mainly focused GDF5 rs143383,and noother variants of GDF5, because our aim is to solely clarify controversial GDF5 rs143383.
To sum it up, our final results, paralleling to previous results, are morereliable by comprehensive collection and assessment of quality. Furthermore, the association between osteoarthritis and GDF5 is more clear than previous studies.