Background
Obesity is a chronic medical condition characterized by the accumulation of excess body fat. It is associated with an increased risk of adverse health events and mortality [
1]. Approximately one-third of adults worldwide were classified as overweight or obese in 2013 [
2]. In the United States, the most recent data in 2018 shows that the prevalence of obesity in adults was 42.4%, representing a significant increase from the 30.5% reported in 2000 [
3]. This is concerning as there is a multitude of health issues linked to obesity, including but not limited to type 2 diabetes mellitus, cardiovascular disease, obstructive sleep apnea, metabolic dysfunction-associated steatotic liver disease (MASLD) [
2,
4].
In the intensive care unit (ICU), the number of obese people getting admitted to ICU for any cause is also increasing and has become a growing concern, with an estimated prevalence of 20% [
5]. The association of obesity with worse outcome in critically ill patient needs further studies, and the association depends on the outcome being measured. Obesity is associated with increased ICU stay and morbidity, but its association with mortality is not as clear [
6,
7]. The term “obesity paradox” refers to this apparent conflicting correlation between obesity and some patient outcomes. Despite the increased risk in debilitating health conditions, research has shown that obese individuals exhibit better survival rates as compared to their non-obese counterparts in certain acute presentations such as sepsis and acute respiratory distress syndrome (ARDS) [
5,
8]. In other words, the paradox is a phenomenon in which obesity appears to be associated with worse outcomes in certain conditions, and with better outcome in others. This paradox has been primarily observed in patients with heart failure [
9], end-stage kidney disease [
10], ARDS [
11], and sepsis [
12], but could also apply to a vast number of other conditions. Pneumonia also seems to be one of those conditions that fit the description of the paradox. Obesity was associated with an increased risk of contracting pneumonia but has been shown to be associated with decrease mortality in this population [
13,
14].
Ventilator-associated pneumonia (VAP) is a type of pneumonia that develops in a patient more than 48 h after endotracheal intubation [
15]. It is estimated that the prevalence of VAP ranges from 5% up to 40% in any given ICU, depending on the setting and diagnostic criteria [
16,
17]. It is also known that VAP increases the length of stay, healthcare costs, morbidity, and mortality of ICU patients [
15,
18].VAP has several well-known risk factors, such as immunodeficiency, supine position, prolonged mechanical ventilation duration, recent antibiotic use, and gastric colonization by pathogenic bacteria [
15,
19]. Very little data exists looking at the relationship between obesity and VAP. Still, studies so far suggest that there is no significant difference in the incidence of VAP between obese and non-obese populations [
20,
21]. This indicates that obesity does not seem to be among the many risk factors for VAP. Despite that, obesity’s effect on VAP patients’ outcomes remains far less clear and not nearly as studied. Our paper presents a nationwide study that utilized a large dataset to delve deeper into the impact of obesity on ICU patients with ventilator-associated pneumonia, specifically analyzing whether obesity increases the length of stay, healthcare costs, and mortality rates or whether the obesity paradox also applies to this specific population.
Discussion
The relationship between obesity and the outcomes of patients with VAP remains unclear. Research on the subject is still scarce, with conflicting results so far. Our study found no significant differences in outcomes between obese and non-obese ICU patients with ventilator-associated pneumonia.
In our study, half of the 4,227 patients with ventilator-associated pneumonia, half of the included patients were of white ethnicity. As per the National Institutes of Health (NIH), obesity affects 17.4 to 42% of the general population, depending on the race. However, in our study, only up to 11% of the patients per ethnic group with VAP were classified as obese. Obesity was noted in 9% of our overall population, with a higher proportion of obesity among women. In contrast, when examining the NIH statistics for the general population, 42% of adults were obese in the USA, with a predominance of obesity among men. Additionally, 56% of obese compared to 42% of non-obese patients, had a Charlson Comorbidity Index score of 3 and more, suggesting a higher burden of comorbidities compared to the non-obese group. Specifically, diabetes, hypertension, and chronic kidney disease were significantly more prevalent among obese patients. A recent retrospective analysis using the NIS captured 33,140 VAP diagnoses [
24]. The authors likely used VAP as a secondary rather than a primary diagnosis. Although they mentioned the principal diagnostic categories of hospitalizations with VAP, since the latter is a secondary diagnosis, it would be challenging to determine whether it occurred during the referenced principal diagnosis or in a previous admission. In their study, obesity was found in 18.1% of cases, but they did not categorize obesity according to different BMI classes. They also did not report the impact of obesity on mortality. Even with a smaller sample size of VAP hospitalizations, our study stands out for several reasons. We classified obesity based on BMI and reported its impact as an independent factor on mortality (adjusted odds ratio). Additionally, we investigated the impact of obesity on in-patient tracheostomy, an indicator of ventilator liberation, which has not been described in VAP hospitalizations before.
We found that obesity did not have a significant effect on mortality in ICU patients with VAP. Though the results were not statistically significant (10% vs. 8.5%,
p-value 0.61, adjusted OR 1.2,
p-value 0.61), obese patients had higher mortality rates. The fact that the adjusted odds ratio was not significant implies the observation of higher mortality in obese patients is not due to obesity itself but rather to its associated comorbidities. This throws doubt on the understanding that obesity is an advantage in the ICU. Obesity has traditionally been associated with lower mortality in ICU patients [
7,
25‐
27]. Although the potential protective mechanisms are not yet fully understood, it has been suggested that the increased energy reserves stored in the excess adipose tissue in obesity may confer an advantage in short-term high-stress situations. Additionally, adipose tissue is now recognized to play more complex roles beyond simple energy storage; these include the release of hormones like leptin and adiponectin, as well as inflammatory mediators like TNF-alpha, IL-6, resistin, and visfatin, which all play essential roles in metabolism and inflammation [
7,
28]. On the other hand, some studies have found that obesity is associated with higher mortality in the ICU population, as one might expect [
29,
30].
Obesity is associated with prolonged duration of mechanical ventilation. The standard practice suggests performing a tracheostomy after prolonged intubation, and institution haves different policies guiding tracheostomy placement as clear-cut recommendations are lacking. Consequently, one might expect a higher likelihood of tracheostomy among obese compared to non-obese patients. Our study echoed this theory, as we found that rates of tracheostomy were higher in the obese group (5 (1.27%) vs. 10 (0.25%), p-value 0.14). This result was not statistically significant, and as such, obesity was not found to be a risk factor for long-term respiratory failure and need for tracheostomy. Our study is the first to address the rate of tracheostomy in obese patients with VAP.
A recent meta-analysis by Peres et al. exploring the risk factor for longer length of stay (LOS) found that mechanical ventilation, hypomagnesemia, delirium, and malnutrition were the most significant variables to have an effect [
31]. Although examined in 7 of the 89 papers studied, BMI did not show a consistent effect [
31]. In our study, the average LOS stay was 11.2 days in non-obese patients vs. 12 days in the obese group. Those with mild and extreme obesity had a shorter stay (8 & 9.6 days, respectively). Paradoxically, patients with a BMI of 40–50 had a longer LOS (17.9 days). In the linear regression analysis, we found that mild obesity contributed to a shorter LOS by 2.98 days on average, as opposed to moderate obesity, which elongated the LOS by an adjusted mean of 7.45 days (
p-value < 0.05). Such inconsistency in the results among the obesity subgroups calls for future studies on the topic of LOS.
We found that the obese group had a higher total admission charge (150,930 vs. 166,876 $). Similar to the LOS, the subgroup analysis had conflicting results. Patients in the BMI group 30–40 and above 50 had a lower total charge (122,920 & 106,048, respectively). The result was higher in the moderately obese group (260,697). In the linear regression analysis none of the BMI subgroups significantly impacted the total charge as an independent factor (Table
4). Direct studies that seek to analyze the effect of obesity on the cost of hospitalization of patients with ventilator-associated pneumonia are non-existent; our study is the first to our knowledge. It is difficult to draw a direct association between obesity, cost, and ventilator-associated pneumonia as our research does not account for other reasons for hospitalization and their potential impact on the cost. However, it remains important to note that mild obesity was shown to have the least overall cost of hospitalization, likely due to fewer comorbidities complicating the hospital stay, in concordance with the estimated BMI-related medical expenditure reported by Ward et al. [
32].
We found that obesity did not have a clear protective or harmful impact on patients with VAP. Such results neither support nor refute the obesity paradox theory. Several theories can explain the reasoning behind such findings. First, many studies in the past have adjusted for ICU admission categories like medical, surgery, and trauma [
7,
25,
26]. However, these broad categorizations and varieties may exist within each one. Even if we were to hypothesize that obesity has some effect on the outcomes of ICU patients, it cannot be assumed that it would have the same impact on all types of ICU patients (those with sepsis, ARDS, heart failure, VAP, etc.). Second, complex confounders such as quality of care can provide the obese group with advantages and disadvantages at the same time. Quality of care depends on factors like geographic location, socioeconomic status, insurance type, hospital quality, aggressiveness of treatment, and systemic biases against obese patients. For example, Arabi et al. found that the decreased mortality in obese sepsis patients can be explained by more aggressive sepsis interventions in the obese [
33]. On the other hand, it has been known for some time that obesity may negatively impact the attitudes of healthcare professionals toward patients and the quality of care they receive [
34‐
36]. Unfortunately, such confounders are very difficult to account for. Our study looked at insurance type, which was not significantly different between the obese and non-obese groups.
Numerous prior studies demonstrate the presence of racial disparities in the healthcare system in the United States. For example, black patients are noted to have higher rates of emergency department visits for both asthma [
37] and heart failure [
38]exacerbations. According to a systematic review published in Critical Care Medicine, Black patients were found to have higher mortality rates [
39]. However, this effect was attributed to age, severity of illness, hospital type, and socioeconomic status rather than race as an independent factor. Our data shows that Black and Hispanic obese patients with VAP had higher mortality rates, but their respective adjusted odds ratios were not statistically significant. Therefore, similar to the data from McGowan et al., the difference in mortality rates is likely related to confounders rather than race. In the non-obese group, Native Americans with VAP had the highest rate of mortality compared to the other races; however, the adjusted odds ratio was again statistically insignificant.
There are some limitations to our study. We conducted a retrospective cohort study using the NIS, an administrative database that limits the uniformity of VAP diagnosis with potential misclassification secondary to using the ICD-10 CM codes. The diagnosis of VAP is extremely variable, enough to render comparisons between hospitals very limited, even when standardized cases eliminate variability in clinical data abstraction [
23]. Second, because of the administrative nature of our database, our analysis did not include the severity of disease at ICU admission or the types, dosages, and frequencies of different VAP-specific treatments during ICU stay. Third, we could not consider different measures of adiposity like waist circumference and waist-to-hip ratio, height estimates, or weight fluctuations during ICU stay. The primary cause of mortality is not specified in NIS; therefore, we presented the total all-cause mortality. Our study continued into 2020, during the time when the COVID-19 pandemic had already begun. Unfortunately, COVID-19 cases were underreported, which limits our ability to accurately determine the frequency of infections. To address this limitation, we conducted a subanalysis over a four-year period. The results indicate no significant difference in the number of VAP patients between the two groups across the four years (
p-value 0.7, Table
1).Since this is not a prospective study, the conclusion that obesity does not significantly impact the outcomes of VAP should be viewed with caution. Despite these limitations, this study is an important contribution to the current understanding of the obesity paradox. We investigated how obesity impacted rates and odds of tracheostomy and the total hospital charges, both of which are not traditionally included in studies of the obesity paradox. We also are the first to report data on racial disparities in the outcomes of VAP in both the obese and non-obese cohorts.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.