In this study, we found different baseline characteristics of obese patients undergoing transcatheter aortic valve implantation (TAVI): higher proportion of female and higher prevalence of diabetes mellitus, coronary artery disease, chronic lung disease and frailty and lower estimated glomerular filtration rate (GFR). However, similar procedural complications rate and device success were found between all body mass index (BMI) groups.
Obesity paradox in the univariate model
In the general population, as BMI increases above 30 kg/m2, the risk of developing cardiovascular disease (CVD) increases, and the risk of mortality is higher 35. Nonetheless, it is often observed that obese individuals with chronic diseases have a reduced risk of mortality compared to non-obese individuals with similar disease characteristics. This positive effect of increased BMI following these interventions has been called the “obesity paradox”. This paradox effect has been described in many different disease categories, including CVD, diabetes mellitus, and chronic kidney disease36-38. The “obesity paradox” has been reported mainly in patients with heart failure, acute coronary syndrome and following percutaneous and surgical coronary interventions[18-20，36，38]. Some studies have explained this protective effect that the soluble TNF-a receptor produced in adipose tissue neutralize the adverse effects of TNF-a on the mortality of patients with chronic inflammatory diseases such as CVD 39. In addition, higher lipoproteins in the circulation may also bind and reduce the role of lipopolysaccharides in stimulating the release of inflammatory cytokines 38.
There is also evidence for the obesity paradox in high-risk groups after TAVI 21,26. Konigstein et al. reported that after adjusting baseline characteristics, increased BMI was independently related to the improvement in survival after TAVI 21. The conclusion of the data analysis from the large FRANCE 2 Registry including 3,072 patients was also consistent with the obesity paradox after TAVI 26. In our univariate model, obesity was found to be closely related to increased survival rates, which was also in line with the obesity paradox.
Disappearance of obesity paradox in the multivariate model
However, in the multivariate model, after adjusting for different baseline characteristics including diabetes mellitus, glomerular filtration rate and frailty, the overall mortality rate showed similar between three BMI groups. That means in contradiction to these previous studies, according to our multivariable model, there is no obesity paradox after TAVI and overall mid-term survival rates are similar among all BMI groups (Fig 1). Previous studies have explained this discrepancy because baseline frailty factor was not included in their research and analysis models. The updated VARC-2 document defined frailty as slowness, weakness, exhaustion, wasting and malnutrition, inactivity, and loss of independence and emphasized that frailty is the most important feature in current risk models 29, which was found to be a strong predictor of mortality and adverse effects after cardiac surgery and TAVI 40-42. Indeed, in this study, frailty was found to be a strong predictor of overall mortality in both the uni- and multi-variable models (HR 3.825; p = 0.003; Table 4). Meanwhile, frailty was found to be present in 57.1% of obese patients vs 20.1% of normal weight patients(p=0.024), which might be a possible confounder of resisting the occurrence of the obesity paradox. In addition, obese patients in this research were significantly older, had less GFR and higher previous percutaneous coronary intervention, diabetes mellitus prevalence, which are also possible confounding factors might provides an explanation for the disappearance of protective effect of obesity on mortality in this study.
Obesity paradox of pulmonary complications
Many studies have consistently found that obese patients are more likely to develop severe hypoxemia after surgery because lung compliance declines particularly significantly in obese patients. Due to their abnormal BMI, obese patients may have some clinical features that require special management during the perioperative period, such as higher doses of medications, greater tidal volume, and higher airway pressure to ensure adequate postoperative ventilation. Therefore, the dyspnea of obese people has increased significantly. In addition, it has been shown that respiratory resistance increases in obesity and that the mechanism of ARDS may be attributed to the imbalance of anti-inflammatory and pro-inflammatory cytokines, as well as oxidants and anti-oxidants 43-46. Most obese patients suffer from chronic and excessive inflammation and oxidative stress 43,44. When pro-inflammatory signaling pathways are significantly upregulated, obese patients are prone to produce more abnormal cytokine products and acute-phase reactants. Furthermore, induction of pro-inflammatory cytokines and mediators has been proved to increase with weight gain 45,46. In addition, obesity can increase oxidative stress and reactive oxygen products, which may lead to direct damage to the cellular membranes, monocytes cellular adhesion, and the release of chemotactic factors and vasoactive substances. The above situations are especially obvious when undergoing cardiopulmonary bypass. However, in this study, although obese patients had lower PaO2/FiO2 than the normal weight patients at pre- and early postoperative time (baseline, transitional extubation and postextubation 12th hour), obesity did not cause a significant effect on severe postoperative hypoxemia and longer intubation time in our multiple regression model. One explanation for this divergence may be that patients with TAVI without cardiopulmonary bypass had not markedly activated the above-mentioned signalling pathways.
Selection bias could occur because this study was based on a retrospective single-centre, which might have limited the conclusions. Furthermore, the differences of a longer-term prognosis of patients is needed. Moreover, the administration of TAVI patients should be time-consuming and a more comprehensive evaluation of the prognostic value of BMI as a predictor of clinical results after TAVI is needed.