top of page

Bariatric surgery can impair midwall left ventricular dysfunction in certain patients one-year after surgery

Updated: Feb 1

Patients with severe obesity and in particular women with hypertension, higher level of inflammation and reduced ejection fraction, have impaired left ventricular (LV) midwall mechanics one year after bariatric surgery, according to researchers from Norway.

Figure 1: Forests plots showing the odds ratios (OR) with 95% CIs for the variables associated with low 1-year MWS in two logistic regression models (A and B). The risk score built on each of the logistic regression model is presented under the respective panel.

It is known that excess adipose tissue impacts every organ system and is associated with increased cardiovascular (CV) risk and subclinical cardiac dysfunction. The Global Burden of Disease Study has identified obesity as the direct cause of 11% of heart failure cases in men and 14% in women. The investigators said Bariatric Burgery on the West Coast of Norway study (a prospective follow-up of patients with obesity referred for Roux-en-Y gastric bypass surgery on the West Coast of Norway (2012–2016)), has shown that impaired cardiac mechanics in the longitudinal or radial direction is largely prevalent in middle-aged patients with severe obesity. Patients with obesity referred to bariatric surgery do not routinely undergo preoperative screening with echocardiography in Norway, despite the fact that 44% of patients that undergo bariatric surgery remain with abnormal cardiac function one year after surgery, most often due to reduced LV midwall mechanics.

Therefore, they analysed data from the Bariatric Surgery on the West Coast of Norway study to determine how preoperative inflammation, findings at the initial electrocardiography (ECG) and echocardiography and traditional CV risk factors may predict longer-term cardiac impairment. Based on the identified predictors, they subsequently sought to develop a simplified risk index to be used in the preoperative assessment of patients with severe obesity.

A total of 121 patients were recruited in the cardiac study arm, none of the participants had a history of CV disease at study inclusion. Biobank samples for analysis of inflammatory markers were available in 76 patients preoperatively. Of these, 75 patients had complete echocardiographic examinations both before and one year after surgery and constituted the present study population.


Before bariatric surgery, 45% (34/75) of patients had low midwall shortening (MWS). Despite comparable age, BMI and LV EF, the low preoperative MWS group had higher heart rate and systolic blood pressure as well as higher HbA1c and serum concentration of inflammatory proteins, particularly hsCRP, total calprotectin, the calprotectin subunit S100A9, and SAA isoforms SAA1.3 and SAA2.2. On ECG, patients with low MWS had comparable P wave (97 vs 94 msek) and PR duration (166 vs 159 msek) (p=0.09) and similar measures of ventricular de- and repolarisation. Myocardial O2 demand was high in 33% of the study population before surgery and correlated with increased serum calprotectin S100A9 (p<0.01). Patients with low preoperative MWS also had smaller LV dimensions with higher relative wall thickness and myocardial O2 demand.

The mean follow-up was 14 ± 3 months. Bariatric surgery induced a significant fall in serum hsCRP, total serum calprotectin and total SAA, as well as an improvement in both GLS and myocardial O2 demand (all p<0.01). High myocardial O2 demand was present in 7% of patients one year after surgery and in this group the preoperative serum levels of all inflammatory proteins were significantly higher than in patients with normalized O2 demand after surgery (all p<0.05).

MWS remained on average unchanged and 35% of patients had low MWS one year after surgery. Of these, 62% also presented with low MWS before surgery. Patients with low one-year LV midwall function had higher heart rate, more often hypertension and higher levels of both hsCRP, calprotectin S100A9 and total SAA preoperatively (all p<0.05).

When preoperative echocardiographic findings were compared, patients with low one-year MWS had lower EF: 59 ± 7% vs. 62 ± 5% and significantly higher myocardial oxygen demand before surgery: 1.85 ± 0.69 vs 1.58 ± 0.53 g kdyne/cm2 bpm × 106 (both p<0.05). Preoperative ECG measures of atrial and ventricular function did not differ between patients with normal and low MWS one year after surgery. Postoperatively, patients with low MWS had higher relative wall thickness, lower GLS and higher myocardial oxygen demand.

Twenty-five percent of women vs. 10% of men had persistently low MWS 1 year after surgery. At baseline, the concentration of inflammatory proteins did not differ between sexes. However, when comparing patients with preoperative BMI below or above the median (41.2 kg/m2), hsCRP was highest in women with BMI above the median (p<0.05). Surgery resulted in a reduction of hsCRP and SAA in both sexes. Compared to patients with persistently normal MWS, women with persistently low MWS had higher preoperative serum calprotectin S100A9, and men with persistently low MWS had higher preoperative serum SAA (both p<0.05).

In logistic regression analysis, low LV MWS one year after surgery was significantly associated with the following preoperative features: hypertension, female sex, low LV EF and hsCRP in the highest tertile (above 5.9 µg/ml) (p= 0.01 for the overall model) (Figure 1). In a similar regression analysis, substituting high hsCRP with high SAA (i.e. above 4.2 µg/ml), SAA was also independently related to low 1-year MWS (p=0.01 for the overall model, Figure 1).

In subsequent analyses, age, preoperative BMI or diabetes, surgery-induced reduction in BMI, as well as use of antihypertensive, antidiabetic and cholesterol lowering drug treatment did not independently predict low one-year MWS.

Two risk indexes for low 1-year MWS were developed based on the identified preoperative predictors: sex, hypertension, EF and elevated hsCRP (risk index 1) or elevated SAA (risk index 2) (Figure 1). A risk index 1 above 33 identified patients with low 1-year MWS with 81% sensitivity and 71% specificity, while a risk index 2 above 32 distinguished patients with low 1-year MWS with 77% sensitivity and 77% specificity.

The researchers added that to identify patients who require closer postoperative follow-up is therefore highly clinically relevant a simple, pragmatic risk score composed of four variables as an easy-to-use tool in the preoperative screening of these patients. They recommended that referral to a preoperative cardiac examination, stricter risk factors control and cardiological follow-up of patients with cardiac dysfunction, in particular women with obesity, hypertension and higher levels of inflammatory markers.

“One-third of patients with severe obesity and in particular women with hypertension, higher level of inflammation and reduced EF, have impaired LV midwall mechanics one year after bariatric surgery,” the authors concluded. “Preoperative assessment by combined echocardiography and measurement of circulating inflammatory proteins may be useful in the routine evaluation of patients referred to bariatric surgery.”

The findings were featured in the paper, “Preoperative risk factors associated with left ventricular dysfunction after bariatric surgery,” published in Nature, Scientific Reports. To access this paper, please click here

bottom of page