Effect of Parenteral Nutrition on Nutrition Status After Bariatric Surgery for Morbid Obesity 
From * Endocrinology and Nutrition, Correspondence: M. A. Martínez Olmos, Endocrinología y Nutrición, Hospital do Meixoeiro, Cno Meixoeiro S/N, 36.200 Vigo (Pontevedra), Spain. Electronic mail may be sent to miguel.angel.martinez.olmos{at}sergas.es. Background: To evaluate the influence of nutrition support (parenteral nutrition [PN] vs no parenteral nutrition [nPN]) on nutrition outcome, complications, and hospital stay after bariatric surgery (BS). Methods: Sixty-seven consecutive BS patients (17 gastric bypass and 50 biliopancreatic diversion). The first 38 received PN and the next 29 did not (nPN) during the fasting postoperative (PO) period. In both groups, after fasting, a progressive oral diet was introduced. Data related to nutrition status, perioperative complications, and postsurgical hospital stay were compared. Results: Sixty-seven patients (58 women), mean age 39.4 ± 11.0 years, body mass index (BMI) 50.7 ± 6.1 kg/m2, were included. Thirty-eight patients received PN during 8.7 ± 2.6 days. Ingestion was initiated at a median 8 PO days in PN vs 6.5 PO days in nPN (p < .04). No significant differences between groups were found in age, final fasting serum albumin (SA), and 30 days postoperative SA, with a similar weight loss. Nonsignificant differences were found in non–catheter-related infectious complications, being mainly urinary tract infections. Catheter-related infections were present in 21.1% in the PN group and 13.8% in the nPN (p= .33). Median hospital stay after surgery was 14 ± 10 days in PN and 12 ± 10 days in nPN (p = .003). Conclusions: (1) Nutrition status after BS PO and 30 days postsurgery was no different between PN and nPN. (2) Postsurgery hospital stay was significantly decreased in the nPN group, without a greater incidence of complications. (3) According to nutrition outcome, PN seems unnecessary at the perioperative period in BS unless there are other postsurgical complications. Bariatric surgery constitutes an alternative treatment used to fight against comorbidity associated with severe obesity. Among bariatric surgery techniques, at medium- and long-term, the most effective are those that act over both the physical capacity of the digestive tract and its absorptive capacity.1 Mixed techniques of bariatric surgery generate a metabolic stress associated with surgery itself and manipulation of the digestive tract that could impact morbidly obese patients with comorbidities2 such as insulin resistance with or without hyperglycemia, respiratory insufficiency, difficulties in venous access, etc. In obese patients (as in nonobese3), metabolic stress response is mediated by contrainsular hormones (glucocorticoids, glucagon, catecholamines, GH) and by cytokines, but the exact effect of these mediators in obesity is unknown. Obese patients with trauma show a greater protein turnover and less lipid mobilization than nonobese patients under similar stress.4,5 Traumatized obese patients have a relative blockade in mobilization and use of lipids and a significantly greater protein and carbohydrate oxidization than nonobese, with a greater nitrogen loss (absolute and related to lean body mass); also, protein synthesis is significantly less efficient in obese than in nonobese individuals after severe trauma. Relative blockade in lipid mobilization and oxidization generates a preferential use of calories from carbohydrates, which accelerates destruction of body proteins in obese with respect to nonobese patients.6 In stress, obese patients do not benefit from lipid reserves but are dependent on efficient use of endogenous glucose synthesized from products of protein degradation. The increase in carbohydrate use generates an increase in respiratory quotient, which, together with muscular destruction, can generate respiratory problems; in fact, obese patients have a greater risk of pneumonia and necessitate a prolonged time of ventilatory support than nonobese.7 For these reasons, nutrition support in patients after bariatric surgery during the postoperative period (especially during postoperative fasting) is controversial.8,9 This work reviews our experience in the use of 2 different nutrition support strategies after this surgical stress: parenteral nutrition (PN) vs no nutrition support (nPN), in order to know which of both is more advantageous.
We studied all successive patients with morbid obesity having bariatric surgery in a 421-bed teaching hospital from 1998 to 2002. Patients were maintained fasting in the postoperative period until radiologic study with Gastrografin (Schering España, Madrid, Spain) demonstrated absence of digestive fistulas.
The biliopancreatic diversion with duodenal switch (BPD-DS) consists of a sleeve gastrectomy, a duodenal switch, a 250-cm alimentary channel, and a 75-cm common channel. This operation divides the duodenum in the distal bulb and the ileum 250 cm proximal to the ileocecal valve with anastomosis of the proximal duodenal segment to the distal ileal segment. The proximal ileal segment, which carries the biliary and pancreatic secretions, is anastomosed end to side for a enteroenterostomy 75 cm proximal to the ileocecal valve. The Roux-en-Y gastric bypass (RYGBP) was constructed by a tubular small gastric pouch on the gastric lesser curvature, the jejunum was transected 40–100 cm distal to the ligament of Treitz, the enteroenterostomy was performed between 100 and 150 cm from the transection line, and a gastrojejunostomy was constructed without tension (Figure 1). The election of which technique used in each patient was made dependent on patient characteristics: age, percentage ideal body weight, eating habits, etc, as previously described.10 The first 38 patients (group 1) received PN by central vein with the following composition: 8945.4 ± 640.2 kJ/d, 64.7 kJ/kg/d, 0.12 g of nitrogen/kg/d, 532 ± 75.3 nonprotein kJ/g of nitrogen/d. If we express its composition related to corrected weight (using the formula corrected weight = ideal body weight + [actual body weight – ideal body weight] x 0.25),11 values are 116.6 ± 14.3 kJ/corrected kg/d, 0.21 g of nitrogen/corrected kg/d. The remaining 29 patients (group 2) received fluid therapy with 5% dextrose and 0.9% saline during the postoperative period. All patients, after recovery of intestinal transit and checking for the absence of anastomotic fistulae, received an oral diet progressive in volume and nutrition intake, with protein supplementation, until a regular diet is started in 4 weeks (Table I).
The more prevalent diagnostic-related groups in our patients were 288 ("obesity surgery," 39 cases) and 565 ("surgery for endocrine, nutrition or metabolic diseases, except amputations, with major complications," 18 cases). We studied perioperative nutrition parameters (especially serum proteins), appearance of complications (suture dehiscence, wound infection, catheter-related infections), and postsurgery hospital stay. Surgical site infections were defined according to Centers for Disease Control and Prevention (CDC) definitions, including superficial incisional, deep incisional, and organ/space infections.12 Catheter-related infections were defined according to CDC definitions, including colonization and bloodstream infections.13 Urinary-tract infections were defined according to previous definitions.14
Statistical Analysis
Basal characteristics and comorbidities of patients are expressed in Table II. Surgical techniques used were (a) RYGBP: 17 patients; (b) BPD-DS: 50 patients. We did not observe differences between groups with respect to age, sex, body mass index, percentage weight excess, waist circumference, waist-to-hip ratio, presence of impaired glucose tolerance, presence of respiratory alterations, or techniques used.
In group 1, PN was used during 8.7 ± 2.6 days. In group 2, ingestion was initiated at 7.5 ± 5.2 days. It was necessary to use some type of peripheral nutrition during 6.62 ± 2.19 days in 10 patients because of delay in initiating oral ingestion (glucose or amino acid peripheral solutions in a minimal amount [12.31 ± 4.72 kJ/corrected kg/d, 0.15 ± 0.05 g of nitrogen/corrected kg/d]). In subgroup analysis, after excluding these 10 cases, the results were similar, as described below. Median postoperative oral ingestion started at 8 days in group 1 and 6.5 days in group 2 (p < .04). In Figure 2, we can see values of some serum protein in the perioperative period in both groups. Mean weight reduction 1 month after surgery with respect to preoperative weight was 8.71 ± 3.03% in group 1 and 11.20 ± 4.76% in group 2 (p < .02).
With respect to postsurgery complications, in group 1 we found 7.9% wound infection, 28.9% other infections, the majority (23.7%) being urinary tract infections. In group 2 we found 17.2% wound infection and 27.6% other infections, including 24.1% urinary tract infections (p = nonsignificant with respect to group 1). Catheter-related infections occurred more frequently in group 1 (21.1%) with respect to group 2 (13.8%), but differences do not reach statistical significance (p = .33). Median postsurgical hospital stay was 14 ± 10.1 days in group 1 and 12 ± 10.1 days in group 2 (p < .003). If we look at the data dividing patients in PN vs no PN, no differences were found between RYGBP and BPD-DS groups in main variables (Table III and Table IV). There were no significant differences in protein markers, postoperative ingestion, or postsurgical hospital stay. The only significant differences observed between groups occurred in age, corrected weight, and the incidence of any infection in nPN patients, mainly due to urinary tract infections.
The more accepted bariatric surgery techniques are mixed techniques, restrictive and partially malabsorptive,1 because they are more effective in inducing medium- and long-term weight loss, but they are more complex technically and have more nutrition risks. In the postoperative state, patients remain fasting for a variable period (generally 5–7 days), during which need for nutrition support therapy is controversial.8 Our study demonstrates that postoperative courses in morbidly obese patients after bariatric surgery are not different whether they receive perioperative PN or maintenance fluid therapy, and a similar nutrition recovery after 30 days postsurgery and a significant reduction in postsurgical hospital stay in fluid therapy group. PN does not add any benefit in bariatric surgery. Obese patients have insulin resistance with a compensatory increase in insulin secretion. In these patients, hyperinsulinemia is the consequence of a diminished sensitivity and hepatic uptake of the hormone and in visceral obesity to the increase in free fatty acids flux. In muscle, insulin resistance is related to a decrease in GLUT4 mRNA levels and a less autophosphorylation of insulin receptor, at least in part due to TNF. Also, postreceptor defects have been described.15 Hyperinsulinemia of critical obese patients (greater than in nonobese) could be responsible for less endogenous lipids turnover in this situation3 because of less lipoprotein lipase activity in adipose tissue due to associated insulin resistance. So in obese individuals under stress, metabolic changes predisposing to adverse consequences are increased,16,17 and nutrition support in these patients must take into account characteristics of hormonal changes and metabolic response in these subjects and its differences with respect to patients not overweight. In this sense, it is of note that our patients (with severe degrees of central obesity) had an elevated frequency of impaired glucose metabolism and associated respiratory alterations, which could have important metabolic and nutrition implications, influencing protein depletion postoperatively.18,19
In our clinical practice, because of this special tendency of central obese patients to deplete protein, PN was initiated after bariatric surgery with a good result, as previously described.20 Afterward, due to central catheter-related infections, we started to use this support only in patients with complications postoperatively, as previously described.8,18 In some patients, peripheral nutrition was used because patients had delays in initiating oral ingestion, but the amount was limited, and this did not alter the analysis between PN vs no PN. After subgroup analysis, statistical significance did not change (data not shown). If we analyzed data dividing patients in PN vs no PN, nutrition status and main outcome variables were similar, except for an increase of infectious episodes in RYGBP patients without PN, mainly urinary tract infections successfully resolved with adequate antibiotic therapy. After digestive tolerance was assured, we started a liquid diet in both groups of patients and maintained it during the first week, and a semiliquid diet until the end of the first month after surgery. The objectives of this dietetic progression are to facilitate digestive tolerance and also to change dietetic habits to adapt to the new "physiology." That is because this diet includes small volumes in each ingestion and is divided in 6-times-a-day, progressively increasing quantity, thickness, and consistency, low in residues and without irritating foods and citric acid. To preclude malnutrition, we added protein supplements to ensure an adequate nitrogen balance because the small volumes of ingestion cannot achieve sufficient protein intake. Also, it is mandatory to add multivitamins and mineral supplements.8 With respect to biochemical parameters, we observed nonsignificant differences in serum albumin and transferrin in both groups after 30 days, indicating that the non-PN group had a similar postoperative nutrition course, in spite of a similar or greater weight loss. We must remark that in the immediate postoperative period, PN patients' values of albumin and prealbumin are significantly greater. Nevertheless, diminished serum albumin was not important in both groups and could be influenced by fluid balance. These differences do not reflect an increase in complications.21 Infectious complications were similar in both groups, predominantly urinary tract infections. Central venous catheter–related bacteremia showed a tendency to decrease in the non-PN group, but that did not reach statistical significance. Many patients do not need central venous catheters if they do not have PN. One of the more important aspects of this study is the significant reduction in postsurgery hospital stay in patients without PN, which we attribute essentially to earlier oral intake in this group. Advantages in reduction of stay are clear, not only relative to less hospital costs but also because of diminished complications associated with hospitalization (nosocomial infections and others). It is also of note that during the study there were no changes in surgical techniques or the surgery team responsible for the patients. Complications in surgical wounds were similar in both groups. In conclusion, we can say from this study of morbidly obese patients after mixed techniques of bariatric surgery that (1) nutrition status after bariatric surgery in the perioperative period and 30 days postsurgery was no different between patients receiving PN or nPN, (2) surgical hospital stay was significantly decreased in the nPN group without a greater incidence of complications, (3) according to nutrition outcome, PN seems unnecessary in the perioperative period in bariatric surgery unless there are other postsurgical complications. Received for publication October 7, 2004. Accepted for publication August 4, 2005.
Journal of Parenteral and Enteral Nutrition, Vol. 29, No. 6,
445-450 (2005)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Top
PATIENTS AND METHODS
RESULTS
2, with Yates' correction for small groups. Search
of association between diverse parameters was made by Spearman lineal
correlation coefficient. For comparison of nonnormal data, we used the
Mann-Whitney test. We considered a p value of < .05 to be a
statistically significant difference. All calculations were made with SPSS v
8.0 for Windows. 