Associations of Bariatric Interventions With Micronutrient and Endocrine Disturbances

Key Points Question Are bariatric interventions associated with serum micronutrients and metabolic hormones changes in the years after intervention, and do trajectories differ between laparoscopic sleeve gastrectomy vs gastric bypass procedures? Findings In this comparative effectiveness study of 499 patients who underwent sleeve gastrectomy and 189 patients who underwent one-anastomosis or Roux-en-Y gastric bypass, parameters associated with calcium metabolism did not differ between the groups. The extent of hemoglobin suppression was greater among patients who underwent gastric bypass compared with those who underwent laparoscopic sleeve gastrectomy, but the difference was not explained by differences in iron stores or vitamin B levels. Meaning The findings of this study suggest that micronutrient deficiencies after bariatric interventions are differentially associated with types of bariatric procedures.

© 2020 Syn NL et al. JAMA Network Open.

Calcium (corrected)
590 patients contributed observations to serum calcium trajectories (eFigure 4). In the overall cohort, serum calcium levels remained relatively constant over time. However, in subgroup analysis, a significant increase in serum calcium (as compared with pre-bariatric levels) was delineated at 3 months post-procedure in the restrictive group (+0.07 mmol/L; 95% CI: 0.004 to 0.039; P = 0.0185) (eFigure 4). No difference between the restrictive and bypass groups was observed at any time during the 5-year follow-up.
Phosphate 582 patients contributed observations to the analysis of serum phosphate levels over a 5-year period (eFigure 4). In the overall cohort, serum phosphate levels were statistically higher at post-bariatric 6 and 12 months (eFigure 4). No difference between the restrictive and bypass groups was observed at any point in time during the 5-year follow-up.

Changes in Iron Metabolism
Serum Iron 565 patients contributed observations to the analysis of serum iron trajectories (eFigure 7). In the overall cohort, iron levels decreased only in the 1 st post-bariatric month, but subsequently increased above baseline levels, especially at 6 through 36 months (eFigure 7). In subgroup analyses, iron levels were higher at 6 through 24 months as compared with baseline in the restrictive group, but higher at 12 through 36 months as compared with baseline in the bypass group (eFigure 7). Iron levels peaked at 12 months and 36 months respectively in the restrictive and bypass groups, which represented an increase of +3.76 mmol/L (95% CI: 2.57 to 4.96; P<0.0001) and +6.86 mmol/L (95% CI: 1.23 to 14.5; P=0.0170) as compared with their pre-bariatric values (eFigure 7).
There was no difference in iron levels between the restrictive and bypass groups over the follow-up, except for the 3 rd month where patients in the restrictive group recorded slightly higher iron levels than those who underwent bypass procedures (+1.60 mmol/L; 95% CI: 0.26 to 2.90; P=0.0190) (eFigure 7).
Amongst women, those undergoing restrictive or bypass procedures had comparable serum iron levels for the first 24 months, but diverged thereafter (eFigure 8). Amongst male bariatric patients, serum iron levels were statistically higher among those undergoing restrictive procedures than bypass procedures at the 1 st month post-operatively, but became comparable in the subsequent months (eFigure 8).

Total Iron-Binding Capacity
Total iron-binding capacity (TIBC) was measured at multiple time-points for 560 patients over a period of 5 years (eFigure 7). In the overall cohort, TIBC was substantially lower at post-bariatric months 1, 6 through 24, and 36 and 48, as compared with baseline  Intriguingly and unexpectedly, within the female subgroup of patients, post-bariatric ferritin levels were statistically higher amongst those who underwent bypass compared to restrictive procedures (average fixed-effects: +31.1 ng/mL, 95% CI: 11.6 to 50.7, P=0.0018) over the course of follow-up (eFigure 6). Amongst the male subgroup, those undergoing restrictive procedures recorded higher ferritin levels between 6 through 18 months compared to male counterparts undergoing bypass procedures (eFigure 6).

Hemoglobin
Haemoglobin was longitudinally assessed over 5 years in 606 patients, and levels in the overall cohort were statistically lower at all post-bariatric time points as compared with prebariatric levels, except for the 3 rd month (eFigure 5).
This observation was largely recapitulated in subgroup analyses of patients who underwent restrictive and bypass procedures; the only exceptions were the 1 st post-operative month in the bypass group and the 9 th post-procedure month in the restrictive group wherein hemoglobin levels did not differ significantly from baseline levels (eFigure 5). During the initial 3 months, the bypass group recorded a steep decline in hemoglobin levels of -0.7 g/dL (95% CI: -1.0 to -0.4; P<0.0001) as compared with pre-surgery levels (eFigure 5). On the other hand, the restrictive group recorded a more tapered decrease in hemoglobin levels up till 18 months post-procedurally (-0.5 g/dL; 95% CI: -0.8 to -0.3; P=0.0001) (eFigure 5). In both the restrictive and bypass groups, trough hemoglobin levels were recorded at month 48, with a decrease of -0.8 g/dL (95% CI: -1.3 to -0.4; P=0.0004) and -0.9 g/dL (95% CI: -1.3 to -0.6; P<0.0001) respectively as compared with pre-bariatric levels.
Over the 5-year follow-up, hemoglobin levels were on average +0.6 g/dL (95% CI: 0.1 to 1.0; P=0.0003) higher in the restrictive group as compared with the bypass group. The difference between the restrictive and bypass groups were statistically significant at postbariatric months 1, 3 through 18, and 36 (eFigure 5).
Amongst the female subgroup of patients, there was evidence of a strongly monotonic, decreasing linear trend in haemoglobin levels over the course of follow-up (P=0.0005).
Amongst female patients, we observed that patients undergoing restrictive procedures had statistically higher post-bariatric haemoglobin levels between 3 through 18 months (eFigure 6). Amongst the male subgroup of patients, patients undergoing restrictive procedures had statistically higher post-bariatric haemoglobin levels at months 3-9, 18, 30, 36, and 60 months. A monotonically-decreasing trend in haemoglobin levels was not detected (P=0.7841), however, post-bariatric levels were statistically lower at all time points compared to baseline (eFigure 6).

Vitamin B Metabolism
Vitamin B12 (Cobalamin) 591 patients contributed observations to the time-course analysis of serum Vitamin B12 (cobalamin) levels (eFigure 9). In the overall cohort, post-bariatric Vitamin B12 levels were statistically higher at 12 through 36 months, and at 60 months as compared with baseline (eFigure 9). On average, there was no difference in serum vitamin B12 between the restrictive and bypass groups at any time point.

Vitamin B9 (Folate)
580 patients contributed observations to the 5-year longitudinal analysis of serum vitamin B9 (folate) levels (eFigure 9). Serum vitamin B9 levels were significantly higher at all ensuing time points after bariatric procedures as compared to baseline levels in the overall cohort (all P < 0.0001) (eFigure 9), an observation which was also recapitulated in subgroup analyses of bypass and restrictive procedures (eFigure 9). Over the 5-year follow up, vitamin B9 levels were on average higher in the bypass groups as compared to the restrictive groups (average fixed-effects = +7.1; 95% CI: 1.8 to 12.4; P=0.0001). The greatest disparity was observed at 30 months, wherein vitamin B9 levels was +13.4 nmol/L (95% CI: 7.0 to 19.9; P<0.0001) higher in the bypass group vs the restrictive group (eFigure 9).

Zinc and Magnesium
199 and 161 patients contributed observations to the analyses of zinc and magnesium respectively, and trajectories are depicted in eFigure 10. In the overall cohort, magnesium concentrations remained relatively stable and only dipped significantly at 1, 30, and 36 months compared to baseline, while post-operative zinc levels did not differ significantly compared to baseline (eFigure 10). Serum zinc concentrations generally did not differ between the restrictive and bypass groups; however, magnesium levels were, on average, higher in the restrictive group than in the bypass group during follow-up (fixed-effects = +0.078; 95% CI: 0.035 to 0.121; P=0.0004).

Global nutrition status: Serum Albumin and %Total Weight Loss
We assessed serum albumin as a marker for global nutrition status, as macronutrient deficiency could potentially mediate or exacerbate micronutrient deficiencies. 595 patients contributed observations to the analyses of serum albumin (eFigure 11). We did not observe any clinically significant difference in albumin levels between the restrictive and bypass groups which could otherwise confound subsequent comparisons of micronutrient levels between surgical procedures (average fixed-effects = 0.6 g/L, 95% CI: -0.07 to 1.27; P=0.0875).
Although there appeared to be a monotonically increasing trend in albumin levels over the course of follow-up, albumin levels were not statistically higher after bariatric surgery compared to baseline values (eFigure 11). Likewise, we also compared percentage change in total weight loss (%TWL) between LSG versus OAGB/LRYGB and found no difference in %TWL over the course of follow-up (MD = +0.00%, 95% CI: -3.48% to 3.48%; P=1.0000) (eFigure 12).