Replication and extension of dietary adherence as a predictor of suboptimal weight-loss outcomes in postbariatric patients.

BACKGROUND

Sarwer et al. found that poor dietary adherence at 6 months postoperatively predicted lower weight loss.

OBJECTIVES

To replicate and extend these findings.

SETTING

University bariatric clinic.

METHODS

Fifty-four adults (72% female; age 51.1 ± 11.3 yr; mean body mass index [BMI] = 43.8 ± 7.4 kg/m; 53.7% = Roux-en-Y gastric bypass, 42.6% = laparoscopic sleeve gastrectomy, and 3.7% = gastric banding) were identified as low or high dietary adherers following the method of Sarwer et al. Patients self-reported dietary adherence with a 9-point Likert scale. Splitting the group at the median, low adherers scored <7 and high dietary adherers ≥7. BMI, percentage excess weight loss (%EWL), and percentage total weight loss (%TWL) were prospectively assessed at 12, 24, and 36 months. Two-tailed independent t tests and Cohen's d effect sizes were used to compare between-group outcomes.

RESULTS

BMI did not differ between low (n = 24) and high (n = 30) dietary adherers at 6 months after surgery. At 12 months, the BMI of low (n = 17) adheres was significantly higher (34.1 ± 4.61 versus 30.3 ± 3.90 kg/m, P = .006, d = 0.90) than that of high (n = 25) adherers, with significantly less %EWL (49.0 ± 24% versus 70.7 ± 21.5%; P = .004; d = 0.95) and %TWL (20.7 ± 11.5% versus 28.9 ± 10.5, P = .02, d = 0.74). At 24 months, BMI remained significantly higher for low (n = 12) versus high (n = 10) adherers (33.7 ± 4.77 versus 29.7 ± 3.82 kg/m, P = .045, d = 0.92), but %EWL and %TWL were not significantly different, despite large effect sizes. At 36 months, moderate effects supported continued higher BMIs and lower %EWL and %TWL for low (n = 5) versus high (n = 8) adherers. Attrition from follow-up was 22.2% (12 mo), 59.3% (24 mo), and 75.9% (36 mo). Post hoc analyses revealed no impact of baseline characteristics on low follow-up rates except younger age (at 1 yr).

CONCLUSIONS

Findings that 6-month postoperative dietary adherence predicts 12-month BMI, %EWL, and %TWL were replicated. Medium to large effects suggest findings extend to 24 and 36 months, with low follow-up rates likely affecting statistical significance.