In this study, we first reported the relationship between skin microbiota, particularly P.acnes, IVDD, LBP, and sciatica, using two-sample bidirectional Mendelian randomization. Our research indicated that three skin microbiota were associated with IVDD, five with LBP, and five with sciatica. Specifically, IVDD was found to increase the abundance of P.acnes.
Propionibacterium acnes plays a pivotal role in maintaining skin health and homeostasis. However, they are also opportunistic pathogens closely associated with acne vulgaris[29]. Notably, Propionibacterium acnes, a species within this bacterial group, can cause infections beyond the skin. For instance, when it infiltrates intervertebral discs, it can lead to IVDD, LBP and sciatica[30]. At the same time, there is currently no direct evidence elucidating the mechanism of how P.acnes enters the intervertebral discs[9]. However, studies have revealed the existence of multiple common microbial communities shared between the intervertebral disc and the skin. This suggests that P.acnes may originate from the skin microbiota[6]. Intriguingly, despite these findings, we have yet to uncover conclusive evidence indicating that skin-derived P.acnes serves as a risk factor for IVDD. This observation contradicts the hypothesis that the presence of Propionibacterium acnes in the intervertebral discs originates solely from the skin or is the causative agent of IVDD.
Furthermore, numerous studies have proposed that the detection of P. acnes in intervertebral discs may be attributed to skin contaminants[16]. Our research adds to this discourse, revealing that IVDD promotes an increased abundance of Propionibacterium acnes on the skin through specific undisclosed mechanisms. This finding supports the contamination theory, suggesting that the high abundance of P.acnes detected in cultures may originate from the skin.
Numerous studies have demonstrated that Coagulase Negative Staphylococcus, akin to Propionibacterium acnes, is a Gram-positive, facultative anaerobic bacterium frequently detected in intervertebral disc (IVD) tissues[13, 31–32]. In our investigation, we identified ASV010 [Staphylococcus (unc.)] as a risk factor for lower back pain (LBP) and sciatica. This bacterial group likely belongs to the Coagulase Negative Staphylococcus species[19], corroborating previous research findings. Hence, Coagulase Negative Staphylococcus in the IVD may be attributed to skin microbiota. Furthermore, a study has revealed the presence of Enhydroxide and Acinetobacter in IVD tissues[6]. Our findings suggest that Enhydroxybacter species originating from different sources exert distinct effects on IVD degeneration and LBP. Unclassified Acinetobacter and Acinetobacter johnsonii from dry skin environments exhibit potential protective effects against LBP and sciatica. Notably, Acinetobacter johnsonii infrequently causes human infections and belongs to the aerobic bacterial group[33], so its colonization in IVD may be difficult. Interestingly, Acinetobacter johnsonii negatively correlates with the secretion of IL-4 and TNF-α[34], suggesting that certain skin-derived Acinetobacter species may possess anti-inflammatory properties. The genus Bacteroides, characterized as Gram-negative, non-spore-forming, and strictly anaerobic bacteria[35], poses a potential risk for colonizing the IVD, thus explaining its association as a risk factor for IVDD. Finegoldia, an anaerobic bacterium considered an opportunistic pathogen[36] associated with physiological joint infections[37], surprisingly demonstrates protective potential against IVDD and LBP in our findings.
This study also has limitations. Firstly, due to data limitations, we can only establish the association between skin microbiota, rather than the microbiota within IVD, and IVDD, LBP, and sciatica through MR. Therefore, we cannot completely deny the role of P.acnes in intervertebral disc degeneration, and previous studies have found that P.acnes in IVD may have other sources. Secondly, strict genome-wide significance thresholds (P < 5×10− 8) for some exposed data limit the number of instrumental variables suitable for analysis. Therefore, the threshold must be relaxed to P < 5×10− 6. However, this may reduce the accuracy of these results. Furthermore, our results only apply to the European population, and their applicability to other races is uncertain.