Obesity has become a dominant public health epidemic with increasing prevalence.1-3 Obesity is associated with cardiovascular disease, coronary artery disease, thromboembolic disease, cerebrovascular disease, some forms of cancer, diabetes mellitus, and obstructive sleep apnea.4-9 Also, population trends toward increased body weight may correlate with an increased incidence of low back pain and degenerative joint disease.3,10 In the obese patient, increased physiological loads on the spine may lead to pain.11,12 The spine sustains increased stress, which may lead to advanced degeneration.11 However, a clear relationship between obesity and low back pain has not been identified.7
Obesity is becoming a common entity in patients undergoing elective thoracic and lumbar spinal surgery. Obese patients have increased risk of complications after spine fusions, including surgical site infections.1,7,13-17 However, increased risk of complications does not represent a contraindication to surgical intervention. Perioperative in-hospital mortality for morbidly obese patients in 1 study was less than 1%.18
Existing studies have focused on the perioperative complications surrounding obesity and spinal surgery.11,18-20 Glassman et al21 used postoperative outcome tools for single- and 2-level fusions. There is a void in the literature regarding long-term assessment of the impact of increased body mass index (BMI) for deformity spinal surgery. This is the first study to look at BMI as an independent variable for spinal fusions of more than 5 levels with 1 and 2 years of follow-up. This study examined BMI and its association with adverse outcomes (increased hospital stay and complications at 1 year).
PATIENTS AND METHODS
A detailed, retrospective review of the inpatient and outpatient medical records from 2007 to 2010 was completed after institutional review board approval. All patients undergoing fusion of 5 or more levels were examined. Patients with a diagnosis of degenerative kyphoscoliosis, lumbar flat-back deformity, and sagittal malalignment were included. Consecutive patients older than 18 years of age who had thoracic, thoracolumbar, or lumbar fusions were reviewed. These cases involved open fusions consisting of interbody and/or posterolateral arthrodesis. Cases of trauma, infection, neoplasm, and nonelective procedures were excluded from our dataset. There was a total of 193 surgical cases; however, 4 were excluded due to unavailable BMI data before to the surgical procedures. Thus, 189 surgical cases were analyzed.
Demographic data were collected for all patients. The American Society of Anesthesiologists class and Charlson Comorbidity Index were used to assess the role of severity of health problems at the time of admission. Outcome variables assessed included perioperative factors and complications before discharge, both major and neurological. Variables thought to affect BMIs were assessed. Major medical complications included myocardial infarction, infection, deep venous thrombosis, pulmonary embolism, pneumonia, stroke, reintubation, optical deficit, and thrombophlebitis. A myocardial infarction was defined as a troponin elevation higher 0.50. An infection was defined as a positive culture (eg, blood, urine, cerebrospinal fluid). Deep venous thrombosis was identified by a positive lower extremity Doppler study. A pulmonary embolus was present if findings on computed tomography angiography were denoted as positive. Pneumonia was present if 3 of the 4 criteria were present: fever higher than 101.5°, positive nonbronchoscopic bronchoalveolar lavage, consolidation on chest radiograph, or productive expectorant. A stroke was defined as restricted diffusion on a magnetic resonance imaging of the brain. Major neurological and surgical complications included motor deficit, bowel/bladder deficit, radiculopathy, sensory deficit, skin complications, wound dehiscence, and death. Motor deficits, bowel/bladder deficits, radiculopathy, and sensory deficits were accounted for by review of the inpatient and outpatient charts. This is also true for any wound-related issues. Complications, revision spinal surgery, and nonspinal surgery were documented at the wound check, early follow-up, 1-year follow-up, and 2-year follow-up period. The incidence of junctional kyphosis and pseudarthrosis at each follow-up interval was recorded. Rehospitalization rates were also calculated. Oswestry Disability Index (ODI) was reviewed at each follow-up visit. The ODI raw number was placed in an ordinal grouping, and the number of individuals in that category was reported. The mean ODI and standard deviation, in addition to the ordinal categories, were included. Each variable raw number was reported, and a percentage was calculated based on the entire cohort for consistency.
We included variables (eg, smoking history, valvular disease, cardiac medications, cardiac laboratory values) in our analyses to identify whether there is a relationship between them and our outcome measures. These particular variables were chosen, as they have an association with cardiac and vascular disease. Estimated blood loss was estimated from the anesthesia record; this value was cross-referenced based on total fluids suctioned minus the irrigation.
Bivariate analyses (t tests, analysis of variance, Kruskal-Wallis, χ2 test, Pearson correlation) were performed on all individual variables and outcomes. One-way analysis of variance examined the outcome variables based on BMI categories. Kruskal-Wallis examined nonparametric variables. Multivariate linear regression analysis was used to evaluate the impact of BMI on hospital stay and complications at 1 and 2 years. Factors potentially influencing the outcome desired were controlled for in the regression models. The models were developed using both forced entry and stepwise progression. The first block of variables were placed in the model regardless of significance (age, sex). The second block of variables were only placed in the model using a forward stepwise progression if they met the statistical requirements for entry (P < .05). A priori knowledge and bivariate analysis were ultimately used to drive model development, not random stepwise inclusion or exclusion criteria exclusively. Age, sex, BMI category, and preoperative ODI were controlled for in both models. The sum of preoperative heart medications and cerebrospinal fluid leak was controlled for in the hospital stay model; and pedicle subtraction osteotomy (PSO) performed was controlled for in complications at 1 and 2 years. All statistical analyses in this study were performed SPSS software version 20) (IBM SPSS Inc, Armonk, New York). A P value <.05 was considered statistically significant.
The National Institutes of Health guidelines categorize weight status using the BMI. BMI is reported in kilograms divided by height in meters squared (kg/m2).22 Underweight patients had a BMI of <19 kg/m2, ideal weight patients had a BMI of 19 to 24.9 kg/m2, and overweight patients had a BMI of 25 to 29.9 kg/m2. Patients with a BMI of 30 to 39 kg/m2 with no significant comorbidity were considered obese.22,23 Morbidly obese patients had a BMI of 35 to 39 kg/m2 with a significant comorbidity or a BMI of 40 kg/m2 or higher.22,23 The presence of a significant comorbidity included having any one of the following disorders: hypertension, diabetes, anticoagulation, asthma/bronchitis, hyperlipidemia, thyroid disease, psychiatric disorder, angina, alcohol consumption, shortness of breath, sleep apnea, and myocardial infarction.22,23
All patients were provided with structured outcome forms during the initial outpatient clinic visit and all subsequent visits. A database was compiled using inpatient and outpatient medical records. Patients were excluded if there was not at least 1-year follow-up. Mean and median follow-up were 2.1 and 2.0 years, respectively. If patients were lost to complete 2-year follow-up after the 1-year follow-up encounter, they were not included in the 2-year assessment. Subgroup analysis was performed of the population who did achieve 2-year follow-up. Missing data points are identified and reflected in the tables. If missing data points were encountered for a particular variable, the sum total of the potential events did not include the missing variable population.
A total of 189 surgeries on 112 patients (30 male/82 female), mean age of 59.5 years and a mean BMI of 29.8 kg/m2, were analyzed. There were 32 smokers. Preoperative heart medications include angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, β-blockers, calcium channel blockers, diuretics, and statins. The mean number of preoperative heart medications was 1.6 (SD = 1.4). There were 5 surgeries in patients who were taking an antiarrhythmic (ie, digoxin), and 7 surgeries in patients who were taking clonidine or minoxidil for hypertension. Fifty-one surgeries and 10 surgeries required aspirin and clopidogrel to be held, respectively. There were 16 surgeries for which the patients had to be taken off of their warfarin preoperatively and 10 surgeries in which cardiac coronary stents were involved (6 drug-eluting and 4 bare metal stent placement). The mean preoperative ODI was 49.7 (SD = 15.7). Preoperative data are shown in Table 1.
Intraoperative data are shown in Table 2. There were 124 staged surgeries (104 inpatient staged surgeries and 20 delayed staged surgeries). The mean number of surgery levels was 8.0 (SD = 4.4), and mean surgery length was 440 minutes (SD = 182). A total of 28 pedicle subtraction osteotomies and 6 vertebral column resections were performed. There were 159 surgeries in which the high-risk spine protocol was instituted.24 This dedicated, multidisciplinary protocol, implemented in January 2007 at our institution, has improved communication and outcomes by providing a management strategy for the preoperative, intraoperative, and postoperative care of spinal surgery patients. There was a total of 22 cerebrospinal fluid leaks (11.7%). The mean estimated blood loss was 2942 mL (SD = 2441). Perioperative data are also highlighted in Table 2. There were 8 intraoperative cardiac changes (4.2%), and 15 unplanned returns to the operating room (8.1%). Based on BMI, the returns to the operating room are stratified as follows: 0 for underweight, 3 for ideal weight, 4 for overweight, 2 for obese, and 6 for morbidly obese (P > .05). The return to the operating room for the ideal weight patients included foraminotomies for muscle weakness and inferior vena cava filter placement for pulmonary embolus. The return to operating room for overweight patients included instrumentation revisions, foraminotomy for pain, and pseudomeningocele repair. The return to the operating room for obese patients was for a retained drain and decompression for radiculopathy. The return to the operating room for the morbidly obese was for pseudomeningocele repair, incision and drainage for infection, wound incision and drainage for seroma, and completion of procedure after intraoperative cardiac changes. The mean number of days spent in the intensive care unit was 3.3, and the mean hospital stay was 13.3 days. Sixty-two of the patients had staged procedures. More than half of the patients required an acute inpatient rehabilitation stay.
Postoperative data are shown in Table 3. Major medical complications and neurological and surgical complications were recorded at the wound check, early follow-up, 1-year follow-up, and the 2-year follow-up. The rehospitalization rates were as follows: 12 (6.8%) since discharge (at wound check between discharge and wound check), 35 (18.6%) since wound check (at early follow-up between wound check and early follow-up), and 36 (19.8%) since early follow-up (at 1-year follow-up between early follow-up and 1-year follow-up). The rates for revision spinal surgery are as follows: 2 (1.1%) since hospital discharge (at the time of wound check), 18 (9.6%) since wound check (at the time of early follow-up, between wound check and early follow-up), 20 (11.0%) since early follow-up (at the time of 1-year follow-up, between early follow-up and 1-year follow-up), and 4 (3.8%) since 1-year follow-up (at the time of 2-year follow-up, between 1-year and 2-year follow-up). The mean ODIs for early follow-up, 1-year follow-up, and 2-year follow-up were 45.6 (SD = 15.6), 32.4 (SD = 19.1), and 26.6 (SD = 18.6), respectively.
Bivariate analysis revealed that the preoperative sum of patient’s heart medications between the 5 BMI categories was statistically significant (P = .001). It also showed that the Charlson Comorbidity Index score was also statistically significant between BMI categories (P = .014). The mean American Society of Anesthesiologists score was statistically significant between BMI categories (P = .034). Analysis of variance revealed the mean preoperative ODI based on BMI category was statistically significant (P = .001). Analysis of variance also showed the mean 1-year ODI and 2-year ODI based on BMI was P = .010 and P = .011, respectively. Underweight individuals compared with those who were not underweight had a larger highest postoperative troponin value, a mean of 0.82, compared with 0.14 ( P = .009).
The mean ODI values at the 1-year and 2-year follow-up time points stratified based on body habitus are significant on bivariate analysis. The difference between the mean preoperative ODI and the mean 1-year postoperative ODI for the weight are as follows for underweight, ideal, overweight, obese, and morbidly obese: 15.2 ± 16.8, 16.9 ± 20.1, 25.2 ± 18.2, 14.8 ± 22.2, 16.5 ± 14.7 (P = .349). The difference between the mean preoperative ODI and the mean 2-year postoperative ODI for weight are as follows: 22 ± 11.9, 19.1 ± 13.2, 27.5 ± 18.9, 22.5 ± 24.5, 17.2 ± 16.2 (P = .690). Bivariate analysis did not reveal a difference between BMI categories for ODI change between the preoperative visit and the 1- and 2-year follow-up dates, although improvement in disability was noted.
The incidence of junctional kyphosis was 18 (9.6%) at early follow-up, 17 (9.3%) at 1-year follow-up, and 14 (13.3%) at 2-year follow-up. The pseudarthrosis rate was 4 (2.1%) at early follow-up, 9 (4.9%) at 1-year follow-up, and 4 (3.8%) at 2-year follow-up. Bivariate analysis showed that morbidly obese patients had a longer hospital stay than ideal weight patients (P = .004), worse ODI scores at 1 and 2 years than ideal weight patients (P = .001 and P = .004, respectively), and more complications at 1 and 2 years than ideal weight patients (P = .011 and P < .001, respectively).
Table 4 shows a multivariate linear regression model revealing sex (P = .02, β = 2.10), cardiac medications (P = .02, β = .64), cerebrospinal fluid leak (P = .01, β = 3.07), and BMI category of ideal vs nonideal weight (P = .04, β = −1.79) influenced length of hospital stay. Table 4 features the category of ideal vs nonideal weight because on bivariate analysis, underweight individuals were similar to the obese and morbidly obese population in terms of surgery length, days in the intensive care unit, and troponin elevation. Each of these variables would have an impact on hospital stay. The multivariate linear regression analysis for all complications at 1 year (major and neurological) is reflected in Table 5, demonstrating that BMI >30 (P = .02, β = .30), preoperative ODI (P < .01, β = .01), and PSO performed (P = .02, β= −.34) influenced all complications at 1 year. BMI is categorized as BMI >30 in Table 5 due to the significant increases in complication rates in the obese and morbidly obese populations after surgery. Mean complications at 2 years for the morbidly obese were 3 times more than underweight, and 8 times more than ideal weight. Controlling for age, sex, and length of stay, morbidly obese patients had worse 2-year ODI (P = .010, β = 18).
Obesity is one of the most important determinants of health-related quality of life.25,26 In a review of the literature, the impact of BMI and obesity on spinal surgery remains unclear. Studies have been performed examining BMI and the development of herniated lumbar discs.27 A cross-sectional study out of Helsinki showed an increased BMI was associated with lumbar disc herniations requiring surgery.28 Cost-utilization projects provide insight to operative and perioperative costs, but without detailing associations beyond BMI. Kalanithi et al18 performed a retrospective, cross-sectional, cost-utilization study using the Healthcare Cost and Utilization Project's California State Inpatient Databases. Evaluating all spinal fusions from 2003 to 2007, they found the mean total hospital charges for posterior lumbar fusion were significantly higher for morbidly obese patients than normal weight individuals ($128 661 vs $108 569, P < .001); mean length of stay was longer for posterior lumbar fusion in morbidly obese patients than normal weight individuals (5.75 vs 4.68, P < .001). Additionally, morbid obesity was associated with higher in-hospital complications after posterior lumbar fusion compared with being of normal weight (odds ratio: 1.497, 95% confidence interval: 1.222-1.833]). However, this study provided no description with regard to surgery length, numbers of levels fused, osteotomy types, and so on.
In discussing fusions, there were conflicting results on the relationship between total complications and BMI. Yadla et al11 found no correlation between patient BMI and the incidence of perioperative minor or major complications in patients undergoing surgery for degenerative thoracolumbar procedures; however, their mean number of levels fused was 0.77 ± 1.99 in the thoracic spine and 1.17 ± 1.36 in the lumbar spine. For minimally invasive lumbar fusions, Rosen et al1 found no correlations between BMI and estimated blood loss, operative time, hospital length of stay, self-reported outcomes, or major or minor complications (P > .05) in patients undergoing minimally invasive lumbar fusions. Our study highlights that BMI has a direct relationship with length of hospital stay and complications at the year anniversary from surgery. Furthermore, age, sex, and preoperative disability index are controlled for, the significance remains.
There are multiple studies that have found a positive correlation of complications with obesity, but without long-term follow-up. Shamji et al4 performed logistic regression controlling for confounders and found that morbidly obese patients were 70% more likely to have a postoperative infection and have twice as many wound complications as the normal body habitus control patients exclusively with posterior approaches (P < .01). This study only looked at the perioperative hospital course and associated charges of the stay. It has the limitation of a national database investigation and without long-term follow-up. Patel et al7 examined the incidence of operative complications related to the presence of obesity where logistic regression analysis revealed that the risk of major complications increased with increasing BMI (P < .04). Unfortunately, the retrospective analysis did not have consistent long-term follow-up.
Many studies pertaining to BMI and complications do not feature more than 2 fusion levels. Vaidya et al23 found a higher incidence of postoperative complications in obese and morbidly obese patients. They had 2-year follow-up, although the majority of their patients had single- or 2-level fusions. Djurasovic et al3 examined the impact of obesity on the clinical results of patients undergoing lumbar spinal fusion. Obese patients had worse ODI scores before and after surgery compared with the nonobese (P < .05), but the degree of improvement with surgery was the same for both groups.3 Both groups showed significant improvement in back and leg pain after surgery, although obese patients did have a higher incidence of wound-related complications.3 This paper only focused on degenerative lumbar conditions and did not specify the number of levels. Peng et al29 examined the perioperative complication rates of anterior lumbar surgery in obese and nonobese patients. There were no significant differences in the overall complication rates or hospitalization length between obese and nonobese patients, although the mean number of operative levels was 1.8 and 1.6, respectively.29 This paper did not discuss long-term follow-up. Gepstein et al8 examined 298 consecutive patients 65 years of age and older. Only patients who had a decompressive laminectomy and/or discectomy were examined. They reported a higher percentage of very dissatisfied patients who were obese.8 Singh et al25 evaluated the results of lumbar fusions on perioperative complications in obese patients and return to work. Their study only looked at single- and 2-level posterior lumbar interbody fusions. The effect of BMI on minimally invasive spinal surgery was evaluated by Park et al30; however, they only looked at single- or 2-level procedures.
The obese patient presents a multitude of challenges during the preoperative, operative, and perioperative periods. The quality of preoperative imaging based on size presents a challenge to magnetic resonance imaging technicians and obtaining adequate range of motion of flexion-extension plain films for radiographers.7 Operating room difficulty in airway management and arterial and venous access increase the challenge for the anesthesiologist.7,19,25 Impaired pulmonary function, increased adiposity for anesthetic distribution, and limited mobility worsen the immediate postoperative care for the anesthesiologist.7,19
Upasani et al31 looked at the effect of obesity on adolescent idiopathic scoliosis and found that obesity did not influence perioperative morbidity with respect to implant failure, wound infection, neurological deficit, pseudarthrosis, or surgical revision. Obesity did not increase perioperative morbidity in this cohort, nor did it affect the ability to maintain or achieve coronal or sagittal realignment. The relationship of obesity to length of hospital stay and complications at 1 year in the adolescent idiopathic scoliosis patient was not looked at in our study.
The burden to the health care system by this epidemic is important for cost utilization. Treating patients and minimizing their risk is at the forefront of the discussion. The goal is improvement of function after spinal surgery for obese and morbidly obese while minimizing complications surrounding the perioperative period. Risk reduction to these patients remains ideal, and further studies are needed to identify whether preoperative weight reduction strategies for non-neurologically compromised individuals will be necessary. Patients are often unable to lose weight because of their inability to exercise. Laparoscopic bariatric treatment of this population may be promising.
Our study evaluated patients’ preoperative antihypertensive, antiarrhythmic, antiplatelet, and anticoagulation medications. We established a sum of preoperative heart medications to categorize the antihypertensive treatment. Bivariate analysis showed angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, β-blockers, calcium channel blockers, and statins influence length of hospital stay (P < .05). The preoperative heart medications sum showed significance for length of hospital stay on linear regression (P = .018). Bivariate analysis showed that aspirin influenced length of hospital stay, but the other antiplatelet, anticoagulants, or antiarrhythmic agents did not. Therefore, based on BMI being a controlled variable, the number of preoperative heart medications is a determinant for hospital stay. This value can be used by surgeons to educate their patients before surgery.
In our regression analyses, age and sex were placed in the model as forced entry irrespective of significance. When BMI was placed in the multivariate linear regression analysis for hospital stay as a continuous variable, it was not significant. The reason is that there is a hyperbolic relationship in BMI with respect to hospital stay on bivariate analysis. That is, underweight individuals behaved as the obese individuals with respect to length of stay. Therefore, we categorized BMI as ideal and not ideal (dichotomization) and found a significant increase in hospital stay. Furthermore, when BMI was placed as a continuous variable in the multivariate linear regression for complications at 1 year, it approached significance, but was not significant (P = .07). The reason that it was not significant was due to the data not being powered to detect a single increase in complication rate based on a single increase in BMI. However, because bivariate analysis showed a linear relationship, if we categorize BMI as greater than 30 or not (dichotomization), we see a clear statistical significance with complications at 1 year.
Our study found rehospitalization rates approaching 20% between follow-up visits. This includes patients requiring revision spinal surgery, hospitalization for unrelated medical issues, and wound care/wound infection issues. Major complications after long-segment fusions are not uncommon, and this provides a benchmark for values for patients between follow-up visits. Classifying patients in BMI categories allows clinicians to counsel patients regarding their potential risk of surgical fusions of more than 5 levels. Ideal body weight individuals will have a shorter hospital stay than those who are not. Obese and morbidly obese patients will have more complications at 1 year than those who are not of this body type. This study provides a framework for surgeons to inform their patients of the effect that weight plays on hospital stay and complications.
There were 15 unplanned returns to the operating room, but no significance was found based on BMI categories. The returns to the operating room were as follows: 6 patients for wound infection, 3 patients for malpositioned screws, 3 patients for foraminotomy, 2 patients for medical events in the perioperative period, and 1 patient for a retained drain.
This study highlights the importance of BMI in spinal fusions, as there is a significant increase in hospital stay and complications at both 1 and 2 years from the time of surgery. This does not discourage surgical intervention for just cause, but does provide insight into the preoperative discussion and potential planning for the postoperative care. Furthermore, although ODI improves at both 1 and 2 years for the obese and morbidly obese, it is worse than the ideal weight patients. The rate of change in the ODI for each weight category overall showed improvement in the interval between follow-up visits. Overall, there is quantitative improvement in the ODI with surgery regardless of BMI; however, having a higher BMI yields a worse ODI at a specific time point. The ODI in the morbidly obese population improved from the preoperative assessment to the early follow-up visit, improvement at the 1-year visit, but slight worsening at the 2-year visit (but not to the score of the preoperative visit). This shows that morbidly obese patients still benefit from surgery; however, their score will not improve as much as if they had a lower BMI. Furthermore, even though complication rates were higher for morbidly obese patients compared with ideal weight patients at 2 years, there is an improvement in ODI.
Although this study had a relatively large sample size, it is limited by being a retrospective study with the inherent selection bias. In retrospective analyses, the investigator depends on the availability and accuracy of the medical record. Additional limitations include surgeon technique and preferences, in-hospital management, and surgical indications, all of which can vary in retrospective studies involving more than 1 surgeon. Linear regression controls for confounders, but data accumulated from a randomized, controlled trial are superior. It was a single-center study with follow-up of at least 1 year with a mean and median follow-up of 2.1 and 2.0 years, respectively. The strengths of this study include the judicious data acquisition and multivariate analysis, controlling for confounders. It features exhaustive data collection regarding potential factors affecting outcomes of hospital stay and complications at 1 and 2 years. This study is the first of its kind in this specific population of patients undergoing fusion of 5 or more levels.
The impact of BMI after major spinal surgery has been examined with improvement of ODI across weight classifications. BMI is an independent predictor of hospital stay and complications at both 1 and 2 years after surgery.
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
1. Rosen DS, Ferguson SD, Ogden AT, et al.. Obesity and self-reported outcome after minimally invasive lumbar spinal fusion surgery. Neurosurgery. 2008;63(5):956–960; discussion 960.
2. Ogden CL, Carroll MD, Curtin LR, et al.. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006;295(13):1549–1555.
3. Djurasovic M, Bratcher KR, Glassman SD, et al.. The effect of obesity on clinical outcomes after lumbar fusion. Spine (Phila Pa 1976). 2008;33(16):1789–1792.
4. Shamji MF, Parker S, Cook C, et al.. Impact of body habitus on perioperative morbidity associated with fusion of the thoracolumbar and lumbar spine. Neurosurgery. 2009;65(3):490–498; discussion 498.
5. Hensrud DD, Klein S. Extreme obesity: a new medical crisis in the United States. Mayo Clin Proc. 2006;81(10 suppl):S5–S10.
6. Pender JR, Pories WJ. Epidemiology of obesity in the United States. Gastroenterol Clin North Am. 2005;34(1):1–7.
7. Patel N, Bagan B, Vadera S, et al.. Obesity and spine surgery: relation to perioperative complications. J Neurosurg Spine. 2007;6(4):291–297.
8. Gepstein R, Shabat S, Arinzon ZH, et al.. Does obesity affect the results of lumbar decompressive spinal surgery in the elderly? Clin Orthop Relat Res. 2004;426:138–144.
9. Upasani VV, Caltoum C, Petcharaporn M, et al.. Does obesity affect surgical outcomes in adolescent idiopathic scoliosis? Spine (Phila Pa 1976). 2008;33(3):295–300.
10. Leboeuf-Yde C, Kyvik KO, Bruun NH. Low back pain and lifestyle. Part II-obesity. Information from a population-based sample of 29,424 twin subjects. Spine (Phila Pa 1976). 1999;24(8):779–783; discussion 783-784.
11. Yadla S, Malone J, Campbell PG, et al.. Obesity and spine surgery: reassessment based on a prospective evaluation of perioperative complications in elective degenerative thoracolumbar procedures. Spine J. 2010;10(7):581–587.
12. Deyo RA, Bass JE. Lifestyle and low-back pain. The influence of smoking and obesity. Spine (Phila Pa 1976). 1989;14(5):501–506.
13. Dindo D, Muller MK, Weber M, et al.. Obesity in general elective surgery. Lancet. 2003;361(9374):2032–2035.
14. Olsen MA, Mayfield J, Lauryssen C, et al.. Risk factors for surgical site infection in spinal surgery. J Neurosurg. 2003;98(2 suppl):149–155.
15. Telfeian AE, Reiter GT, Durham SR, et al.. Spine surgery in morbidly obese patients. J Neurosurg. 2002;97(1 suppl):20–24.
16. Wimmer C, Gluch H, Franzreb M, et al.. Predisposing factors for infection in spine surgery: a survey of 850 spinal procedures. J Spinal Disord. 1998;11(2):124–128.
17. Cole JS, Jackson TR. Minimally invasive lumbar discectomy in obese patients. Neurosurgery. 2007;61(3):539–544; discussion 544.
18. Kalanithi PA, Arrigo R, Boakye M. Morbid obesity increases cost and complication rates in spinal arthrodesis. Spine (Phila Pa 1976). 2012;37(11):982–988.
19. Andreshak TG, An HS, Hall J, et al.. Lumbar spine surgery in the obese patient. J Spinal Disord. 1997;10(5):376–379.
20. Hanigan WC, Elwood PW, Henderson JP, et al.. Surgical results in obese patients with sciatica. Neurosurgery. 1987;20(6):896–899.
21. Glassman S, Gornet MF, Branch C, et al.. MOS short form 36 and Oswestry Disability Index outcomes in lumbar fusion: a multicenter experience. Spine J. 2006;6(1):21–26.
22. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. National Institutes of Health. Obes Res. 1998;(6 suppl 2):51S–209S.
23. Vaidya R, Carp J, Bartol S, et al.. Lumbar spine fusion in obese and morbidly obese patients. Spine (Phila Pa 1976). 2009;34(5):495–500.
24. Halpin RJ, Sugrue PA, Gould RW, et al.. Standardizing care for high-risk patients in spine surgery: the Northwestern high-risk spine protocol. Spine (Phila Pa 1976). 2010;35(25):2232–2238.
25. Singh AK, Ramappa M, Bhatia CK, et al.. Less invasive posterior lumbar interbody fusion and obesity: clinical outcomes and return to work. Spine (Phila Pa 1976). 2010;35(24):2116–2120.
26. Friedmann JM, Elasy T, Jensen GL. The relationship between body mass index and self-reported functional limitation among older adults: a gender difference. J Am Geriatr Soc. 2001;49(4):398–403.
27. Heliovaara M. Body height, obesity, and risk of herniated lumbar intervertebral disc. Spine (Phila Pa 1976). 1987;12(5):469–472.
28. Bostman OM. Body mass index and height in patients requiring surgery for lumbar intervertebral disc herniation. Spine (Phila Pa 1976). 1993;18(7):851–854.
29. Peng CW, Bendo JA, Goldstein JA, et al.. Perioperative outcomes of anterior lumbar surgery in obese versus non-obese patients. Spine J. 2009;9(9):715–720.
30. Park P, Upadhyaya C, Garton HJ, et al.. The impact of minimally invasive spine surgery on perioperative complications in overweight or obese patients. Neurosurgery. 2008;62(3):693–699; discussion 699.
31. Upasani VV, Caltoum C, Petcharaporn M, et al.. Adolescent idiopathic scoliosis patients report increased pain at five years compared with two years after surgical treatment. Spine (Phila Pa 1976). 2008;33(10):1107–1112.
The authors report a retrospective review of spine surgery patients undergoing >5 levels of fusion at a tertiary care facility, classify their patients by BMI, and assess the impact of BMI on complication occurrence. In this report, the authors note that BMI is an independent predictor of length of stay and increased complication occurrence.
The authors address current problem in spinal surgery, the risks associated with long fusion procedures on an overweight or obese patient population. Their results indicate a higher risk of complications and revision surgery for nonideal weight patients. The current literature on the topic is lacking such study as stated by the authors as most published studies focused on short-level constructs. Controlling for comorbidities that accompany the obesity on a given patient is also important to isolate the risk of a procedure.
The greatest weakness of the analysis is its retrospective nature. This limits any conclusions that may be drawn, especially appreciating the poor quality of complication data captured in retrospective chart reviews. The literature clearly shows that retrospective reviews underestimate complication incidence.
Perhaps the authors will address in future reports the impact of BMI on overall patient outcomes or the impact of complication occurrence and patient clinical results. Each of these would be interesting further analyses. A prospective approach to capturing complication occurrence would be optimal. With those allowances, this is a worthwhile contribution to the spinal surgery literature and provides some insight into a significant public health concern.
Nicholas G. Phaneuf
The authors are to be applauded for their thorough analysis how body habitus affects perioperative and postoperative outcomes after multilevel (>5 levels) spinal fusion procedures. This work is of significance, and although the perioperative course of morbidly obese patients is well reported to be more complicated than their normal body habitus counterparts, this group also reports the long-term disability functional outcomes that are needed for both physicians and patients to make informed surgical decisions.
Through a retrospective review, the authors have concluded that increasing BMI is predictive of both duration of hospitalization as well as overall complications sustained by 1 year in the context of patients undergoing long-segment spinal fusion constructs. Further, they comment that although substantial Oswestry Disability Index (ODI) benefit is conferred to underweight through obese patients, morbid obesity is associated with a clinical improvement albeit to a lesser extent. All body habitus groups, including the morbidly obese, did meet minimum clinically important differences in ODI scores after these interventions. Should such patients then be offered surgery because of the observed clinical benefits, or should their care for degenerative pathology be delayed until weight loss measures can be implemented to decrease their perioperative risk and resource utilization? That is the essence of informed consent and evidence-based surgeon decision making.
Although this is a retrospective study based on 1 center's experience, the data are important to better understand how body habitus affects long-segment spinal fusion complications and outcomes. What is lacking and for which this study must serve as an impetus is developing strategies for complication avoidance if trying to surgically manage the complex, morbidly obese patient.
Mohammed F. Shamji