Anna Maria Bombardieri, MD, PhD is an Anesthesiologist who specializes in Regional Anesthesiology.
She completed her residency training and PhD at the University of Florence.
She completed a fellowship in Regional Anesthesia and Acute Pain Medicine at the Hospital for Special Surgery in New York and a Critical Care Medicine fellowship at Université Libre de Bruxelles, in Belgium.
She was faculty in the Department of Anesthesiology at the Universitè Sophia Antipolis in Nice, France, and at Washington University School of Medicine in St. Louis before joining the department of Anesthesiology at Stanford.
- Regional Anesthesia
Honors & Awards
Best Poster Award, European Society of Neurosonology and Cerebral Hemodynamics (2013)
Charles L. Christian Award for Excellence in Musculoskeletal Research, Hospital for Special Surgery, NY (2012)
Award for Academic Excellence, Hospital for Special Surgery, NY (2012)
Resident Research Contest Award, Post Graduate Assembly in Anesthesiology, NY (2011)
Boards, Advisory Committees, Professional Organizations
Committee on Regional Anesthesia and Acute Pain Medicine, American Society of Anesthesiology (2021 - Present)
Committee on Communications, American Society of Anesthesiology (2021 - Present)
Committee of Women in Anesthesia, American Society of Anesthesiology (2021 - Present)
Member, California Society of Anesthesiology (2020 - Present)
Member, American Society of Regional Anesthesia (2019 - Present)
Member, American Society of Anesthesiology (2018 - Present)
Member, International Anesthesia Research Society (2018 - Present)
Member, European Society of Neurosonolgy and Cerebral Hemodynamics (2014 - Present)
Member, Cerebral Autoregulation Research Network (2012 - Present)
Fellowship: University Libre De Bruxelles (2014) Belgium
PhD, University of Florence School of Medicine, Anesthesiology, Italy (2013)
Fellowship: Hospital for Special Surgery Anesthesia Fellowships (2012) NY
Fellowship, Hospital for Special Surgery, Research, NY (2011)
Board Certification: University of Florence, Anesthesiology (2009)
Residency: University of Florence (2009) Italy
Medical Education: University of Rome "La Sapienza" (2005)
Current Research and Scholarly Interests
My overall research goal is to advance clinical practice by providing anesthesiologists with data to most effectively maintain cerebral blood flow in the perioperative period.
I am interested in the effect of the autonomic nervous system on cerebral blood flow regulation.
I intend to combine regional anesthetic techniques and noninvasive bedside cerebral blood flow monitoring to understand the effect of the sympathetic system on cerebral blood flow.
A secondary goal is to apply this new knowledge to investigate whether cervical sympathetic blocks improve long term neurological outcomes.
Cerebral Blood Flow Evaluation With Trancranial Doppler After Interscalene Nerve Block
The purpose of this study is to assess cerebral blood flow velocity (CBFV) following regional anesthesia for shoulder surgery. The results of this study will help to further understand the physiologic effect of the interscalene nerve block on cerebral blood flow.
Stanford is currently not accepting patients for this trial. For more information, please contact Anna Maria Bombardieri, MD, PhD, 314-326-9107.
Cervical Sympathetic Block in Patients With Cerebral Vasospasm
The purposes of this study are twofold: 1) to assess the effect of a cervical sympathetic block on cerebral blood flow in patients suffering from cerebral vasospasm, after aneurysmal subarachnoid hemorrhage; 2) to evaluate the effect of the sympathetic block on the recovery of the neurological function.
Stanford is currently not accepting patients for this trial. For more information, please contact Anna Maria Bombardieri, MD, PhD, (314) 326-9107.
- Vasopressor Effects on Cerebral Microcirculation: Comment. Anesthesiology 2022; 136 (5): 867-868
The Regional Cerebral Oxygen Saturation Effect of Inotropes/Vasopressors Administered to Treat Intraoperative Hypotension: A Bayesian Network Meta-analysis.
Journal of neurosurgical anesthesiology
One of the main concerns of intraoperative hypotension is adequacy of cerebral perfusion, as cerebral blood flow decreases passively when mean arterial pressure falls below the lower limit of cerebral autoregulation. Treatment of intraoperative hypotension includes administration of drugs, such as inotropes and vasopressors, which have different pharmacological effects on cerebral hemodynamics; there is no consensus on the preferred drug to use. We performed a network meta-analysis (NMA) to pool and analyze data comparing the effect on cerebral oxygen saturation (ScO2) measured by cerebral oximetry of various inotropes/vasopressors used to treat intraoperative hypotension. We searched randomized control trials in Embase, Ovid Medline, Scopus, Cochrane Central Register of Controlled Trials, and Web of Science. We included studies that enrolled adult patients undergoing surgery under general/spinal anesthesia that compared at least 2 inotropes/vasopressors to treat hypotension. We reviewed 51 full-text manuscripts and included 9 randomized controlled trials in our study. The primary outcome was change in ScO2. Our results showed the likelihood that dopamine, ephedrine, and norepinephrine had the lowest probability of decreasing ScO2. The suggested rank order to maintain ScO2, from higher to lower, was dopamine
View details for DOI 10.1097/ANA.0000000000000783
View details for PubMedID 34116546
The effect on cerebral oxygen saturation of vasopressors administered to treat intraoperative hypotension: a Bayesian network meta-analysis
LIPPINCOTT WILLIAMS & WILKINS. 2021: 553-554
View details for Web of Science ID 000752526600235
Practical Training of Anesthesia Clinicians in Electroencephalogram-Based Determination of Hypnotic Depth of General Anesthesia.
Anesthesia and analgesia
2020; 130 (3): 777-786
Electroencephalographic (EEG) brain monitoring during general anesthesia provides information on hypnotic depth. We hypothesized that anesthesia clinicians could be trained rapidly to recognize typical EEG waveforms occurring with volatile-based general anesthesia.This was a substudy of a trial testing the hypothesis that EEG-guided anesthesia prevents postoperative delirium. The intervention was a 35-minute training session, summarizing typical EEG changes with volatile-based anesthesia. Participants completed a preeducational test, underwent training, and completed a posteducational test. For each question, participants indicated whether the EEG was consistent with (1) wakefulness, (2) non-slow-wave anesthesia, (3) slow-wave anesthesia, or (4) burst suppression. They also indicated whether the processed EEG (pEEG) index was discordant with the EEG waveforms. Four clinicians, experienced in intraoperative EEG interpretation, independently evaluated the EEG waveforms, resolved disagreements, and provided reference answers. Ten questions were assessed in the preeducational test and 9 in the posteducational test.There were 71 participants; 13 had previous anesthetic-associated EEG interpretation training. After training, the 58 participants without prior training improved at identifying dominant EEG waveforms (median 60% with interquartile range [IQR], 50%-70% vs 78% with IQR, 67%-89%; difference: 18%; 95% confidence interval [CI], 8-27; P < .001). In contrast, there was no significant improvement following the training for the 13 participants who reported previous training (median 70% with IQR, 60%-80% vs 67% with IQR, 67%-78%; difference: -3%; 95% CI, -18 to 11; P = .88). The difference in the change between the pre- and posteducational session for the previously untrained versus previously trained was statistically significant (difference in medians: 21%; 95% CI, 2-28; P = .005). Clinicians without prior training also improved in identifying discordance between the pEEG index and the EEG waveform (median 60% with IQR, 40%-60% vs median 100% with IQR, 75%-100%; difference: 40%; 95% CI, 30-50; P < .001). Clinicians with prior training showed no significant improvement (median 60% with IQR, 60%-80% vs 75% with IQR, 75%-100%; difference: 15%; 95% CI, -16 to 46; P = .16). Regarding the identification of discordance, the difference in the change between the pre- and posteducational session for the previously untrained versus previously trained was statistically significant (difference in medians: 25%; 95% CI, 5-45; P = .012).A brief training session was associated with improvements in clinicians without prior EEG training in (1) identifying EEG waveforms corresponding to different hypnotic depths and (2) recognizing when the hypnotic depth suggested by the EEG was discordant with the pEEG index.
View details for DOI 10.1213/ANE.0000000000004537
View details for PubMedID 31880629
Intraoperative Awareness With Recall: A Descriptive, Survey-Based, Cohort Study.
Anesthesia and analgesia
2019; 129 (5): 1291-1297
Unintended intraoperative awareness with recall (AWR) is a potential complication of general anesthesia. Patients typically report recollections of (1) hearing sounds or conversations, (2) being unable to breathe or move, (3), feeling pain, and/or (4) experiencing emotional distress. The purpose of the current study was to identify and further characterize AWR experiences identified through postoperative surveys of a large unselected adult surgical cohort.This is a substudy of a prospective registry study, which surveys patients on their health and well-being after surgery. Responses to 4 questions focusing on AWR were analyzed. Patients who reported AWR with pain, paralysis, and/or distress were contacted by telephone to obtain more information about their AWR experience. The interview results for patients who received general anesthesia were sent to 3 anesthesiologists, who adjudicated the reported AWR episodes.Of 48,151 surveys sent, 17,875 patient responses were received. Of these respondents, 622 reported a specific memory from the period between going to sleep and waking up from perceived general anesthesia and 282 of these reported related pain, paralysis, and/or distress. An attempt was made to contact these 282 patients, and 149 participated in a telephone survey. Among the 149 participants, 87 endorsed their prior report of AWR. However, only 22 of these patients had received general anesthesia, while 51 received only sedation and 14 received regional anesthesia. Three anesthesiologists independently adjudicated the survey results of the 22 general anesthesia cases and assigned 6 as definite AWR, 8 as possible AWR, and 8 as not AWR episodes. Of the 65 patients who confirmed their report of AWR after regional or sedation anesthesia, 37 (31 with sedation and 6 with regional anesthesia) had not expected to be conscious during surgery.The complication of AWR continues to occur during intended general anesthesia. Many reports of AWR episodes occur in patients receiving sedation or regional anesthesia and relate to incorrect expectations regarding anesthetic techniques and conscious experiences, representing a potential target for intervention.
View details for DOI 10.1213/ANE.0000000000004358
View details for PubMedID 31453869
An Observational Study of Cerebral Blood Flow Velocity Evaluation in the Prone Position During Posterior Lumbar Surgery.
Anesthesia and analgesia
2019; 129 (2): 487-492
Prone positioning (PP) is necessary for surgical access during posterior spine procedure. However, physiological changes occur in the PP. Typical findings are a decrease in arterial blood pressure and in cardiac output that could potentially lead to an alteration in cerebral perfusion. Therefore, we decided to study cerebral blood flow velocity (CBFV) with transcranial Doppler ultrasonography to evaluate the effect of the PP on cerebral hemodynamics.Twenty-two patients undergoing spine surgery in the PP were studied. General anesthesia was induced using 250 μg of fentanyl, 2 mg/kg of propofol, and 0.1 mg/kg of vecuronium, and was maintained with 0.25%-0.5% isoflurane, 50% nitrous oxide in oxygen, continuous infusion of 100 μg/kg/min of propofol, 1.5 μg/kg/h of fentanyl, and 0.15 mg/kg/h of ketamine. Continuous invasive arterial blood pressure, heart rate, electrocardiogram, and end-tidal carbon dioxide were monitored. CBFV with transcranial Doppler in the middle cerebral artery was first measured with the patients under general anesthesia in the supine position. Patients were then placed in the PP and remained in this position throughout surgery. CBFV, end-tidal carbon dioxide, heart rate, and blood pressure were measured continuously for 75 minutes after initiation of PP. This coincided with surgical exposure and minimal blood loss. Data were analyzed every 15 minutes for statistical significant change over time.Mean arterial blood pressure decreased 15 minutes after the installation of the PP and onward, but this decrease was not statistically significant. CBFVsyst (the maximal CBFV during the systolic phase of a cardiac cycle) and CBFVmean (the time averaged value of the maximal velocity envelope over 1 cardiac cycle) did not vary at any time points. CBFVdiast (the CBFV just before the acceleration phase [systole] of the next waveform) was lower at T3 (30 minutes after PP) compared to T1 (value derived averaging the first measure in the PP with the ones at 5 and 10 minutes) (P = .01), and the pulsatility index was higher at T5 (60 minutes after PP) compared to T0 (baseline, patient supine under general anesthesia) (P = .04). Data were analyzed at specific time points (T0 and T1). This value was derived by computing an average of the CBFV values collected at the first measure in the PP and at 5 and 10 minutes thereafter: T2, 15 minutes after PP; T3, 30 minutes after PP; T4, 45 minutes after PP; T5, 60 minutes after PP; and T6, 75 minutes after PP.Our data on CBFV during PP for spine surgery demonstrate preservation of cerebral perfusion during stable systemic hemodynamic conditions. The present results do not allow us to determine whether the PP would be similarly tolerated with increasing length of surgery, variations in systemic hemodynamics, and in different patient populations.
View details for DOI 10.1213/ANE.0000000000003892
View details for PubMedID 30418236
A comparison of two different concentrations and infusion rates of ropivacaine in perineural infusion administered at the same total dose for analgesia after foot and ankle surgery: a randomized, double blinded, controlled study.
2019; 85 (2): 139-147
Continuous popliteal nerve block is utilized for postoperative analgesia after foot and ankle surgery. Whether only the total dose of local anesthetic or the combination of concentration and volume determine the characteristics of a continuous popliteal nerve infusion remains currently unknown. We hypothesized a reduction of the incidence of insensate extremity in patients given ropivacaine 0.4% at 4 mL/h compared to patients given ropivacaine 0.2% at 8 mL/h.Sixty-four patients scheduled for major foot and ankle surgery requiring a continuous popliteal catheter infusion for postoperative analgesia were studied. Thirty-three patients were randomized to receive a continuous popliteal nerve block with 0.2% (basal 8 mL/h) and thirty-one with 0.4% (basal 4 mL/h) ropivacaine, reaching the same total dose (16 mg/h). The primary outcome was the incidence of persistent sensory block in the distal sciatic nerve distributions in the postoperative period. Secondary outcomes were the incidence of motor block, NRS pain scores at rest in the postoperative period up to 48 hours after surgery, opioid use and related side effects, patients' satisfaction.The incidence of persistent sensory block was similar in patients given 0.2% and in patients receiving 0.4% ropivacaine. The incidence of motor block, postoperative pain scores at rest, use of oral opioids, side effects and patients' satisfaction with the quality of recovery were also similar in both groups.Our results suggest that local anesthetic total dose is the primary determinant of continuous popliteal perineural infusion effects.
View details for DOI 10.23736/S0375-9393.18.12266-8
View details for PubMedID 29633812
An Observational Study of Cerebral Blood Flow Velocity During Hypotensive Epidural Anesthesia.
Anesthesia and analgesia
2016; 122 (1): 226-33
Hypotensive epidural anesthesia (HEA), as practiced at our institution, uses sympathetic blockade to achieve mean arterial blood pressure (MAP) of ≤50 mm Hg while administering epinephrine by infusion to support the circulation. HEA has not been associated with gross adverse effects on neurologic outcome or cognitive function in the postoperative period, suggesting adequate cerebral blood flow (CBF). However, the use of MAPs well below the commonly accepted lower limit of CBF autoregulation suggests that CBF should be significantly reduced below normal levels. To examine these conflicting hypotheses, we performed a prospective investigation of the effects of HEA on CBF velocity (CBFV), an accepted index of cerebral perfusion.Fifty-two hip replacement patients were studied. HEA was induced by lumbar epidural injection of local anesthetic and infusion of epinephrine to achieve an MAP of ≤50 mm Hg. Propofol/midazolam sedation was administered. Baseline CBFV was recorded pre-HEA (after sedation and before local anesthetic injection) and continuously thereafter.During HEA, MAP decreased by 40% and was stable throughout. The CBFVmean at baseline and at 3 HEA intervals during surgery was 46 ± 12 (SD), 45 ± 12, 47 ± 14, and 47 ± 14 cm·s, respectively. Although mean CBFVmean did not vary, there was considerable heterogeneity among patients. Twelve patients (23%) experienced reductions of CBFVmean of >20% during HEA intervals (99% lower confidence limit: 9%) and 6 (12%) reductions of >30% (99% lower confidence limit: 1%). There was no correlation between CBFVmean and MAP for MAPs between 100 and 40 mm Hg (R = 0.0015, P = 0.44). There were no instances of gross postoperative neurologic injury.Both hypotheses proved partially correct. CBFV was sometimes well maintained during HEA, despite MAPs well below the commonly accepted lower limit of autoregulation. However, there was considerable interindividual heterogeneity with 23% of subjects having CBFV reductions >20% (99% lower confidence limit: 9%), with some reductions approaching the threshold for ischemic injury. The present data do not allow us to determine whether hypotension would be similarly tolerated in other circumstances.
View details for DOI 10.1213/ANE.0000000000000985
View details for PubMedID 26465933
Does limb preconditioning reduce pain after total knee arthroplasty? A randomized, double-blind study.
Clinical orthopaedics and related research
2014; 472 (5): 1467-74
Total knee arthroplasty (TKA) can be associated with considerable postoperative pain. Ischemic preconditioning of tissue before inducing procedure-related underperfusion may reduce the postoperative inflammatory response, which further may reduce associated pain.In this prospective, randomized study, we aimed at evaluating the impact of ischemic preconditioning on postoperative pain at rest and during exercise; use of pain medication; levels of systemic prothrombotic and local inflammatory markers; and length of stay and achievement of physical therapy milestones.Sixty patients undergoing unilateral TKA under tourniquet were enrolled with half (N = 30) being randomized to an episode of limb preconditioning before induction of ischemia for surgery (tourniquet inflation). Pain scores, analgesic consumption, markers of inflammation (interleukin-6 [IL-6], tumor necrosis factor [TNF]-α in periarticular drainage), and periarticular circumference were measured at baseline and during 2 days postoperatively. Changes in prothrombotic markers were evaluated.Patients in the preconditioning group had significantly less pain postoperatively at rest (mean difference = -0.71, 95% confidence interval [CI] = -1.40 to -0.02, p = 0.043) and with exercise (mean difference = -1.38, 95% CI = -2.32 to -0.44, p = 0.004), but showed no differences in analgesic consumption. No differences were seen between the study and the control group in terms of muscle oxygenation and intraarticular levels of IL-6 and TNF-α as well as levels of prothrombotic markers. No differences were found between groups in regard to hospitalization length and time to various physical therapy milestones.Ischemic preconditioning reduces postoperative pain after TKA, but the treatment effect size we observed with the preconditioning routine used was modest.Given the ease of this intervention, ischemic preconditioning may be considered as part of a multimodal analgesic strategy. However, more study into the impact of different preconditioning strategies, elucidation of mechanisms, safety profiles, and cost-effectiveness of this maneuver is needed.
View details for DOI 10.1007/s11999-013-3106-4
View details for PubMedID 23761178
View details for PubMedCentralID PMC3971250
Pulmonary circulatory changes after bilateral total knee arthroplasty during regional anesthesia.
Journal of clinical anesthesia
2013; 25 (1): 4-8
To monitor the pulmonary hemodynamics of patients undergoing bilateral total knee arthroplasty (BTKA) intraoperatively and up to 24 hours following surgery.Prospective observational study.University-affiliated teaching hospital.30 ASA physical status 2 and 3 patients scheduled for single-stage, cemented BTKA during epidural anesthesia.Pulmonary artery catheters were in all patients.Systemic vascular resistance (SVR), pulmonary vascular resistance (PVR), the ratio of PVR to SVR at baseline, at the beginning of surgery, and after each knee implantation were recorded and compared with measurements taken one day postoperatively (POD 1).On POD 1, PVR/SVR was increased by 30% compared with baseline (P < 0.0001) and by 20% versus the end of surgery (P < 0.0001). Systemic vascular resistance decreased during surgery and was significantly lower than baseline at 24 hours after surgery (P < 0.0001). No significant change in PVR was noted during surgery.The PVR/SVR ratio on the day following BTKA was increased. This change may represent the different effects of inflammatory perioperative stresses on the pulmonary and systemic vasculature.
View details for DOI 10.1016/j.jclinane.2012.05.004
View details for PubMedID 23391339
View details for PubMedCentralID PMC4009487
Comparative utility of centrally versus peripherally transduced venous pressure monitoring in the perioperative period in spine surgery patients.
Journal of clinical anesthesia
2012; 24 (7): 542-8
To compare central venous pressure (CVP) with peripheral venous pressure (PVP) monitoring during the intraoperative and postoperative periods in patients undergoing spine surgery.Prospective observational study.University-affiliated teaching hospital.35 ASA physical status 1, 2, and 3 patients.A peripheral catheter in the forearm or hand and a central catheter into the internal jugular vein were placed for PVP and CVP monitoring, respectively.CVP and PVP values were collected simultaneously and recorded electronically at 5-minute intervals throughout surgery and in the recovery room. The number of attempts for catheter placement, ease of use, maintenance, and interpretation were recorded. Patient comfort, frequency of complications, and cost were analyzed.The correlation coefficient between CVP and PVP was 0.650 in the operating room (P < 0.0001) and 0.388 in the recovery room (P < 0.0001). There was no difference between groups in number of attempts to place either catheter, maintenance, and interpretation with respect to PVP and CVP monitoring in the operating room. In the recovery room, the nurses reported a higher level of difficulty in interpretation of PVP than CVP, but no differences were noted in ease of maintenance. There were no complications related to either central or peripheral catheter placement. Patient comfort and cost efficiency were higher with a peripheral than a central catheter.During clinically relevant conditions, there was limited correlation between PVP and CVP in the prone position during surgery and postoperatively in the recovery room.
View details for DOI 10.1016/j.jclinane.2012.03.005
View details for PubMedID 22999983
View details for PubMedCentralID PMC3483409
The effect of low versus high tidal volume ventilation on inflammatory markers in healthy individuals undergoing posterior spine fusion in the prone position: a randomized controlled trial.
Journal of clinical anesthesia
2012; 24 (4): 263-9
To evaluate the effect of ventilation strategy on markers of inflammation in patients undergoing spine surgery in the prone position.Randomized controlled trial.University-affiliated teaching hospital.26 ASA physical status 1 and 2 patients scheduled for elective primary lumbar decompression and fusion in the prone position.Patients were randomized to receive mechanical ventilation with either a tidal volume (V(T)) of 12 mL/kg ideal body weight with zero positive end-expiratory pressure (PEEP) or V(T) of 6 mL/kg ideal body weight with PEEP of 8 cm H(2)O.Plasma levels of interleukin (IL)-6 and IL-8 were determined at the beginning of ventilation and at 6 and 12 hours later. Urinary levels of desmosine were determined at the beginning of ventilation and on postoperative days 1 and 3.A significant increase in IL-6, IL-8, and urine desmosine levels was noted over time compared with baseline (P < 0.01). However, no significant difference in the levels of markers was seen between the groups at any time point when controlling for demographics, ASA physical status, body mass index, duration of ventilation, or estimated blood loss.Although markers of inflammation are increased after posterior spine fusion surgery, ventilation strategy has minimal impact on markers of systemic inflammation.
View details for DOI 10.1016/j.jclinane.2011.08.003
View details for PubMedID 22001758
View details for PubMedCentralID PMC3288806
Mortality of patients with respiratory insufficiency and adult respiratory distress syndrome after surgery: the obesity paradox.
Journal of intensive care medicine
2011; 27 (5): 306-11
Obesity has long been considered a risk factor for the development of various pathologies, yet evidence supporting increased risk of perioperative mortality in obese individuals developing postoperative complications is limited. Therefore, we sought to characterize the demographics of obese and nonobese individuals developing postoperative respiratory insufficiency (RI)/adult respiratory distress syndrome (ARDS) and to quantify the impact of obesity on in-hospital mortality among this patient population utilizing data collected for the Nationwide Inpatient Sample (NIS).Nationwide Inpatient Sample data for each year between 1998 and 2007 were accessed. Entries were included if they underwent a surgical procedure and had a diagnosis of RI/ARDS following surgery. Patients fulfilling entry criteria were divided into those with and without obesity. In-hospital mortality was the primary outcome. A logistic regression model was fitted to elucidate if obesity was associated with increased odds for the outcome while controlling for age, gender, admission and procedure type, and comorbidity burden.We identified 9 149 030 admissions that underwent the included surgical procedures between 1998 and 2007. Of those, 5.48% had a diagnosis of obesity. The incidence of RI/ARDS was 1.82% among obese and 2.01% among nonobese patients. Obese patients whose postoperative course was complicated by RI/ARDS had a significantly lower incidence of the need for mechanical ventilation (50% vs 55%). In-hospital mortality was significantly lower compared to nonobese patients (5.45% vs 18.72%). For those patients with RI/ARDS requiring intubation, the in-hospital mortality rate was 11% for obese and 25% for nonobese patients. In the multivariate regression analysis, obesity was associated with a 69% reduction in the odds of in-hospital mortality in postoperative patients with RI/ARDS.In our analysis, obesity was associated with a decreased incidence and adjusted odds for in-hospital mortality after surgery. Our results support the emerging concept of the "obesity paradox."
View details for DOI 10.1177/0885066611411410
View details for PubMedID 21778465