Preoperative low absolute and relative handgrip strength as predictors of postoperative short-term outcomes: a prospective study based on patients aged 60years and older with gastric cancer.

Abstract

This study aimed to determine the association of absolute and relative HGS indicators with postoperative short-term outcomes in older patients with gastric cancer. The prospective, single-center, observational study involved 230 patients (age ≥ 60years) who underwent radical gastrectomy. Absolute HGS was directly obtained from results measured by a hand dynamometer, while relative HGS was calculated as the value of absolute HGS divided by body mass index (BMI), appendicular skeletal muscle mass index (ASMI) and age, respectively. Low absolute HGS was defined according to the criterion in AWGS 2019 consensus and the first gender-specific quartile. Low relative handgrip strength was defined if it was below the first gender-specific quartile of the distribution. Postoperative outcomes included total complications, malnutrition and length of hospital stay. Univariate and multivariate regression models were performed to investigate and compare the predictive values of different HGS indicators. For the five HGS indicators, multivariate analyses showed that low absolute HGS (AWGS consensus), low HGS/BMI and HGS/age were independent risk factors for both postoperative total complications (absolute HGS(AWGS consensus): OR 2.03, 95%CI:1.05-3.93; HGS/BMI: OR 2.12, 95%CI 1.05-4.28; HGS/age: OR 2.79, 95%CI 1.40-5.54) and malnutrition (absolute HGS(AWGS consensus): OR 2.01, 95%CI 1.10-3.67; HGS/BMI: OR 2.28, 95%CI 1.20-4.33; HGS/age: OR 2.70, 95%CI 1.42-5.14). Low absolute HGS (quartile) was an independent risk factor for malnutrition (OR 1.96, 95%CI 1.04-3.71). Moreover, Low HGS/age was associated with lengthened postoperative hospital stay (OR 2.07, 95%CI 1.12-3.85). Except HGS/ASMI, both absolute and relative HGS indicators were associated with postoperative short-term outcomes. Particularly, HGS/age revealed relatively better predictive value for the studied outcomes.