Laparoscopic Gastric Banding Prevents Type 2 Diabetes and Arterial Hypertension and Induces Their Remission in Morbid Obesity: A 4-year case-controlled study

Since June 1996, LAGB (LAP-BAND; Inamed, Santa Barbara, CA) has been performed at Istituto Clinico Sant Ambrogio, Ospedale San Paolo, and Ospedale San Raffaele, in Milan, Italy, in patients with morbid obesity (i.e., grade 3 obesity as determined by World Health Organization criteria) (19,21,28).

In this study, we considered obese patients who were recruited in 1999 and followed for 4 years. According to the common protocol approved by the local ethics committees, patients were eligible for LAGB if they met the following criteria: age 18–66 years inclusive and BMI >40.0 kg/m² alone or >35.0 kg/m² in the presence of comorbidities (19,21,28). The clinical protocol, including a psychological and psychiatric evaluation, has been previously described in detail (21,29,30). Two 75-g OGTTs were performed in all patients, with blood glucose levels being determined at 0 and 120 min; glucose tolerance was labeled as normal (NGT), impaired (IGT), or diabetic (31). Blood pressure was measured twice on each occasion using a sphygmomanometer with an adequate cuff and after an adequate resting period. Patients were considered to be hypertensive if their systolic/diastolic blood pressure was >140/90 mmHg (World Health Organization criteria) or when they were taking stable antihypertensive medication.

After their diagnostic workup was completed, 73 eligible patients (12 men and 61 women, age 47.5 ± 1.21 years, BMI 45.9 ± 0.88 kg/m²) underwent LAGB; 49 eligible patients (19 men and 30 women, age 50.0 ± 1.83 years, BMI 45.2 ± 1.03 kg/m²) refused LAGB for personal reasons but continued in the follow-up study and were considered as the control group (No-LAGB). All surgical procedures were performed by two staff senior surgeons. The clinical and dietary follow-up protocol has been described elsewhere (21).

Obese patients were divided according to the absence (primary intervention study) or presence (secondary intervention study) of type 2 diabetes (Table 1). BMI, A1C, and glucose tolerance (determined by OGTT) (31) were evaluated at yearly intervals. Arterial hypertension was scored as present or absent; during the 4-year follow-up study, obese hypertensive subjects were instructed to continue their antihypertensive treatment unless symptomatic hypotension (withdrawal of treatment) or worsening of arterial hypertension (increase of treatment) occurred.

Blood glucose, insulin, and A1C levels were measured as previously described (21,32). Whole-body resistance and reactance (capacitance) were measured using a tetrapolar bioelectrical impedance analyzer (SoftTissue Analyzer; Akern Bioresearch, Firenze, Italy); data were analyzed with the Bodygram software for Windows 95–98-NT (33).

Data are given as means ± SE or frequencies. Patients were considered as long as they remained in the study. Two-sided t tests, repeated-measures ANOVA, reporting F for the interaction between time and groups, and χ² tests were used to analyze the differences between the groups at baseline and during follow-up. Survival curves were calculated to estimate the cumulative incidence and the remission of type 2 diabetes and arterial hypertension. The difference between the groups in the incidence and remission of diabetes and hypertension was tested by means of the two-sided log-rank test. All analyses of end points were based on the intention-to-treat principle. Stata’s survival analysis (Stata Statistical Software, Release 8.0; Stata, College Station, TX) was used to derive the basic estimates, such as the Kaplan-Meier product-limit survivor function, testing equality of survivor functions, and incidence rate comparisons. The proportional hazard assumption was tested by the Schoenfeld option. Only six patients (4.9%) had censored times. To estimate the incidence as well as the remission of diabetes and hypertension, subjects were scored at the yearly follow-up with 0 if the condition was not present or 1 if it was. For each subgroup, the proportion of subjects who developed (or recovered from) diabetes and hypertension was calculated. P values <0.05 were considered statistically significant.

Stoma regulation was required in 43 of the 73 patients (1–10 regulations/patient; 1.8 ± 0.24 in the entire group). Reintervention was required in eight patients (laparotomic in six patients) due to slippage and pouch dilation. Removal of the LAGB was performed in four patients: one at 1 year, one at 3 years, and two at 5 years, i.e., after completion of the study.

The clinical and metabolic variables of patients undergoing surgery (LAGB group) and refusing surgery (No-LAGB group) did not differ at the subjects’ entry into the study (Table 1), except for a slight difference in the sex ratio in the secondary intervention study. By the 2nd year of the study, six No-LAGB patients were lost to follow-up.

In the primary intervention study, body weight, A1C, and systolic and diastolic blood pressure significantly decreased in the LAGB but not in the No-LAGB group (Table 2). Figure 1 shows that type 2 diabetes developed during the 4-year follow-up period in none of the LAGB patients (0.0%) but in five of the No-LAGB patients, all of whom had IGT at baseline (17.2%). Figure 1 also shows that arterial hypertension developed in one patient in the LAGB group (2.5%) and in eight patients in the No-LAGB group (27.6%). Kaplan-Meier survival estimates showed that type 2 diabetes and hypertension appeared at a rate of 4.0 and 16.0 per 100 person-years, respectively, in the No-LAGB patients and 0.0 and 1.0 per 100 person-years, respectively, in the LAGB patients.

In the secondary intervention study, body weight, A1C, and systolic and diastolic blood pressure significantly decreased in the LAGB group but not in the No-LAGB group (Table 2). Remission of type 2 diabetes occurred in 45% of LAGB patients and 4% of No-LAGB patients; remission of arterial hypertension occurred in 51% of LAGB patients and 4% of No-LAGB patients (Fig. 1). Kaplan-Meier survival estimates showed that type 2 diabetes and arterial hypertension decreased at the rate of 1.2 and 1.8 per 100 person-years, respectively, in No-LAGB patients and 26.7 and 20.2 per 100 person-years, respectively, in LAGB patients. Figure 2 shows changes in BMI, A1C, systolic and diastolic blood pressure, fat-free mass, and fat mass in the whole cohort of LAGB and No-LAGB patients during the 4-year study. BMI, A1C, systolic and diastolic blood pressure, and fat mass stably and significantly decreased in the LAGB as compared with the No-LAGB group; fat-free mass showed a temporary, but not significant, decrease in the LAGB group and was unchanged in the No-LAGB group.

In the LAGB group, from baseline to the 4th year of the study, glucose tolerance deteriorated (from NGT to IGT) in 1 patient (1.4%), improved (from IGT to NGT and from type 2 diabetes to IGT or NGT) in 20 patients (27.4%), and remained unchanged in 52 patients (71.2%). In contrast, in the No-LAGB group available for follow-up (n = 43), glucose tolerance deteriorated (from NGT to IGT or from IGT to type 2 diabetes) in 8 patients (18.6%), improved (from IGT to NGT, and from type 2 diabetes to IGT or NGT) in 4 patients (9.3%), and remained unchanged in 31 patients (72.1%). The difference between the groups was statistically significant (χ² = 14.65, P = 0.0007).

Arterial hypertension appeared de novo in 1 LAGB patient (1.4%; plus transiently in another LAGB patient) and in 11 No-LAGB patients (25.6%); disappeared in 15 LAGB patients (20.5%) and 1 No-LAGB patient (2.3%); and remained unchanged in 57 LAGB (78.1%) and 31 No-LAGB (72.1%; χ² = 21.97, P = 0.0001) patients.

In this study, the weight loss obtained through LAGB prevented type 2 diabetes in morbidly obese patients for at least 4 years. LAGB also prevented the incidence of arterial hypertension and actually decreased the prevalence of hypertension already present. In patients with type 2 diabetes at entry into the study, the remission of type 2 diabetes and arterial hypertension was more frequent in the LAGB than in the No-LAGB group. In both the primary and the secondary intervention studies, the pattern of body weight, A1C, and arterial blood pressure was significantly different between the LAGB and No-LAGB groups. Therefore, these data suggest that LAGB is effective in preventing and promoting the remission of established type 2 diabetes and arterial hypertension.

An important concern in massive weight loss is that, together with fat mass, fat-free mass might also be reduced (33). We evaluated body composition through bioelectrical impedance analysis, a method that has a reliability comparable with that of other methods, such as anthropometric measurements and deuterium oxide dilution (34). As previously shown for biliary diversion (35), we observed an impressive reduction of fat mass and a transitory and not significant reduction of fat-free mass.

In some studies, BPD and, to a lesser extent, GBP have been shown to be more effective than LAGB in inducing weight loss and promoting the remission of type 2 diabetes and hypertension (20–25). However, these studies used different criteria to establish decreased prevalence of type 2 diabetes, measuring efficacy with either fasting blood glucose and A1C levels (20,22–24) or OGTTs (21). In this study, we showed that LAGB is effective in preventing type 2 diabetes, as has already been described for GBP and vertical-banded gastroplasty (26,27). In the Swedish Obese Subjects (SOS) study (27), the presence or absence of type 2 diabetes was verified only through patient self-report, which, without a direct assessment with OGTT, does not allow improvements in glucose tolerance to be ascertained. The difference between LAGB and GBP is that the former is a minimally invasive surgical procedure, is reversible, and has a negligible surgery-related morbidity and mortality (20–27). LAGB requires some stoma regulations and occasional reinterventions, and a few patients even require or ask for the removal of the banding (21,22,25).

With regard to the prevention or improvement of arterial hypertension, we noted remission of arterial hypertension in a fair proportion of patients, similar to that reported in other studies (20–25). In contrast, the SOS study showed that gastric surgery prevented arterial hypertension at 2 but not at 8 years (27). This difference is likely due to the crucial role of body weight in the disease; in the SOS study, a significant regain of body weight was observed in surgery patients, whereas in the present study, weight loss remained stable at ∼20% in the primary intervention study.

Similar to other studies involving surgery, this study was not randomized; however, all our patients underwent the same diagnostic workup and made the decision to undergo or refuse LAGB. Randomized, but not blinded, surgical studies could be both feasible and ethical and would further support these results. Before this study, evidence that bariatric surgery is beneficial for type 2 diabetes and hypertension was available only for gastric bypass, vertical-banded gastroplasty, and biliary diversion (20,23–25).

Pharmacological and lifestyle modifications can prevent type 2 diabetes (8–12). The limit to this approach is that these approaches work as long as they are applied, and it can be questioned whether they prevent or simply postpone type 2 diabetes (11,13). Another difference between pharmacological and lifestyle modifications compared with surgery is the efficacy of type 2 diabetes prevention. First, it seems that the annual incidence rate of new cases of type 2 diabetes is higher in subjects with BMI 31–34 kg/m² than in those with grade 3 obesity; this difference may also be due to the different ages of patients in the studies (8–12,26,27). Second, in some studies (8–12), risk reduction was 31–58% (average 45%), meaning that ∼55% treated patients became diabetic. In other studies (26,27) and this study, risk reduction was 80–100% (average 90%), meaning that <20% of treated patients became diabetic. From this viewpoint, surgery is preferable in morbid obesity, as it is effective for at least 10 years (LAGB) or indefinitely (gastric bypass). The duration of type 2 diabetes is a risk factor for complications as is the duration of obesity (e.g., left ventricular hypertrophy) (4,36). It remains to be shown if these complications are prevented by weight loss obtained through LAGB, although evidence exists to this effect with regard to the reduction of total and cardiovascular mortality for GBP (37,38).

This study had some limitations. First, compared with intervention studies based on dietary restrictions, drugs, and/or lifestyle modifications, our study was relatively small; however, it should be recalled that morbid obesity represents ∼5% of all cases of obesity (39). Second, even though the rate of dropouts was low in this study, obesity studies are usually affected by high rates of dropouts, whether they are diet (40) or surgery (23) studies. Third, this was an observational study, not a randomized clinical trial, but we have no reason to suspect that the large differences between the groups were the result of confounding factors.

In conclusion, this study showed that in grade 3 obesity, weight loss obtained through LAGB is able to prevent type 2 diabetes and hypertension for at least 4 years.

This research was supported by Grant FIRST 2002 from the Università degli Studi di Milano, the Ministero della Salute (grant RF-199/02), and the Ministero dell’Università e della Ricerca Scientifica e Tecnologica 2002 (grant 2002064582-003) to A.E.P. and from the Ministero della Salute (grant RF-02/224) to F.F.

We thank Tara Zoll Folli for her assistance in editing the manuscript.