The Effects of Obesity on Oesophageal Function, Acid Exposure and the Symptoms of Gastro-Oesophageal Reflux Disease

R. Anggiansah; R. Sweis; A. Anggiansah; T. Wong; D. Cooper; M. Fox

Disclosures

Aliment Pharmacol Ther. 2013;37(5):555-563. 

In This Article

Results

Demographic, anthropometric and physiological measurements were collected from 582 patients [female 56%: male 44%, median age 48 years (range 14–89)] referred for investigation of typical reflux symptoms (Table 1). The predominant symptoms of all patients were heartburn, acid regurgitation and retrosternal chest pain. There was a tendency for more women to have BMI > 30 kg/m2 (23% women vs. 16% men; P = 0.056); however, a greater proportion of men had abdominal obesity (WC ≥ 99 cm; 28% women vs. 41% men; P = 0.001). Figure S1 demonstrates the rise in BMI and WC with age in both women and men (P < 0.001). The median oesophageal acid exposure time was 4.7% time pH < 4 (IQR: 1.5–9.3%). Just over half (52.5%) had pathological acid exposure time pH <4 for >4% of the 24-h recording period. Completed questionnaires were available for 406/582 (70%). There were no differences in demographical, clinical or baseline oesophageal acid exposure characteristics between patients with complete and incomplete questionnaire data (data not shown).

Association of Obesity With GERD

Oesophageal acid reflux (log percentage time pH <4 during the study time period) increased with BMI (R = 0.236, P < 0.001) and WC (R = 0.284, P < 0.001). These effects were evident also for the number of reflux events (Figure 1) and were present in both the erect and the supine positions (all P < 0.001, data not shown). The findings persisted although to a reduced extent (BMI: R = 0.211, P < 0.001; WC: R = 0.189, P < 0.001) when adjusted for age and gender (Table S1). Separate multivariate analysis revealed additional, independent effects of demographic variables on oesophageal acid exposure (see below).

Figure 1.

Association of BMI (a) and WC (b) with oesophageal acid exposure and the number of reflux episodes. This unadjusted data demonstrate a positive, linear association between obesity and the severity of acid reflux assessed by ambulatory pH studies (see also Table S1).

Interaction of Obesity, Reflux and Manometric Assessment of Oesophageal Function

The severity of reflux disease as assessed by oesophageal acid exposure was significantly associated with a reduction in LOS pressure relative to intra-gastric pressure (R = −0.343; P < 0.001), abdominal LOS length (R = −0.343; P < 0.001) and contractile amplitude in the distal oesophagus (R = 0.207; P < 0.001). The findings persisted when adjusted for age and gender (Table S1).

Analysing the association of obesity with oesophageal function, there was a negative association of BMI with LOS pressure relative to intra-gastric pressure (R = −0.136; P = 0.001) but no significant link with abdominal LOS length (R = −0.095; P = 0.055). In contrast, there was a relatively strong, negative association of WC with both LOS pressure (R = −0.221; P < 0.001) and abdominal LOS length (R = −0.209; P < 0.001). Neither measure of obesity was associated with contractile amplitude. The findings persisted when adjusted for age and gender (Table S1). Unadjusted data illustrating the association of obesity on oesophageal function are presented in Figure 2.

Figure 2.

Association of obesity with lower oesophageal sphincter (LOS) pressure (a), and abdominal LOS length (b). This unadjusted data demonstrate a negative, linear association between obesity and LOS pressure and abdominal LOS length (i.e. the integrity of the oesophago-gastric reflux barrier). No association between obesity and peristaltic function was present (c) (see also Tables S2 and S3).

These findings are consistent with the 'mechanical hypothesis' linking obesity, especially abdominal obesity (WC), with oesophageal acid exposure by disruption of OGJ structure and function. We investigated further whether the association between obesity (BMI, WC) and acid exposure would be weakened or lost by correcting for manometric parameters affecting oesophageal acid exposure. This multivariate analysis revealed that the strength of association between acid exposure and obesity decreased but remained significant (P < 0.001 for both BMI and WC) after introduction of LOS pressure, intra-abdominal LOS length and distal contractile amplitude (Table S2). These findings persisted after adjustment for age and gender. In addition, we tested whether the effect of obesity on oesophageal function had an effect on oesophageal acid exposure by analysing whether the strength of association between obesity and acid exposure decreased when LOS pressure or intra-abdominal LOS length were entered into the regression model. This showed that obesity and oesophageal function had independent effects on oesophageal acid exposure. Specifically, the inclusion of the interaction terms relating obesity and oesophageal function into the regression had no significant effect on the model for BMI (change in adjusted R 2 = 0.008, P = 0.314) or WC (change in adjusted R 2 = 0.004, P = 0.616). Finally, we checked whether the 'goodness of fit' of the regression model (adjusted R 2) for BMI or WC was changed by adding the other anthropometric measure of obesity into the regression model; however, this did not show significant effects (data not shown).

Concerning the association of obesity and oesophageal acid exposure, introduction of age and gender into the regression models consistently increased the 'goodness of fit'. For the model including WC and manometric variables, demographic factors increased the adjusted R 2 by 0.029 (P = 0.001) and 0.023 (P = 0.004) respectively. The multivariable regression is presented in Table S3. The results show independent associations with anthropometric (BMI, WC), demographic (age, gender) and physiology (LOS pressure, LOS abdominal length) variables (all P < 0.05).

Symptom Severity

There was an increase in reflux symptom severity with oesophageal acid exposure (R = 0.300, P < 0.001). Obese patients complained of more severe reflux symptoms than thin patients; however, this effect was not seen after adjustment for age, gender and acid exposure by multivariable analysis (BMI P = 0.605, WC P = 0.320; Table S4). Thus, the severity of reflux symptoms at any given level of oesophageal acid exposure was independent of BMI or WC.

Further analysis revealed interactions between age, gender, acid exposure and symptom severity. After adjustment for acid exposure, younger patients and women complained of somewhat greater symptom severity than older patients and men at any given level of oesophageal acid exposure. For females, the association between acid exposure and symptoms was significantly affected by age [unstandardised coefficient −0.047 (CI: −0.82 to −0.12); P = 0.008]. For males, the association was independent of age [unstandardised coefficient −0.024 (CI: −0.63 to 0.15); P = 0.226]. The relationship between age, obesity and symptom severity is graphically presented in Figure 3.

Figure 3.

Effects of abdominal obesity and age on oesophageal acid exposure and symptom severity. This unadjusted data demonstrate that the increase in oesophageal acid exposure and reflux symptoms observed in obese patients was present across the age range of patients referred for investigation. The data are divided into tertile groups as defined in methods: lower tertile [WC <89 cm (group 1)], mid-tertile [WC 89–99 cm (group 2)] and upper tertile [WC>99 cm (group 3)]. Note that, despite the increase in oesophageal acid exposure with age, there was no concomitant increase in reflux symptoms in older patients (see also Table S4).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....