The Impact of Malnutrition on Healing After Arthroscopic Rotator Cuff Repair

Commentary Arthroscopic rotator cuff repair (ARCR) has been shown to reliably reduce pain and improve shoulder function. However, the retear rate following ARCR can range from 10% to 94%1 depending on tear characteristics such as size, degree of retraction, and presence of fatty infiltration. While these and other risk factors for retears have been identified in the literature1, very few can be modified in an attempt to improve surgical healing rates. One modifiable factor, preoperative malnutrition, has been shown to lead to worse outcomes and increased complications after other types of surgery2–5, but there are few reported studies on the impact of this factor on healing rates after ARCR. Shitara et al. attempted to elucidate the potential role of nutritional status as a modifiable risk factor after ARCR. The authors measured the preoperative Geriatric Nutritional Risk Index (GNRI), a useful tool for evaluating nutritional status in geriatric patients and in the surgical setting4,5, in a retrospective cohort of 143 patients aged 65 years and older who underwent ARCR and had a minimum 2-year clinical and imaging follow-up. Magnetic resonance imaging (MRI) was performed 2 years after surgery in all patients to assess rotator cuff repair healing. The GNRI was assessed as a risk factor for retears using an extensive multivariable analysis that adjusted for other confounding or independent risk factors for retears, including tear characteristics (size, retraction, fatty infiltration, critical shoulder angle, repair technique), age, serum albumin level, and presence of diabetes, osteoporosis, or hyperlipidemia. The GNRI was found to be a significant independent risk factor for retears as seen on MRI at 2-year follow-up (odds ratio [OR]: 3.39, 95% confidence interval [CI]: 1.22 to 9.41) when using established thresholds for low nutritional risk (GNRI: 92 to ≤98) versus no nutritional risk (GNRI: >98)4,5; this result was not distorted by collinearity with other independent variables and was validated by sensitivity analyses. Stratum-specific likelihood ratio (SSLR) analysis was then employed to establish data-driven GNRI thresholds (GNRI < 103, 103 ≤ GNRI < 109, and GNR ≥ 109) in the study cohort. GNRI < 103 was found to be a significant independent risk factor for retears seen on MRI at 2-year follow-up when compared with the other 2 cutoff ranges (103 ≤ GNRI < 109 and GNR ≥ 109). Shitara et al. should be commended for this well-performed study evaluating the effect of preoperative nutritional status on healing after ARCR. As stated, there are few reported studies on this topic in the published literature, and this is the first study to identify malnutrition as defined by the GNRI as an independent risk factor for retears after ARCR in patients 65 years and older. Interestingly, the degree of malnutrition that affects rotator cuff healing may be less than previously thought, as the authors found that the GNRI threshold that was a significant risk factor for retears in their study cohort was higher than the previously established threshold (GNRI < 103 versus GNRI ≤ 98). If appropriately validated, a threshold such as the one proposed by Shitara et al. has the potential to be used as target for optimizing a modifiable factor preoperatively to improve surgical outcome. This is similar to the use of hemoglobin A1c thresholds in diabetic patients to optimize glucose control prior to surgery. Despite advances in surgical techniques for ARCR, reported retear rates have remained relatively stable in the literature. Therefore, investigating modifiable factors that may influence the biology of rotator cuff healing, such as nutritional status, may be more impactful in improving rotator cuff repair healing rates over time than further improvements in the mechanical strength of the repair. Additional data are needed to build upon the findings of the study by Shitara et al., which was retrospective and had a relatively small sample size. While it is impressive that the authors brought back 80% of the patients with complete preoperative data for the minimum 2-year clinical and MRI follow-up, 52 patients were excluded from the study because incomplete preoperative data prevented calculation of the GNRI. Larger, prospective studies, including those aiming to improve the GNRI preoperatively to directly assess its impact on rotator cuff repair healing, are needed in the future to determine if the GNRI or other nutritional measures should be routinely obtained to assess healing potential in patients undergoing ARCR.