- Email: [email protected]
Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients
Clinical Nutrition ESPEN xxx (2018) 1e6
Contents lists available at ScienceDirect
Clinical Nutrition ESPEN journal homepage: http://www.clinicalnutritionespen.com
Original article
Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients Amartya Mukhopadhyay a, *, Yanika Kowitlawakul b, Jeyakumar Henry c, Venetia Ong a, Claudia Shu-Fen Leong c, Bee Choo Tai d a
Division of Respiratory and Critical Care Medicine, University Medicine Cluster, National University Health System and National University of Singapore, Singapore Alice Lee Centre for Nursing Studies, National University Health System and National University of Singapore, Singapore c Clinical Nutrition Research Centre, Singapore Institute for Clinical Sciences, Singapore d Saw Swee Hock School of Public Health, National University of Singapore and National University Health System, Singapore b
a r t i c l e i n f o
s u m m a r y
Article history: Received 19 August 2018 Accepted 21 August 2018
Background & aims: Asians with similar body mass index (BMI) as the Caucasians are at higher health risk as compared to their counterparts. Although the mean weight of patients admitted to the intensive care unit (ICU) is increasing but the relation between BMI with 28-day mortality and length of stay (LOS) following ICU discharge in Asian patients is not well studied. Methods: We included all adult patients admitted to the ICU of a tertiary hospital who received mechanical ventilation (MV) for at least 48 hours between October 2013 and September 2014. Demographics, BMI, MV, comorbidities, ICU scores (Acute Physiology And Chronic Health Evaluation (APACHE) II and sequential organ failure assessment (SOFA)), use of vasopressor, renal replacement therapy and calorie supplementation were collected from the ICU database. BMI was categorized into four groups according to the World Health Organization's Asian BMI recommendation. Post-ICU LOS (days) was calculated from ICU discharge to hospital discharge in hospital survivors. We used multivariable logistic regression to identify factors associated with 28-day mortality and post-ICU LOS of more than 7 days. Results: In a cohort of 273 patients (male 62%, mean age 58.4 ± 17 years), the prevalence of overweight/ obesity was 53%. In the bivariate analysis, 28-day mortality was lower (p ¼ 0.014) and post-ICU LOS longer (p ¼ 0.01) in the overweight/obese groups. In the multivariable logistic regression analysis, APACHE II (Odds ratio, OR 1.10, CI 1.05e1.16), SOFA (OR 1.17, CI 1.05e1.31), duration of MV (days, OR 1.14, CI 1.05e1.25) were associated with increased and higher BMI groups (p < 0.001) with decreased 28-day mortality. Further analysis of 196 hospital survivors showed age (OR 1.04, CI 1.02e1.06), duration of MV (days, OR 1.14, CI 1.02e1.27) and higher Asian BMI (p ¼ 0.042) were associated with longer post-ICU LOS. The odds of longer post-ICU LOS amongst overweight and obese patients were 1.27 (CI 0.59e2.73) and 1.62 (CI 0.69e3.81) times that of those with normal BMI respectively. Conclusion: In multiethnic critically ill Asian patients, the prevalence of overweight/obesity was high. Although higher BMI was associated with reduced risk of 28-day mortality, obese patients stayed significantly longer in the hospital following ICU discharge. © 2018 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.
Keywords: Body mass index Intensive care unit Mortality Length of stay
1. Introduction The world population is becoming heavier and the prevalence of obesity has increased over the last 3 decades [1,2]. Between 1980 * Corresponding author. NUHS Tower block, Level 10, 1E Kent Ridge Road 119228, Singapore. E-mail address: [email protected] (A. Mukhopadhyay).
and 2008, the global mean BMI has risen by 0.4 kg/m2/decade in men and 0.5 kg/m2/decade in women respectively [1]. Asia, the most populous continent, is the new epicenter of the obesity epidemic [3] with the rise of BMI above the global mean. The increase in BMI amongst Asia-Pacific males was 0.6 kg/m2/decade and 1 kg/m2/decade amongst Southeast Asian females [1]. Obesity in Asians differs from those of European origin in terms of higher percentage of body fat even with similar body mass index (BMI). As
https://doi.org/10.1016/j.clnesp.2018.08.009 2405-4577/© 2018 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
2
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
such, the World Health Organization (WHO) has recommended different BMI cut-offs for the Asian population [4]. Higher visceral fat content in Asian population is associated with increased insulin resistance and predisposes the Asians to distinctly higher degree of cardiovascular and diabetes risk [5]. In parallel with the rising obesity in the community, there is a trend of increased BMI amongst patients admitted to the intensive care unit (ICU). A study on 5-year trend of ICU practices in the USA has shown annualized BMI increase of 0.071 kg/m2 [6]. Although epidemiological studies have suggested that both low and high BMIs were associated with increased all-cause mortality [7,8], patients admitted to ICUs with higher BMI were found to have survival advantage [9,10], a phenomenon often termed “obesity paradox”. Data from Asian ICU in this regard is sparse; a Korean study had shown higher mortality in underweight and lower mortality in obese respectively [11]. The trajectories of patients after ICU discharge, particularly the relationship between obesity and post-ICU length of stay (LOS) in the hospital is also unknown. In this study, we investigated the relation between Asian BMI with (i) 28-day mortality in mechanically ventilated patients, and their (ii) post-ICU LOS. 2. Materials and methods 2.1. Patient and setting This is a post-hoc study of a subgroup of patients who received mechanical ventilation (MV). Full detail of the method was described in our previously published paper [12]. Briefly, all adult patients (aged 18 years and above) who were admitted to the medical ICU (MICU) of a tertiary university affiliated hospital between October 2013 and September 2014 and received MV for at least 48 hours were included in the study. According to the MICU protocol, we routinely attempt to start enteral feeding by nasogastric tube within the first 48 hours of MV, unless contraindicated. Energy, ranging from 25 to 30 kcal/kg/day, was given via continuous enteral or parenteral feeding. For obese patients, we implemented adjusted body weight for calorie requirements. Length of stay following ICU discharge (post-ICU LOS) was calculated as days from ICU discharge to hospital discharge in hospital survivors only. 2.2. Data collection IntelliSpace Critical Care and Anesthesia (ICCA, Philips Healthcare) was used as the clinical information system in ICU where all data including demographics, height and weight on admission, ICU scores (Acute Physiology And Chronic Health Evaluation (APACHE) II and sequential organ failure assessment (SOFA)), use of vasopressor and renal replacement therapy (RRT) were recorded.
Comorbidities, diagnosis, outcomes and LOS were extracted from a computerized database used in the hospital. We calculated calorie adequacy as calorie received divided by calorie recommended during MV or up to 12 days, whichever occurred first [13]. The study was approved by institutional review board and ethics committee (Domain Specific Review Board, National Healthcare Group ref: 2014/00209). 2.3. Statistical analysis The cohort was divided according to WHO Asian BMI groups (underweight: BMI < 18.5 kg/m2, normal: BMI 18.5e22.9 kg/m2, overweight: BMI 23e24.9 kg/m2 and obese: BMI 25 kg/m2) and differences between these subgroups were compared using c2 test or one-way analysis of variance (ANOVA) as appropriate. For continuous variables that were not normally distributed, the medians were compared using KruskaleWallis test. To determine factors associated with post-ICU LOS, we dichotomized the hospital survivors based on median post-ICU LOS. The median LOS following ICU discharge amongst hospital survivors was 7 days. Variables which were significant in the bivariate analyses (p 0.01) were considered for further inclusion in the multivariable logistic regression analysis to identify significant predictors of 28-day mortality and post-ICU LOS. Model selection was formally conducted using the likelihood ratio test. The final model was adjusted for the possible confounding effect of age which was classified into 4 categories: 18e40, 41e60, 61e80 and > 80 years. Based on the final logistic regression model that was generated for each outcome, cubic splines with 10 knots were plotted to depict the association between Asian BMI and the predicted probabilities of 28-day mortality and post-ICU LOS respectively. All statistical analyses were evaluated assuming a 2-sided test at the 5% level of significance, using Stata version 15 (STATA Corp., College Station, TX). 3. Results 3.1. Study population A total of 273 patients fulfilled the inclusion/exclusion criteria for this study. As shown in Fig. 1, patients were classified according to Asian BMI groups [4] as follows: underweight (n ¼ 40, 14.7%), normal (n ¼ 88,32.2%), overweight (n ¼ 89,32.6%) and obese (n ¼ 56,20.5%). Their baseline characteristics were summarized according to Asian BMI groups in Table 1. The cohort was predominantly male (61.9%) and of Chinese ethnicity (53%). Prevalence of combined overweight and obese group was 53%. No significant differences were noted amongst the groups in terms of
Fig. 1. Patient inclusion divided by Asian body mass index groups and 28-day outcome.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
3
Table 1 Patient characteristics by Asian BMI. Characteristics
Gender, n(%) Male Female Age, mean (SD),years Age group, years, n(%) 18e40 41e60 61e80 >80 Race, n(%) Chinese Malay Indian Others Comorbidities n(%) 0e1 2e3 >3 ICU severity scores mean, (SD) APACHE II SOFA Treatment, n(%) Vasopressor Renal replacement therapy Calorie adequacy during ventilation, mean (SD) Length of stay, median (IQR), days Pre-ICU ICU Post ICUa Hospitala Duration of Mechanical ventilation, median (IQR), days 28-day mortality, n(%)
All
Asian BMI Underweight
Normal
Overweight
Obese
P
N ¼ 273
N ¼ 40
N ¼ 88
N ¼ 89
N ¼ 56
169 (61.9) 104 (38.1) 58.39 (17.07)
25 (62.50) 15 (37.50) 56.89 (18.36)
48 (54.55) 40 (45.45) 56.76 (18.94)
64 (28.09) 25 (71.91) 59.79 (16.42)
32 (57.14) 24 (42.86) 59.84 (13.9)
44 (16.12) 93 (34.07) 105 (38.46) 31 (11.36)
7 (17.50) 15 (37.50) 12 (30.00) 6 (15.00)
18 (20.45) 31 (35.23) 30 (34.09) 9 (10.23)
14 24 41 10
(15.73) (26.97) (46.07) (11.24)
5 (8.93) 23 (41.07) 22 (39.29) 6 (10.71)
145 (53.11) 65 (23,81) 33 (12.09) 30 (10.99
24 (60.00) 6 (15.00) 5 (12.50) 5 (12.50)
54 (61.36) 20 (22.73) 7 (7.95) 7 (7.95)
49 (55.06) 19 (21.35) 12 (13.48) 9 (10.11)
18 (12.41) 20 (30.77) 9 (27.27) 9 (30.00)
150 (54.95) 111 (40.66) 12 (4.4)
23 (57.50) 15 (37.50) 2 (5.00)
56 (63.64) 29 (32.95) 3 (3.41)
46 (51.69) 38 (42.70) 5 (5.62)
25 (44.64) 29 (51.79) 2 (3.57)
27.3 (8.03) 9.94 (3.52)
27.67 (7.92) 9.75 (3.65)
25.45 (8.15) 9.39 (3.74)
28.46 (7.76) 10.29 (3.56)
28.02 (8.01) 10.39 (2.95)
0.973 0.278
156 (57.14) 60 (21.98) 0.454 (0.35)
23 (57.50) 6 (15.00) 0.611 (0.439)
46 (52.27) 19 (21.59) 0.529 (0.343)
51 (57.30) 20 (22.47) 0.361 (0.329)
36 (64.29) 15 (26.79) 0.375 (0.219)
0.568 0.591 <0.001
0 (0e1) 4 (2e7) 7 (3.5e17) 15 (8e29) 2.5 (1.33e4.62)
0 (0e3) 5 (2e7.5) 5 (1e7) 14 (8e19) 2.23 (1.27e4.78)
0 (0e1) 4.5 (3e8.5) 6 (2e17) 11 (7e30) 2.75 (1.33e4.60)
0 (0e2) 4 (2e7) 9 (4e19) 15 (7e32) 2.2 (1.33e4.62)
0 (0e1) 4 (2e7) 10 (5e32) 17 (8.5e35.5) 3.08 (1.39e5.00)
0.692 0.961 0.010 0.560 0.730
69 (25.27)
14 (35)
30 (34.09)
15 (16.85)
10 (17.86)
0.014
0.096
0.081 0.505
0.080
0.390
Abbreviations: BMI Body mass index, ICU Intensive care unit, SD Standard deviation, APACHE Acute Physiology and Chronic Health Evaluation, SOFA Sequential organ failure assessment, IQR Interquartile range. a Hospital survivors only (N ¼ 196).
gender, age, ethnicity, comorbidities, APACHE II and SOFA scores, use of vasopressor and RRT, duration of MV and pre-ICU/ICU/ hospital LOS in the 4 Asian BMI groups. However, 28-day mortality (p ¼ 0.014) and calorie adequacy (p < 0.001) were reduced in patients with higher BMI while post-ICU LOS in the hospital was longer (p ¼ 0.01).
3.2. 28-day mortality In the unadjusted analysis, higher BMI was associated with lower 28-day mortality (Table 1). The 28-day mortality was 16.9% and 17.9% in overweight and obese patients respectively as compared to 34.1% in patients with normal BMI (Fig. 1 and Table 1). Age-adjusted multivariable analysis (Table 2) shows that APACHE II score (Odds ratio, OR 1.10, 95% CI 1.05e1.16), SOFA score (OR 1.17, 95% 1.05e1.31), duration of MV (days, OR 1.14, 95% CI 1.05e1.25) and Asian BMI were independent predictors of 28-day mortality. Taking normal BMI as the reference, both overweight (OR 0.22, 95% CI 0.10e0.51) and obese (OR 0.25, 95% CI 0.10e0.63) had significantly reduced mortality. From the above multivariable logistic regression model, we obtained the predicted probability of 28-day mortality by Asian BMI groups for each of the four age categories (18e40, 41e60, 61e80 and 80 years). Fig. 2 suggests that amongst the younger patients (aged 18e40 years) the decline in predicted probability of
28-day mortality with increased BMI was rather gradual. However, with advancing age, a distinct U shape was observed, where the points indicating lowest mortality (open circles) were shifted toward higher BMI for age groups 41e60, 61e80 and 80 years respectively. This suggests that elderly patients with higher BMI have better survival advantage as compared to their younger counterparts. Table 2 Multivariable regression of 28-day mortality. 28-day Mortality (n ¼ 273)
APACHE II SOFA Duration of MV, day Asian BMI Normal Underweight Overweight Obese Age 18-40 40-60 60-80 >¼ 80
Odds Ratio
95% CI
p
1.10 1.17 1.14
1.05e1.16 1.05e1.31 1.05e1.25
<0.001 0.006 0.003 <0.001
1.00 0.85 0.22 0.25
Ref 0.34e2.13 0.10e0.51 0.10e0.63
1.00 1.83 1.92 4.21
Ref 0.56e5.95 0.61e6.10 1.05e16.88
0.735 <0.001 0.003 0.225 0.315 0.267 0.042
Abbreviations: APACHE Acute Physiology And Chronic Health Evaluation, SOFA Sequential organ failure assessment, BMI Body mass index.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
4
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6 Table 4 Multivariable regression analysis of post ICU hospital stay. Post-ICU length of stay in hospital survivors (n ¼ 196) Age Duration of MV, day Asian BMI Normal Underweight Overweight Obese
1.04 1.14
1.02e1.06 1.02e1.27
1.00 0.31 1.27 1.62
Ref 0.10e0.98 0.59e2.73 0.69e3.81
<0.001 0.021 0.042 0.045 0.545 0.267
Abbreviation: ICU Intensive care unit, MV mechanical ventilation, BMI Body mass index.
Fig. 2. Relation of Asian BMI and predicted 28-day mortality by age groups.
both overweight (OR 1.27, CI 0.59e2.73) and obese (OR1.62, CI 0.69e3.81) patients have higher odds of longer post-ICU LOS. Fig. 3 shows relation of predicted 28-day mortality and post-ICU LOS by Asian BMI. With increased BMI, predicted probability of 28day mortality was decreased while that of post-ICU LOS was increased.
3.3. Post-ICU hospital stay 4. Discussion One hundred and ninety-six patients who survived till hospital discharge were included in the analysis of post-ICU LOS. Their median (IQR) post-ICU LOS was 7 (3.5e17) days. In the bivariate analysis age, comorbidity, Asian BMI, use of vasopressor and duration of MV ventilation were significantly associated with longer post-ICU LOS (Table 3). In the multivariable analysis, only age, duration of MV and Asian BMI were significantly associated with longer post-ICU LOS (Table 4). With normal BMI as reference,
In this study, higher BMI was found to have a protective effect on 28-day mortality. With normal BMI as the reference, both overweight (p < 0.001) and obese (p ¼ 0.003) patients have significantly reduced risk of 28-day mortality (Table 2). Other significant predictors of 28-day mortality were APACHE and SOFA scores, and duration of mechanical ventilation. The predicted probability of 28day mortality from our multivariable logistic regression model
Table 3 Factors affecting post ICU hospital stay. Characteristics
Gender, n(%) Male Female Age, mean (SD), years Age group, years, n(%) <65 years 65 years Race, n(%) Chinese Malay Indian Others Comorbidities, n(%) 0e1 2e3 >3 BMI, mean (SD), kg/m2 Asian BMI, n(%) Underweight Normal Overweight Obese ICU severity scores, mean (SD) APACHE II SOFA Treatment, n(%) Vasopressor Renal replacement therapy Calorie adequacy during ventilation, mean (SD) ICU Length of stay, median (IQR),days Duration of Mechanical ventilation, median (IQR), days
All
Post ICU Length of stay 7 days
>7 days
N ¼ 196
N ¼ 99 (50.51%)
97 (49.49%)
120 (61.22) 76 (38.78) 56.89 (17.23)
60 (60.61) 39 (39.39) 52.24 (18.84)
60 (61.86) 37 (38.14) 62.77 (13.07)
128 (65.31) 68 (34.69)
75 (75.76) 24 (24.24)
53 (54.64) 44 (45.36)
98 48 26 24
50 26 12 11
48 22 14 13
P
0.858
<0.001 0.002
0.879 (50.00) (24.49) (13.27) (12.24)
(50.51) (26.26) (12.12) (11.11)
(49.48) (22.68) (14.43) (13.40) 0.014
118 (60.20) 71 (36.22) 7 (3.57) 24.8 (6.68)
69 (69.70) 26 (26.26) 4 (4.04) 23.65 (6.45)
49 (50.52) 45 (46.39) 3 (3.09) 25.99 (6.73)
25 57 70 44
19 32 31 17
6 (6.19) 25 (25.77) 39 (40.21) 27 (27.84)
(12.76) (29.08) (35.71) (22.45)
(19.19) (32.32) (31.31) (17.17)
0.014 0.013
25.60 (7.4) 9.24 (3.32)
24.16 (7.39) 8.51 (3.16)
27.07 (7.15) 9.98 (9.32)
0.997 0.999
97 (50.51) 35 (17.86) 0.45 (0.33) 4 (2e7) 2.08 (1.25e4.02)
42 (42.42) 13 (13.13) 0.44 (0.33) 3 (2e6) 1.75 (1.13e3.17)
55 (56.70) 22 (22.68) 0.47 (0.34) 5 (3e8) 2.88 (1.46e4.89)
0.046 0.081 0.765 0.003 <0.001
Abbreviation: ICU Intensive Care Unit; BMI body mass index; SD standard deviation; APACHE Acute Physiology and Chronic Health Evaluation; SOFA Sequential Organ Failure Assessment score; IQR interquartile range.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
Fig. 3. Predicted 28-day mortality and post-ICU LOS by Asian BMI.
showed elderly patients with higher BMI had mortality advantage over their younger counterparts (Fig. 2). On the other hand, compared to normal BMI, overweight and obese patients who survived till hospital discharge have higher risk of longer stay following ICU discharges. With normal BMI as the reference, overweight and obese patients have 1.27 and 1.62 times increased odds of longer post-ICU LOS respectively (Table 4). One year increase in age and each additional day of MV also increased the risk of longer post-ICU LOS by 4% and 14% respectively. In summary, as the BMI increases the predicted 28-day mortality was reduced (p < 0.001) but predicted post-ICU LOS increased (p ¼ 0.042) (Fig. 3). Large epidemiological studies on the association between bodyweight and all-cause mortality have yielded mixed results. While some studies have shown increased mortality in obese patients [7,8], a recent meta-analysis has suggested a protective effect of overweight (BMI 25-<30) for all-cause mortality [14]. The protective effect of higher BMI in ICU population has been documented previously [10,15,16]; however, data from Asian countries is relatively sparse [11] and with mixed results being reported [17]. Asians have higher abdominal adiposity even at similar BMI as compared to their European counterparts and such abdominal obesity has been shown to be associated with higher risk of ICU mortality [18]; it is possible that the protective effect of obesity seen in the critically ill western patients may not be seen in Asian patients. However, our study results are in line with previous western studies showing that overweight and obesity indeed conferred protective effect on short-term mortality in critically ill Asian patients. Age had a profound effect on the relationship between weight and mortality: a U-shaped curve was observed in elderly patients while this curve was flattened in younger subjects, an effect also seen in previous cohort [19]. However, our data on the subgroup analysis showing differential effect of age on the short-term mortality is merely hypothesis-generating at this stage; larger studies using different ICU patient population and various case-mix will be necessary to prove this effect. We also found that overweight and obese patients stayed longer in the hospital following ICU discharges as compared to patients with normal BMI. Overweight and obese patients usually have baseline functional limitations before ICU admissions [20] and it is likely that they will have further functional decline during the ICU stay. However, we did not find any difference in the duration of MV and ICU LOS amongst patients with higher BMI, although a previous meta-analysis had showed the contrary [15]. Functional outcome status at ICU discharge was not collected as part of this study and
5
hence future studies may be warranted to investigate their possible effects on post-ICU length LOS in overweight and obese patients. A study has also shown greater dependence of overweight patients on MV [21] although this effect was not observed in our cohort. Obese patients were also at higher risk of ICU-acquired catheter and blood stream infections [22,23] that might prolong their postICU stay. Why obese patients in the ICU have an overall protective effect with respect to short-term mortality outcome is not clearly understood and several possible explanations have been postulated. It is possible that higher BMI indeed offer survival benefit because of the “nutritional reserve” in these patients [24]. Fat cells secrete several anti-inflammatory cytokines including adipokines, leptin and interleukin-10 that may offer survival advantage in critically ill patients [25]. Leptin has regulatory effect of interferon production and T lymphocyte population, and higher leptin is found in sepsis survivors. Lower BMI can be a manifestation of chronic underlying disease and may offer survival disadvantage to this group (reverse causation) [26]. Indeed improved survival of obese patients is also seen in other chronic wasting diseases including heart failure, chronic obstructive pulmonary disease and chronic kidney disease [27e29]. It is postulated that in wasting diseases, short-term effect of malnutrition-inflammation-cachexia complex in low BMI patients outweighs the detrimental effect of obesity-overnutrition in patients with high BMI [30]. In our study, we have included a multiethnic patient cohort, therefore the data will represent most Asian population. Overweight/obese patients formed a large part of the sample, making it relevant to the internal population. The data collection was exhaustive and to ensure data integrity and quality, we checked the data extensively for accuracy and completeness, including verification of outliers. In particular, we recorded the actual bodyweight of the patient for the computation of BMI before re-categorization. The association between post-ICU LOS and BMI is not well studied particularly in the Asian population and the current study opens up further research possibilities on this important question. Our study has several limitations. This was a single center observational study involving mainly medical patients. This may introduce selection bias in a study with small sample size. Although the data collection was extensive, we can not rule out the possibility of any unobserved confounder. However, our results are in line with previous literature on this topic therefore reassuring us about the validity of the outcomes. 5. Conclusion In a multiethnic cohort of critically ill patients in Asia, overweight and obesity were highly prevalent. Although overweight and obese patients may encounter other medical issues in the ICU, they experienced improved survival outcome at 28-day but required longer post-ICU stay in the hospital. Clinicians should be vigilant about modifiable risk factors for post-ICU hospital stay. It is likely that age modifies the beneficial effect of higher BMI with elderly patients having better 28-day outcomes as compared to younger patients. Funding sources This study was done without any external funding. Conflicts of interest Authors declare no conflict of interest.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
6
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
Statement of authorship
Amartya Mukhopadhyay Yanika Kowitlawakul Jeyakumar Henry Venetia Ong Claudia Shu-Fen Leong Tai Bee Choo
Conception and design of the study
Acquisition of data
Analysis and interpretation of data
Drafting the article
Revising it critically for important intellectual content
Final approval of the version to be submitted
x x x x x x
x x
x
x
x
x
x x x x x x
x x x x x x
x x
Acknowledgments We thank ICU nurse and pharmacist for their help during the study period.
[14]
[15]
References [1] Finucane MM, Stevens GA, Cowan MJ, Danaei G, Lin JK, Paciorek CJ, et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet (London, England) 2011;377:557e67. [2] Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000. Jama 2002;288:1723e7. [3] Yoon KH, Lee JH, Kim JW, Cho JH, Choi YH, Ko SH, et al. Epidemic obesity and type 2 diabetes in Asia. Lancet (London, England) 2006;368:1681e8. [4] Consultation WHOE. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet (London, England) 2004;363:157e63. [5] Park YW, Allison DB, Heymsfield SB, Gallagher D. Larger amounts of visceral adipose tissue in Asian Americans. Obes Res 2001;9:381e7. [6] Lilly CM, Swami S, Liu X, Riker RR, Badawi O. Five-year trends of critical care practice and outcomes. Chest 2017;152:723e35. [7] Zheng W, McLerran DF, Rolland B, Zhang X, Inoue M, Matsuo K, et al. Association between body-mass index and risk of death in more than 1 million Asians. N Engl J Med 2011;364:719e29. [8] Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath Jr CW. Body-mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med 1999;341: 1097e105. [9] Pepper DJ, Sun J, Welsh J, Cui X, Suffredini AF, Eichacker PQ. Increased body mass index and adjusted mortality in ICU patients with sepsis or septic shock: a systematic review and meta-analysis. Crit Care (London, England) 2016;20:181. [10] Sakr Y, Alhussami I, Nanchal R, Wunderink RG, Pellis T, Wittebole X, et al. Being overweight is associated with greater survival in ICU patients: results from the intensive care over Nations audit. Crit Care Med 2015;43:2623e32. [11] Lim SY, Choi WI, Jeon K, Guallar E, Koh Y, Lim CM, et al. Body mass index and mortality in Korean intensive care units: a prospective multicenter cohort study. PLoS One 2014;9, e90039. [12] Mukhopadhyay A, Henry J, Ong V, Leong CS, Teh AL, van Dam RM, et al. Association of modified NUTRIC score with 28-day mortality in critically ill patients. Clin Nutr (Edinburgh, Scotland) 2017;36:1143e8. [13] Heyland DK, Dhaliwal R, Jiang X, Day AG. Identifying critically ill patients who benefit the most from nutrition therapy: the development and initial
[16]
[17] [18]
[19] [20]
[21]
[22] [23]
[24]
[25] [26] [27]
[28]
[29]
[30]
validation of a novel risk assessment tool. Crit Care (London, England) 2011;15:R268. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. Jama 2013;309:71e82. Akinnusi ME, Pineda LA, El Solh AA. Effect of obesity on intensive care morbidity and mortality: a meta-analysis. Crit Care Med 2008;36:151e8. Hogue Jr CW, Stearns JD, Colantuoni E, Robinson KA, Stierer T, Mitter N, et al. The impact of obesity on outcomes after critical illness: a meta-analysis. Intensive Care Med 2009;35:1152e70. Lim SY, Kim SI, Ryu YJ, Lee JH, Chun EM, Chang JH. The body mass index as a prognostic factor of critical care. Kor J Intern Med 2010;25:162e7. Paolini JB, Mancini J, Genestal M, Gonzalez H, McKay RE, Samii K, et al. Predictive value of abdominal obesity vs. body mass index for determining risk of intensive care unit mortality. Crit Care Med 2010;38:1308e14. Hughes V. The big fat truth. Nature 2013;497:428e30. Prescott HC, Chang VW, O'Brien Jr JM, Langa KM, Iwashyna TJ. Obesity and 1year outcomes in older Americans with severe sepsis. Crit Care Med 2014;42: 1766e74. Porhomayon J, Papadakos P, Singh A, Nader ND. Alteration in respiratory physiology in obesity for anesthesia-critical care physician. HSR Proc Intensive Care Cardiovasc Anesth 2011;3:109e18. Serrano PE, Khuder SA, Fath JJ. Obesity as a risk factor for nosocomial infections in trauma patients. J Am Coll Surg 2010;211:61e7. Dossett LA, Dageforde LA, Swenson BR, Metzger R, Bonatti H, Sawyer RG, et al. Obesity and site-specific nosocomial infection risk in the intensive care unit. Surg Infect 2009;10:137e42. Hurt RT, Frazier TH, McClave SA, Kaplan LM. Obesity epidemic: overview, pathophysiology, and the intensive care unit conundrum. JPEN J Parenter Enter Nutr 2011;35:4Se13S. Hauner H. Adipose tissue inflammation: are small or large fat cells to blame? Diabetologia 2010;53:223e5. Manson JE, Stampfer MJ, Hennekens CH, Willett WC. Body weight and longevity. A reassessment. Jama 1987;257:353e8. Marti S, Munoz X, Rios J, Morell F, Ferrer J. Body weight and comorbidity predict mortality in COPD patients treated with oxygen therapy. Eur Respir J 2006;27:689e96. Kovesdy CP, Anderson JE, Kalantar-Zadeh K. Paradoxical association between body mass index and mortality in men with CKD not yet on dialysis. Am J Kidney Dis Off J Natl Kidney Found 2007;49:581e91. Davos CH, Doehner W, Rauchhaus M, Cicoira M, Francis DP, Coats AJ, et al. Body mass and survival in patients with chronic heart failure without cachexia: the importance of obesity. J Card Fail 2003;9:29e35. Kalantar-Zadeh K, Horwich TB, Oreopoulos A, Kovesdy CP, Younessi H, Anker SD, et al. Risk factor paradox in wasting diseases. Curr Opin Clin Nutr Metab Care 2007;10:433e42.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
Contents lists available at ScienceDirect
Clinical Nutrition ESPEN journal homepage: http://www.clinicalnutritionespen.com
Original article
Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients Amartya Mukhopadhyay a, *, Yanika Kowitlawakul b, Jeyakumar Henry c, Venetia Ong a, Claudia Shu-Fen Leong c, Bee Choo Tai d a
Division of Respiratory and Critical Care Medicine, University Medicine Cluster, National University Health System and National University of Singapore, Singapore Alice Lee Centre for Nursing Studies, National University Health System and National University of Singapore, Singapore c Clinical Nutrition Research Centre, Singapore Institute for Clinical Sciences, Singapore d Saw Swee Hock School of Public Health, National University of Singapore and National University Health System, Singapore b
a r t i c l e i n f o
s u m m a r y
Article history: Received 19 August 2018 Accepted 21 August 2018
Background & aims: Asians with similar body mass index (BMI) as the Caucasians are at higher health risk as compared to their counterparts. Although the mean weight of patients admitted to the intensive care unit (ICU) is increasing but the relation between BMI with 28-day mortality and length of stay (LOS) following ICU discharge in Asian patients is not well studied. Methods: We included all adult patients admitted to the ICU of a tertiary hospital who received mechanical ventilation (MV) for at least 48 hours between October 2013 and September 2014. Demographics, BMI, MV, comorbidities, ICU scores (Acute Physiology And Chronic Health Evaluation (APACHE) II and sequential organ failure assessment (SOFA)), use of vasopressor, renal replacement therapy and calorie supplementation were collected from the ICU database. BMI was categorized into four groups according to the World Health Organization's Asian BMI recommendation. Post-ICU LOS (days) was calculated from ICU discharge to hospital discharge in hospital survivors. We used multivariable logistic regression to identify factors associated with 28-day mortality and post-ICU LOS of more than 7 days. Results: In a cohort of 273 patients (male 62%, mean age 58.4 ± 17 years), the prevalence of overweight/ obesity was 53%. In the bivariate analysis, 28-day mortality was lower (p ¼ 0.014) and post-ICU LOS longer (p ¼ 0.01) in the overweight/obese groups. In the multivariable logistic regression analysis, APACHE II (Odds ratio, OR 1.10, CI 1.05e1.16), SOFA (OR 1.17, CI 1.05e1.31), duration of MV (days, OR 1.14, CI 1.05e1.25) were associated with increased and higher BMI groups (p < 0.001) with decreased 28-day mortality. Further analysis of 196 hospital survivors showed age (OR 1.04, CI 1.02e1.06), duration of MV (days, OR 1.14, CI 1.02e1.27) and higher Asian BMI (p ¼ 0.042) were associated with longer post-ICU LOS. The odds of longer post-ICU LOS amongst overweight and obese patients were 1.27 (CI 0.59e2.73) and 1.62 (CI 0.69e3.81) times that of those with normal BMI respectively. Conclusion: In multiethnic critically ill Asian patients, the prevalence of overweight/obesity was high. Although higher BMI was associated with reduced risk of 28-day mortality, obese patients stayed significantly longer in the hospital following ICU discharge. © 2018 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.
Keywords: Body mass index Intensive care unit Mortality Length of stay
1. Introduction The world population is becoming heavier and the prevalence of obesity has increased over the last 3 decades [1,2]. Between 1980 * Corresponding author. NUHS Tower block, Level 10, 1E Kent Ridge Road 119228, Singapore. E-mail address: [email protected] (A. Mukhopadhyay).
and 2008, the global mean BMI has risen by 0.4 kg/m2/decade in men and 0.5 kg/m2/decade in women respectively [1]. Asia, the most populous continent, is the new epicenter of the obesity epidemic [3] with the rise of BMI above the global mean. The increase in BMI amongst Asia-Pacific males was 0.6 kg/m2/decade and 1 kg/m2/decade amongst Southeast Asian females [1]. Obesity in Asians differs from those of European origin in terms of higher percentage of body fat even with similar body mass index (BMI). As
https://doi.org/10.1016/j.clnesp.2018.08.009 2405-4577/© 2018 European Society for Clinical Nutrition and Metabolism. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
2
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
such, the World Health Organization (WHO) has recommended different BMI cut-offs for the Asian population [4]. Higher visceral fat content in Asian population is associated with increased insulin resistance and predisposes the Asians to distinctly higher degree of cardiovascular and diabetes risk [5]. In parallel with the rising obesity in the community, there is a trend of increased BMI amongst patients admitted to the intensive care unit (ICU). A study on 5-year trend of ICU practices in the USA has shown annualized BMI increase of 0.071 kg/m2 [6]. Although epidemiological studies have suggested that both low and high BMIs were associated with increased all-cause mortality [7,8], patients admitted to ICUs with higher BMI were found to have survival advantage [9,10], a phenomenon often termed “obesity paradox”. Data from Asian ICU in this regard is sparse; a Korean study had shown higher mortality in underweight and lower mortality in obese respectively [11]. The trajectories of patients after ICU discharge, particularly the relationship between obesity and post-ICU length of stay (LOS) in the hospital is also unknown. In this study, we investigated the relation between Asian BMI with (i) 28-day mortality in mechanically ventilated patients, and their (ii) post-ICU LOS. 2. Materials and methods 2.1. Patient and setting This is a post-hoc study of a subgroup of patients who received mechanical ventilation (MV). Full detail of the method was described in our previously published paper [12]. Briefly, all adult patients (aged 18 years and above) who were admitted to the medical ICU (MICU) of a tertiary university affiliated hospital between October 2013 and September 2014 and received MV for at least 48 hours were included in the study. According to the MICU protocol, we routinely attempt to start enteral feeding by nasogastric tube within the first 48 hours of MV, unless contraindicated. Energy, ranging from 25 to 30 kcal/kg/day, was given via continuous enteral or parenteral feeding. For obese patients, we implemented adjusted body weight for calorie requirements. Length of stay following ICU discharge (post-ICU LOS) was calculated as days from ICU discharge to hospital discharge in hospital survivors only. 2.2. Data collection IntelliSpace Critical Care and Anesthesia (ICCA, Philips Healthcare) was used as the clinical information system in ICU where all data including demographics, height and weight on admission, ICU scores (Acute Physiology And Chronic Health Evaluation (APACHE) II and sequential organ failure assessment (SOFA)), use of vasopressor and renal replacement therapy (RRT) were recorded.
Comorbidities, diagnosis, outcomes and LOS were extracted from a computerized database used in the hospital. We calculated calorie adequacy as calorie received divided by calorie recommended during MV or up to 12 days, whichever occurred first [13]. The study was approved by institutional review board and ethics committee (Domain Specific Review Board, National Healthcare Group ref: 2014/00209). 2.3. Statistical analysis The cohort was divided according to WHO Asian BMI groups (underweight: BMI < 18.5 kg/m2, normal: BMI 18.5e22.9 kg/m2, overweight: BMI 23e24.9 kg/m2 and obese: BMI 25 kg/m2) and differences between these subgroups were compared using c2 test or one-way analysis of variance (ANOVA) as appropriate. For continuous variables that were not normally distributed, the medians were compared using KruskaleWallis test. To determine factors associated with post-ICU LOS, we dichotomized the hospital survivors based on median post-ICU LOS. The median LOS following ICU discharge amongst hospital survivors was 7 days. Variables which were significant in the bivariate analyses (p 0.01) were considered for further inclusion in the multivariable logistic regression analysis to identify significant predictors of 28-day mortality and post-ICU LOS. Model selection was formally conducted using the likelihood ratio test. The final model was adjusted for the possible confounding effect of age which was classified into 4 categories: 18e40, 41e60, 61e80 and > 80 years. Based on the final logistic regression model that was generated for each outcome, cubic splines with 10 knots were plotted to depict the association between Asian BMI and the predicted probabilities of 28-day mortality and post-ICU LOS respectively. All statistical analyses were evaluated assuming a 2-sided test at the 5% level of significance, using Stata version 15 (STATA Corp., College Station, TX). 3. Results 3.1. Study population A total of 273 patients fulfilled the inclusion/exclusion criteria for this study. As shown in Fig. 1, patients were classified according to Asian BMI groups [4] as follows: underweight (n ¼ 40, 14.7%), normal (n ¼ 88,32.2%), overweight (n ¼ 89,32.6%) and obese (n ¼ 56,20.5%). Their baseline characteristics were summarized according to Asian BMI groups in Table 1. The cohort was predominantly male (61.9%) and of Chinese ethnicity (53%). Prevalence of combined overweight and obese group was 53%. No significant differences were noted amongst the groups in terms of
Fig. 1. Patient inclusion divided by Asian body mass index groups and 28-day outcome.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
3
Table 1 Patient characteristics by Asian BMI. Characteristics
Gender, n(%) Male Female Age, mean (SD),years Age group, years, n(%) 18e40 41e60 61e80 >80 Race, n(%) Chinese Malay Indian Others Comorbidities n(%) 0e1 2e3 >3 ICU severity scores mean, (SD) APACHE II SOFA Treatment, n(%) Vasopressor Renal replacement therapy Calorie adequacy during ventilation, mean (SD) Length of stay, median (IQR), days Pre-ICU ICU Post ICUa Hospitala Duration of Mechanical ventilation, median (IQR), days 28-day mortality, n(%)
All
Asian BMI Underweight
Normal
Overweight
Obese
P
N ¼ 273
N ¼ 40
N ¼ 88
N ¼ 89
N ¼ 56
169 (61.9) 104 (38.1) 58.39 (17.07)
25 (62.50) 15 (37.50) 56.89 (18.36)
48 (54.55) 40 (45.45) 56.76 (18.94)
64 (28.09) 25 (71.91) 59.79 (16.42)
32 (57.14) 24 (42.86) 59.84 (13.9)
44 (16.12) 93 (34.07) 105 (38.46) 31 (11.36)
7 (17.50) 15 (37.50) 12 (30.00) 6 (15.00)
18 (20.45) 31 (35.23) 30 (34.09) 9 (10.23)
14 24 41 10
(15.73) (26.97) (46.07) (11.24)
5 (8.93) 23 (41.07) 22 (39.29) 6 (10.71)
145 (53.11) 65 (23,81) 33 (12.09) 30 (10.99
24 (60.00) 6 (15.00) 5 (12.50) 5 (12.50)
54 (61.36) 20 (22.73) 7 (7.95) 7 (7.95)
49 (55.06) 19 (21.35) 12 (13.48) 9 (10.11)
18 (12.41) 20 (30.77) 9 (27.27) 9 (30.00)
150 (54.95) 111 (40.66) 12 (4.4)
23 (57.50) 15 (37.50) 2 (5.00)
56 (63.64) 29 (32.95) 3 (3.41)
46 (51.69) 38 (42.70) 5 (5.62)
25 (44.64) 29 (51.79) 2 (3.57)
27.3 (8.03) 9.94 (3.52)
27.67 (7.92) 9.75 (3.65)
25.45 (8.15) 9.39 (3.74)
28.46 (7.76) 10.29 (3.56)
28.02 (8.01) 10.39 (2.95)
0.973 0.278
156 (57.14) 60 (21.98) 0.454 (0.35)
23 (57.50) 6 (15.00) 0.611 (0.439)
46 (52.27) 19 (21.59) 0.529 (0.343)
51 (57.30) 20 (22.47) 0.361 (0.329)
36 (64.29) 15 (26.79) 0.375 (0.219)
0.568 0.591 <0.001
0 (0e1) 4 (2e7) 7 (3.5e17) 15 (8e29) 2.5 (1.33e4.62)
0 (0e3) 5 (2e7.5) 5 (1e7) 14 (8e19) 2.23 (1.27e4.78)
0 (0e1) 4.5 (3e8.5) 6 (2e17) 11 (7e30) 2.75 (1.33e4.60)
0 (0e2) 4 (2e7) 9 (4e19) 15 (7e32) 2.2 (1.33e4.62)
0 (0e1) 4 (2e7) 10 (5e32) 17 (8.5e35.5) 3.08 (1.39e5.00)
0.692 0.961 0.010 0.560 0.730
69 (25.27)
14 (35)
30 (34.09)
15 (16.85)
10 (17.86)
0.014
0.096
0.081 0.505
0.080
0.390
Abbreviations: BMI Body mass index, ICU Intensive care unit, SD Standard deviation, APACHE Acute Physiology and Chronic Health Evaluation, SOFA Sequential organ failure assessment, IQR Interquartile range. a Hospital survivors only (N ¼ 196).
gender, age, ethnicity, comorbidities, APACHE II and SOFA scores, use of vasopressor and RRT, duration of MV and pre-ICU/ICU/ hospital LOS in the 4 Asian BMI groups. However, 28-day mortality (p ¼ 0.014) and calorie adequacy (p < 0.001) were reduced in patients with higher BMI while post-ICU LOS in the hospital was longer (p ¼ 0.01).
3.2. 28-day mortality In the unadjusted analysis, higher BMI was associated with lower 28-day mortality (Table 1). The 28-day mortality was 16.9% and 17.9% in overweight and obese patients respectively as compared to 34.1% in patients with normal BMI (Fig. 1 and Table 1). Age-adjusted multivariable analysis (Table 2) shows that APACHE II score (Odds ratio, OR 1.10, 95% CI 1.05e1.16), SOFA score (OR 1.17, 95% 1.05e1.31), duration of MV (days, OR 1.14, 95% CI 1.05e1.25) and Asian BMI were independent predictors of 28-day mortality. Taking normal BMI as the reference, both overweight (OR 0.22, 95% CI 0.10e0.51) and obese (OR 0.25, 95% CI 0.10e0.63) had significantly reduced mortality. From the above multivariable logistic regression model, we obtained the predicted probability of 28-day mortality by Asian BMI groups for each of the four age categories (18e40, 41e60, 61e80 and 80 years). Fig. 2 suggests that amongst the younger patients (aged 18e40 years) the decline in predicted probability of
28-day mortality with increased BMI was rather gradual. However, with advancing age, a distinct U shape was observed, where the points indicating lowest mortality (open circles) were shifted toward higher BMI for age groups 41e60, 61e80 and 80 years respectively. This suggests that elderly patients with higher BMI have better survival advantage as compared to their younger counterparts. Table 2 Multivariable regression of 28-day mortality. 28-day Mortality (n ¼ 273)
APACHE II SOFA Duration of MV, day Asian BMI Normal Underweight Overweight Obese Age 18-40 40-60 60-80 >¼ 80
Odds Ratio
95% CI
p
1.10 1.17 1.14
1.05e1.16 1.05e1.31 1.05e1.25
<0.001 0.006 0.003 <0.001
1.00 0.85 0.22 0.25
Ref 0.34e2.13 0.10e0.51 0.10e0.63
1.00 1.83 1.92 4.21
Ref 0.56e5.95 0.61e6.10 1.05e16.88
0.735 <0.001 0.003 0.225 0.315 0.267 0.042
Abbreviations: APACHE Acute Physiology And Chronic Health Evaluation, SOFA Sequential organ failure assessment, BMI Body mass index.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
4
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6 Table 4 Multivariable regression analysis of post ICU hospital stay. Post-ICU length of stay in hospital survivors (n ¼ 196) Age Duration of MV, day Asian BMI Normal Underweight Overweight Obese
1.04 1.14
1.02e1.06 1.02e1.27
1.00 0.31 1.27 1.62
Ref 0.10e0.98 0.59e2.73 0.69e3.81
<0.001 0.021 0.042 0.045 0.545 0.267
Abbreviation: ICU Intensive care unit, MV mechanical ventilation, BMI Body mass index.
Fig. 2. Relation of Asian BMI and predicted 28-day mortality by age groups.
both overweight (OR 1.27, CI 0.59e2.73) and obese (OR1.62, CI 0.69e3.81) patients have higher odds of longer post-ICU LOS. Fig. 3 shows relation of predicted 28-day mortality and post-ICU LOS by Asian BMI. With increased BMI, predicted probability of 28day mortality was decreased while that of post-ICU LOS was increased.
3.3. Post-ICU hospital stay 4. Discussion One hundred and ninety-six patients who survived till hospital discharge were included in the analysis of post-ICU LOS. Their median (IQR) post-ICU LOS was 7 (3.5e17) days. In the bivariate analysis age, comorbidity, Asian BMI, use of vasopressor and duration of MV ventilation were significantly associated with longer post-ICU LOS (Table 3). In the multivariable analysis, only age, duration of MV and Asian BMI were significantly associated with longer post-ICU LOS (Table 4). With normal BMI as reference,
In this study, higher BMI was found to have a protective effect on 28-day mortality. With normal BMI as the reference, both overweight (p < 0.001) and obese (p ¼ 0.003) patients have significantly reduced risk of 28-day mortality (Table 2). Other significant predictors of 28-day mortality were APACHE and SOFA scores, and duration of mechanical ventilation. The predicted probability of 28day mortality from our multivariable logistic regression model
Table 3 Factors affecting post ICU hospital stay. Characteristics
Gender, n(%) Male Female Age, mean (SD), years Age group, years, n(%) <65 years 65 years Race, n(%) Chinese Malay Indian Others Comorbidities, n(%) 0e1 2e3 >3 BMI, mean (SD), kg/m2 Asian BMI, n(%) Underweight Normal Overweight Obese ICU severity scores, mean (SD) APACHE II SOFA Treatment, n(%) Vasopressor Renal replacement therapy Calorie adequacy during ventilation, mean (SD) ICU Length of stay, median (IQR),days Duration of Mechanical ventilation, median (IQR), days
All
Post ICU Length of stay 7 days
>7 days
N ¼ 196
N ¼ 99 (50.51%)
97 (49.49%)
120 (61.22) 76 (38.78) 56.89 (17.23)
60 (60.61) 39 (39.39) 52.24 (18.84)
60 (61.86) 37 (38.14) 62.77 (13.07)
128 (65.31) 68 (34.69)
75 (75.76) 24 (24.24)
53 (54.64) 44 (45.36)
98 48 26 24
50 26 12 11
48 22 14 13
P
0.858
<0.001 0.002
0.879 (50.00) (24.49) (13.27) (12.24)
(50.51) (26.26) (12.12) (11.11)
(49.48) (22.68) (14.43) (13.40) 0.014
118 (60.20) 71 (36.22) 7 (3.57) 24.8 (6.68)
69 (69.70) 26 (26.26) 4 (4.04) 23.65 (6.45)
49 (50.52) 45 (46.39) 3 (3.09) 25.99 (6.73)
25 57 70 44
19 32 31 17
6 (6.19) 25 (25.77) 39 (40.21) 27 (27.84)
(12.76) (29.08) (35.71) (22.45)
(19.19) (32.32) (31.31) (17.17)
0.014 0.013
25.60 (7.4) 9.24 (3.32)
24.16 (7.39) 8.51 (3.16)
27.07 (7.15) 9.98 (9.32)
0.997 0.999
97 (50.51) 35 (17.86) 0.45 (0.33) 4 (2e7) 2.08 (1.25e4.02)
42 (42.42) 13 (13.13) 0.44 (0.33) 3 (2e6) 1.75 (1.13e3.17)
55 (56.70) 22 (22.68) 0.47 (0.34) 5 (3e8) 2.88 (1.46e4.89)
0.046 0.081 0.765 0.003 <0.001
Abbreviation: ICU Intensive Care Unit; BMI body mass index; SD standard deviation; APACHE Acute Physiology and Chronic Health Evaluation; SOFA Sequential Organ Failure Assessment score; IQR interquartile range.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
Fig. 3. Predicted 28-day mortality and post-ICU LOS by Asian BMI.
showed elderly patients with higher BMI had mortality advantage over their younger counterparts (Fig. 2). On the other hand, compared to normal BMI, overweight and obese patients who survived till hospital discharge have higher risk of longer stay following ICU discharges. With normal BMI as the reference, overweight and obese patients have 1.27 and 1.62 times increased odds of longer post-ICU LOS respectively (Table 4). One year increase in age and each additional day of MV also increased the risk of longer post-ICU LOS by 4% and 14% respectively. In summary, as the BMI increases the predicted 28-day mortality was reduced (p < 0.001) but predicted post-ICU LOS increased (p ¼ 0.042) (Fig. 3). Large epidemiological studies on the association between bodyweight and all-cause mortality have yielded mixed results. While some studies have shown increased mortality in obese patients [7,8], a recent meta-analysis has suggested a protective effect of overweight (BMI 25-<30) for all-cause mortality [14]. The protective effect of higher BMI in ICU population has been documented previously [10,15,16]; however, data from Asian countries is relatively sparse [11] and with mixed results being reported [17]. Asians have higher abdominal adiposity even at similar BMI as compared to their European counterparts and such abdominal obesity has been shown to be associated with higher risk of ICU mortality [18]; it is possible that the protective effect of obesity seen in the critically ill western patients may not be seen in Asian patients. However, our study results are in line with previous western studies showing that overweight and obesity indeed conferred protective effect on short-term mortality in critically ill Asian patients. Age had a profound effect on the relationship between weight and mortality: a U-shaped curve was observed in elderly patients while this curve was flattened in younger subjects, an effect also seen in previous cohort [19]. However, our data on the subgroup analysis showing differential effect of age on the short-term mortality is merely hypothesis-generating at this stage; larger studies using different ICU patient population and various case-mix will be necessary to prove this effect. We also found that overweight and obese patients stayed longer in the hospital following ICU discharges as compared to patients with normal BMI. Overweight and obese patients usually have baseline functional limitations before ICU admissions [20] and it is likely that they will have further functional decline during the ICU stay. However, we did not find any difference in the duration of MV and ICU LOS amongst patients with higher BMI, although a previous meta-analysis had showed the contrary [15]. Functional outcome status at ICU discharge was not collected as part of this study and
5
hence future studies may be warranted to investigate their possible effects on post-ICU length LOS in overweight and obese patients. A study has also shown greater dependence of overweight patients on MV [21] although this effect was not observed in our cohort. Obese patients were also at higher risk of ICU-acquired catheter and blood stream infections [22,23] that might prolong their postICU stay. Why obese patients in the ICU have an overall protective effect with respect to short-term mortality outcome is not clearly understood and several possible explanations have been postulated. It is possible that higher BMI indeed offer survival benefit because of the “nutritional reserve” in these patients [24]. Fat cells secrete several anti-inflammatory cytokines including adipokines, leptin and interleukin-10 that may offer survival advantage in critically ill patients [25]. Leptin has regulatory effect of interferon production and T lymphocyte population, and higher leptin is found in sepsis survivors. Lower BMI can be a manifestation of chronic underlying disease and may offer survival disadvantage to this group (reverse causation) [26]. Indeed improved survival of obese patients is also seen in other chronic wasting diseases including heart failure, chronic obstructive pulmonary disease and chronic kidney disease [27e29]. It is postulated that in wasting diseases, short-term effect of malnutrition-inflammation-cachexia complex in low BMI patients outweighs the detrimental effect of obesity-overnutrition in patients with high BMI [30]. In our study, we have included a multiethnic patient cohort, therefore the data will represent most Asian population. Overweight/obese patients formed a large part of the sample, making it relevant to the internal population. The data collection was exhaustive and to ensure data integrity and quality, we checked the data extensively for accuracy and completeness, including verification of outliers. In particular, we recorded the actual bodyweight of the patient for the computation of BMI before re-categorization. The association between post-ICU LOS and BMI is not well studied particularly in the Asian population and the current study opens up further research possibilities on this important question. Our study has several limitations. This was a single center observational study involving mainly medical patients. This may introduce selection bias in a study with small sample size. Although the data collection was extensive, we can not rule out the possibility of any unobserved confounder. However, our results are in line with previous literature on this topic therefore reassuring us about the validity of the outcomes. 5. Conclusion In a multiethnic cohort of critically ill patients in Asia, overweight and obesity were highly prevalent. Although overweight and obese patients may encounter other medical issues in the ICU, they experienced improved survival outcome at 28-day but required longer post-ICU stay in the hospital. Clinicians should be vigilant about modifiable risk factors for post-ICU hospital stay. It is likely that age modifies the beneficial effect of higher BMI with elderly patients having better 28-day outcomes as compared to younger patients. Funding sources This study was done without any external funding. Conflicts of interest Authors declare no conflict of interest.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009
6
A. Mukhopadhyay et al. / Clinical Nutrition ESPEN xxx (2018) 1e6
Statement of authorship
Amartya Mukhopadhyay Yanika Kowitlawakul Jeyakumar Henry Venetia Ong Claudia Shu-Fen Leong Tai Bee Choo
Conception and design of the study
Acquisition of data
Analysis and interpretation of data
Drafting the article
Revising it critically for important intellectual content
Final approval of the version to be submitted
x x x x x x
x x
x
x
x
x
x x x x x x
x x x x x x
x x
Acknowledgments We thank ICU nurse and pharmacist for their help during the study period.
[14]
[15]
References [1] Finucane MM, Stevens GA, Cowan MJ, Danaei G, Lin JK, Paciorek CJ, et al. National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet (London, England) 2011;377:557e67. [2] Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000. Jama 2002;288:1723e7. [3] Yoon KH, Lee JH, Kim JW, Cho JH, Choi YH, Ko SH, et al. Epidemic obesity and type 2 diabetes in Asia. Lancet (London, England) 2006;368:1681e8. [4] Consultation WHOE. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet (London, England) 2004;363:157e63. [5] Park YW, Allison DB, Heymsfield SB, Gallagher D. Larger amounts of visceral adipose tissue in Asian Americans. Obes Res 2001;9:381e7. [6] Lilly CM, Swami S, Liu X, Riker RR, Badawi O. Five-year trends of critical care practice and outcomes. Chest 2017;152:723e35. [7] Zheng W, McLerran DF, Rolland B, Zhang X, Inoue M, Matsuo K, et al. Association between body-mass index and risk of death in more than 1 million Asians. N Engl J Med 2011;364:719e29. [8] Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath Jr CW. Body-mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med 1999;341: 1097e105. [9] Pepper DJ, Sun J, Welsh J, Cui X, Suffredini AF, Eichacker PQ. Increased body mass index and adjusted mortality in ICU patients with sepsis or septic shock: a systematic review and meta-analysis. Crit Care (London, England) 2016;20:181. [10] Sakr Y, Alhussami I, Nanchal R, Wunderink RG, Pellis T, Wittebole X, et al. Being overweight is associated with greater survival in ICU patients: results from the intensive care over Nations audit. Crit Care Med 2015;43:2623e32. [11] Lim SY, Choi WI, Jeon K, Guallar E, Koh Y, Lim CM, et al. Body mass index and mortality in Korean intensive care units: a prospective multicenter cohort study. PLoS One 2014;9, e90039. [12] Mukhopadhyay A, Henry J, Ong V, Leong CS, Teh AL, van Dam RM, et al. Association of modified NUTRIC score with 28-day mortality in critically ill patients. Clin Nutr (Edinburgh, Scotland) 2017;36:1143e8. [13] Heyland DK, Dhaliwal R, Jiang X, Day AG. Identifying critically ill patients who benefit the most from nutrition therapy: the development and initial
[16]
[17] [18]
[19] [20]
[21]
[22] [23]
[24]
[25] [26] [27]
[28]
[29]
[30]
validation of a novel risk assessment tool. Crit Care (London, England) 2011;15:R268. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. Jama 2013;309:71e82. Akinnusi ME, Pineda LA, El Solh AA. Effect of obesity on intensive care morbidity and mortality: a meta-analysis. Crit Care Med 2008;36:151e8. Hogue Jr CW, Stearns JD, Colantuoni E, Robinson KA, Stierer T, Mitter N, et al. The impact of obesity on outcomes after critical illness: a meta-analysis. Intensive Care Med 2009;35:1152e70. Lim SY, Kim SI, Ryu YJ, Lee JH, Chun EM, Chang JH. The body mass index as a prognostic factor of critical care. Kor J Intern Med 2010;25:162e7. Paolini JB, Mancini J, Genestal M, Gonzalez H, McKay RE, Samii K, et al. Predictive value of abdominal obesity vs. body mass index for determining risk of intensive care unit mortality. Crit Care Med 2010;38:1308e14. Hughes V. The big fat truth. Nature 2013;497:428e30. Prescott HC, Chang VW, O'Brien Jr JM, Langa KM, Iwashyna TJ. Obesity and 1year outcomes in older Americans with severe sepsis. Crit Care Med 2014;42: 1766e74. Porhomayon J, Papadakos P, Singh A, Nader ND. Alteration in respiratory physiology in obesity for anesthesia-critical care physician. HSR Proc Intensive Care Cardiovasc Anesth 2011;3:109e18. Serrano PE, Khuder SA, Fath JJ. Obesity as a risk factor for nosocomial infections in trauma patients. J Am Coll Surg 2010;211:61e7. Dossett LA, Dageforde LA, Swenson BR, Metzger R, Bonatti H, Sawyer RG, et al. Obesity and site-specific nosocomial infection risk in the intensive care unit. Surg Infect 2009;10:137e42. Hurt RT, Frazier TH, McClave SA, Kaplan LM. Obesity epidemic: overview, pathophysiology, and the intensive care unit conundrum. JPEN J Parenter Enter Nutr 2011;35:4Se13S. Hauner H. Adipose tissue inflammation: are small or large fat cells to blame? Diabetologia 2010;53:223e5. Manson JE, Stampfer MJ, Hennekens CH, Willett WC. Body weight and longevity. A reassessment. Jama 1987;257:353e8. Marti S, Munoz X, Rios J, Morell F, Ferrer J. Body weight and comorbidity predict mortality in COPD patients treated with oxygen therapy. Eur Respir J 2006;27:689e96. Kovesdy CP, Anderson JE, Kalantar-Zadeh K. Paradoxical association between body mass index and mortality in men with CKD not yet on dialysis. Am J Kidney Dis Off J Natl Kidney Found 2007;49:581e91. Davos CH, Doehner W, Rauchhaus M, Cicoira M, Francis DP, Coats AJ, et al. Body mass and survival in patients with chronic heart failure without cachexia: the importance of obesity. J Card Fail 2003;9:29e35. Kalantar-Zadeh K, Horwich TB, Oreopoulos A, Kovesdy CP, Younessi H, Anker SD, et al. Risk factor paradox in wasting diseases. Curr Opin Clin Nutr Metab Care 2007;10:433e42.
Please cite this article in press as: Mukhopadhyay A, et al., Higher BMI is associated with reduced mortality but longer hospital stays following ICU discharge in critically ill Asian patients, Clinical Nutrition ESPEN (2018), https://doi.org/10.1016/j.clnesp.2018.08.009