Overweight negatively affects pediatric respiratory function. In this study, we evaluate if overweight is associated with more severe bronchiolitis in hospitalized infants.
This retrospective cohort study analyzed infants aged 30 to 365 days hospitalized for bronchiolitis from September 2019 to April 2020. Exclusion criteria included known risk factors for severe bronchiolitis, asthma treatment, or bacterial pneumonia. Weight-for-length z-score was categorized per the World Health Organization’s growth assessments as overweight (z-score >2), underweight (z-score <−2), and standard weight (between −2 and ≤2). Primary outcomes included respiratory support, ICU stay, and local bronchiolitis score. Secondary outcomes included supplemental interventions.
After exclusion criteria, 385 of 644 infants were categorized as overweight (n = 24), standard (n = 335), or underweight (n = 26). There were differences in need for respiratory support (overweight, 100%; standard weight, 81.8%; underweight, 76.9%; P = .03), highest support of high-flow nasal cannula (overweight, 75%; standard weight, 48%; underweight, 42%; P = .03), admission to ICU (overweight, 54.2%; standard weight, 21.5%; underweight, 34.7%; P < .001), and median bronchiolitis score (overweight, 8 [interquartile range 5–10]; standard weight, 4 [3–7]; underweight, 4 [3–7]; P = .01). Findings remained significant after age adjustments. Additionally, overweight experienced higher frequency of certain treatments.
This study suggests overweight is associated with more severe bronchiolitis in hospitalized infants supported by increased respiratory support level, bronchiolitis scores, and interventions. Higher need for ICU admission may be related to high-flow nasal cannula limitations on the acute care floor.
Overweight is associated with worse outcomes in children and adults with respiratory diseases. The impact of overweight in infants remains less studied. In children and adults, obesity leads to decreased respiratory function, decreased chest wall compliance, and increased upper airway resistance.1–3 For infants, studies have demonstrated negative effects, including decreased ventilation,4–6 increased obstructive sleep events,7 and an association with wheezing in infancy.8,9 A recent study found that increased adiposity was associated with longer lengths of stay (LOS) for infants with bronchiolitis,10 the most common cause for hospitalization in infants.11 Otherwise, little is known about the effects of overweight on infants with bronchiolitis. Severe disease in bronchiolitis is associated with other risk factors, including prematurity, cardiopulmonary disease, low birth weight, young age, low weight, and tobacco exposure.11–14 Our aim was to evaluate the association between overweight infants hospitalized for bronchiolitis, their disease severity, and used medical interventions.
Methods
Design and Population
We conducted a retrospective cohort study of infants admitted to a tertiary children’s hospital in the Midwest with their first episode of bronchiolitis from September 2019 to April 2020. Data were collected from the electronic health record for patients aged 30 to 365 days with the International Classification of Diseases, 10th Revision, code for bronchiolitis. Exclusion criteria consisted of known risk factors for severe disease, including prematurity (aged <37 weeks), chronic lung disease or bronchopulmonary dysplasia, neuromuscular disease, trisomy 21, immunodeficiency, and reactive airway disease or asthma. Additional exclusion criteria included a diagnosis of bacterial pneumonia, use of an inhaled corticosteroid, or placement on our institution’s asthma protocol (which excludes them from the bronchiolitis protocol).
Demographic and background data collected included age, sex, race, ethnicity, admission weight, admission length, gestational age, birth weight, and nutrition at time of hospitalization (breastfed, formula-fed, combination, or cow’s milk).
Weight Definitions
In line with the World Health Organization (WHO) recommendations for those aged <2 years, we used weight-for-length z-scores calculated on the basis of WHO growth charts. We categorized patients as overweight (z-score >2), underweight (z-score <−2), and standard weight (−2≤ z-score ≤2) on the basis of the WHO’s child growth standards.15,16
Primary Outcomes
Primary outcomes included highest respiratory support need (room air, nasal cannula [NC], high flow nasal cannula [HFNC], noninvasive positive pressure ventilation [bilevel positive airway pressure (PAP) or continuous PAP], or mechanical ventilation), LOS, and admission or transfer to the PICU. Our institution’s HFNC protocol created a maximum support of 8 liters per minute (LPM) permitted on the acute care floor; therefore, patients requiring >8 LPM are admitted or transferred to the ICU. For patients who did not require ICU care during their stay, we recorded the highest bronchiolitis score, a modified tool based on a locally validated study that is part of our institution’s acute care bronchiolitis protocol.17 The summative score helps determine respiratory support needs and ranges from 0 to 18 on the basis of the individual scores for 6 categories: breath sounds, retractions, respiratory effort, respiratory rate, respiratory support, and aeration.
Secondary Outcomes
Secondary outcomes focused on supplemental evaluation and treatment, including placement of intravenous line or nasogastric tube, frequency of albuterol treatments, nasal and nasopharyngeal (NP) suctioning with wall-mounted suction devices, intermittent PAP therapy (where positive pressure via mask connected to wall oxygen is applied to patient while they breathe for up to 15 minutes), and chest radiographs.
Data Management and Statistics
Study data were collected and managed using a secure, Web-based software platform.18 The study was approved by the hospital’s institutional review board.
Data were summarized by median and interquartile range or n (%). χ2 for Fisher’s exact test was used to compare categorical variables and Kruskal-Wallis test was used for continuous variables. Poisson regression with a robust error variance was used to examine the effect of weight-for-length categories on respiratory support and admission to ICU, including age as a covariate. Adjusted relative risk (aRR) and 95% confidence interval (CI) were estimated. Generalized linear model was performed to evaluate the association between weight-for-length categories and highest bronchiolitis score, with age included in the model. Among the potential confounders (age, sex, race, ethnicity, birth weight, gestational age, nutrition), age was found to be significantly different among the 3 groups. Therefore, it was accounted for in the final models for the primary outcomes of respiratory support needs, ICU admission, and bronchiolitis score. P < .05 was considered significant. SAS version 9.4 (SAS Institute Inc., Cary, NC) and R were used for statistical analysis.
Results
Demographics
During the study period, there were 644 infants hospitalized for their first episode of bronchiolitis. Two hundred fifty-nine were excluded (Supplemental Table 3); the majority (166 infants) was for prematurity. Thus, 385 infants were categorized into 3 groups: overweight (n = 24), standard weight (n = 335), or underweight (n = 26). Age in days differed significantly between the groups; P = .002. There were no other differences (Table 1).
. | Overweight (n = 24) . | Standard Weight (n = 335) . | Underweight (n = 26) . | P . |
---|---|---|---|---|
Z-score | 2.4 (2.2–2.8) | 0.0 (−0.7 to 0.8) | −2.5 (−2.8 to −2.4) | — |
Weight-for-length percentile | 99.1 (98.6–99.7) | 50.7 (25.3–78.8) | 0.7 (0.3–0.9) | <.001 |
Age, d | 148.0 (86.5–247.0) | 145.0 (79.0–225.0) | 77.0 (42.0–148.0) | .002 |
Sex | .55 | |||
Female | 12 (50.0) | 141 (42.1) | 9 (34.6) | |
Male | 12 (50.0) | 194 (57.9) | 17 (65.4) | |
Race | .27 | |||
Asian | 1 (4.2) | 15 (4.5) | — | |
Black or African American | 7 (29.2) | 89 (26.6) | 8 (30.8) | |
Native Hawaiian or other Pacific Islander | — | — | 1 (4) | |
Patient refused | 2 (8.3) | 9 (2.7) | — | |
Unknown | — | 14 (4.2) | — | |
White | 14 (58.3) | 208 (62.1) | 17 (65.4) | |
Ethnicity | .89 | |||
Hispanic or Latino | 4 (16.7) | 49 (14.6) | 5 (19.2) | |
Not Hispanic or Latino | 20 (83.3) | 271 (80.9) | 21 (80.8) | |
Patient refused | — | 2 (0.6) | — | |
Unknown | — | 13 (3.9) | — | |
Birth weight, kga | 3.3 (2.7–3.7) | 3.3 (3.0–3.6) | 3.2 (3.0–3.6) | .95 |
Nutrition | .61 | |||
Breastfed | 6 (25.0) | 75 (22.4) | 10 (38.5) | |
Formula | 15 (62.5) | 225 (67.2) | 14 (53.9) | |
Combination | 3 (12.5) | 31 (9.2) | 2 (7.7) | |
Transitioned to cow’s milk | — | 4 (1.2) | — | |
Gestational age, wka | 39.0 (37.0–39.0) | 39.0 (38.0–39.0) | 38.0 (37.5–39.0) | .11 |
. | Overweight (n = 24) . | Standard Weight (n = 335) . | Underweight (n = 26) . | P . |
---|---|---|---|---|
Z-score | 2.4 (2.2–2.8) | 0.0 (−0.7 to 0.8) | −2.5 (−2.8 to −2.4) | — |
Weight-for-length percentile | 99.1 (98.6–99.7) | 50.7 (25.3–78.8) | 0.7 (0.3–0.9) | <.001 |
Age, d | 148.0 (86.5–247.0) | 145.0 (79.0–225.0) | 77.0 (42.0–148.0) | .002 |
Sex | .55 | |||
Female | 12 (50.0) | 141 (42.1) | 9 (34.6) | |
Male | 12 (50.0) | 194 (57.9) | 17 (65.4) | |
Race | .27 | |||
Asian | 1 (4.2) | 15 (4.5) | — | |
Black or African American | 7 (29.2) | 89 (26.6) | 8 (30.8) | |
Native Hawaiian or other Pacific Islander | — | — | 1 (4) | |
Patient refused | 2 (8.3) | 9 (2.7) | — | |
Unknown | — | 14 (4.2) | — | |
White | 14 (58.3) | 208 (62.1) | 17 (65.4) | |
Ethnicity | .89 | |||
Hispanic or Latino | 4 (16.7) | 49 (14.6) | 5 (19.2) | |
Not Hispanic or Latino | 20 (83.3) | 271 (80.9) | 21 (80.8) | |
Patient refused | — | 2 (0.6) | — | |
Unknown | — | 13 (3.9) | — | |
Birth weight, kga | 3.3 (2.7–3.7) | 3.3 (3.0–3.6) | 3.2 (3.0–3.6) | .95 |
Nutrition | .61 | |||
Breastfed | 6 (25.0) | 75 (22.4) | 10 (38.5) | |
Formula | 15 (62.5) | 225 (67.2) | 14 (53.9) | |
Combination | 3 (12.5) | 31 (9.2) | 2 (7.7) | |
Transitioned to cow’s milk | — | 4 (1.2) | — | |
Gestational age, wka | 39.0 (37.0–39.0) | 39.0 (38.0–39.0) | 38.0 (37.5–39.0) | .11 |
Data are n (%) or median (interquartile range).
Exact birth weight known for n = 346; exact gestational age known for n = 358 (otherwise documented as “term”).
Primary Outcomes
Highest Respiratory Support
All overweight infants required some form of respiratory support (NC, HFNC, bilevel PAP/continuous PAP, or mechanical ventilation) compared with 81.8% of standard weight and 76.9% of underweight infants; P = .03. After adjusting for age, overweight were more likely to have any respiratory support compared with standard weight (aRR, 1.22; 95% CI, 1.16–1.29; P < .001) or underweight (aRR, 1.29; 95% CI, 1.04–1.60; P = .02).
A significant association (P = .009) was also found among highest respiratory support needs (Fig 1). Overweight infants used HFNC the most (75.0%) compared with standard weight (48.1%) and underweight (42.3%); P = .03. After adjusting for age, overweight were more likely to have HFNC as their highest respiratory support than standard weight (aRR, 1.56; 95% CI, 1.21–2.02; P < .001) and underweight (aRR, 1.75; 95% CI, 1.05–2.91; P = .03).
ICU Admission and Transfers
For admission to the ICU, 54.2% of overweight infants required admission, whereas 21.5% of standard-weight infants and 34.7% of underweight infants were admitted (P < .001). There was no difference in transfers to the ICU from the acute care floor (Table 2). After adjusting for age, overweight were more likely to require ICU admission compared with standard weight (aRR, 2.52; 95% CI, 1.65–3.85; P < .001).
Primary Outcomes . | Overweight . | Standard Weight . | Underweight . | P . |
---|---|---|---|---|
Any respiratory supporta | 24 (100.0) | 274 (81.8) | 20 (76.9) | .03 |
ICU admissiona | 13 (54.2) | 72 (21.5) | 9 (34.7) | <.001 |
ICU transfers | 0 (0) | 36 (10.7) | 2 (7.7) | .29 |
Bronchiolitis scorea | 8 (5–10) | 4 (3–7) | 4 (3–7) | .01 |
LOS, d | 2.8 (1.8–3.9) | 2.0 (1.4–3.1) | 2 (1.2–4.4) | .07 |
Secondary outcomes | ||||
Placement of IV or NG | 19 (79.2) | 217 (64.8) | 19 (73.1) | .26 |
Number of NP suctioning per patient | 11.5 (7.5–16) | 7 (3–13) | 6 (3–18) | .04 |
Number of nasal suctioning per patient | 13.5 (10.5–26) | 12 (7–20) | 13 (9–16) | .18 |
Number of PAP treatments per patient | 5 (0–8.5) | 0 (0–6) | 0 (0–3) | .05 |
Number of albuterol treatments/patient | 5.5 (1–11) | 2 (1–10) | 7 (1–19) | .62 |
Number of CXR per patient | 0 (0–1) | 0 (0–1) | 0 (0–1) | .85 |
Primary Outcomes . | Overweight . | Standard Weight . | Underweight . | P . |
---|---|---|---|---|
Any respiratory supporta | 24 (100.0) | 274 (81.8) | 20 (76.9) | .03 |
ICU admissiona | 13 (54.2) | 72 (21.5) | 9 (34.7) | <.001 |
ICU transfers | 0 (0) | 36 (10.7) | 2 (7.7) | .29 |
Bronchiolitis scorea | 8 (5–10) | 4 (3–7) | 4 (3–7) | .01 |
LOS, d | 2.8 (1.8–3.9) | 2.0 (1.4–3.1) | 2 (1.2–4.4) | .07 |
Secondary outcomes | ||||
Placement of IV or NG | 19 (79.2) | 217 (64.8) | 19 (73.1) | .26 |
Number of NP suctioning per patient | 11.5 (7.5–16) | 7 (3–13) | 6 (3–18) | .04 |
Number of nasal suctioning per patient | 13.5 (10.5–26) | 12 (7–20) | 13 (9–16) | .18 |
Number of PAP treatments per patient | 5 (0–8.5) | 0 (0–6) | 0 (0–3) | .05 |
Number of albuterol treatments/patient | 5.5 (1–11) | 2 (1–10) | 7 (1–19) | .62 |
Number of CXR per patient | 0 (0–1) | 0 (0–1) | 0 (0–1) | .85 |
Data are n (%) or median (interquartile range). CXR, chest radiograph; IV, intravenous; NG, nasogastric tube.
See text for age adjustments.
Bronchiolitis Score
For patients who only required treatment on the acute care floor (n = 255), the overweight infant median bronchiolitis scores (8) were double the scores of standard weight (4) and underweight (4); P = .01 (Table 2). After adjusting for age, the expected highest bronchiolitis score in overweight was 1.51 times that of standard weight (P = .006) and 1.58 times that of underweight (P = .03).
LOS
There was no significant difference in LOS in days among the 3 groups; P = .07.
Secondary Outcomes
Significant differences were found in number of NP suction events per patient and number of PAP treatments per patient among weight-for-length categories (Table 2). No differences were seen in intravenous or nasogastric tube use, nasal suctioning, albuterol treatments, or number of radiographs.
Discussion
Overweight infants hospitalized for bronchiolitis were associated with more respiratory support and experienced more interventions, suggesting more severe disease. One notable finding was that all overweight infants required some form of support compared with standard-weight and underweight infants. Respiratory support is used for supportive treatment of bronchiolitis for respiratory distress with and without hypoxia. This may imply that overweight infants demonstrate more respiratory distress, which is also supported by the higher bronchiolitis scores found in the overweight group.
The most used form of respiratory support for overweight infants was HFNC with 75% utilization. This may be explained by the benefits of HFNC compared with standard NC. HFNC in children improves washout of NP dead-space, reduces energy expenditure, and likely provides some positive pharyngeal pressure.19 Overweight infants may require increased energy for effective ventilation and experience more frequent upper airway obstruction events,4,7 thus HFNC may help optimize these infants’ respiratory effort.
Diminished respiratory function in overweight infants may explain the increase in PAP treatments and NP suction events. PAP treatments are recommended in our protocol for increasing bronchiolitis score, diminished aeration, or increasing hypoxia, which could be expected in infants with severe bronchiolitis and worse underlying respiratory function. Moreover, NP suction is performed at our institution when patients have visible or audible upper airway congestion that is not relieved by nasal suctioning. If similar to older patients,3 overweight infants may experience increased upper airway resistance, which could predispose them to worse upper airway obstruction from congestion.
More overweight infants required an ICU stay; however, a significant limitation to this metric exists at our institution. There is a finite number of patients requiring HFNC allowed on the acute care floors, and a max setting of 8 LPM is permitted. Given standard of care of weight-based settings for HFNC, it can be expected that heavier infants require higher flow settings. As such, overweight infants may have required flows that exceeded our acute care floor limit, making it difficult to interpret the ICU data. Nonetheless, other indications of illness severity suggested overweight infants had more severe disease. Additionally, our study suggests that weight-based HFNC protocols may be useful and provide a more appropriate amount of flow.
Our study has many inherent limitations, including the retrospective study design and smaller sample size. The single-site study was influenced by our institution’s practice patterns and limits generalizability. Additionally, excluding the patients with reactive airway disease or asthma (total of 19) may have introduced a selection bias on the acuity of illness, despite our center’s practice habits of using scheduled albuterol and/or inhaled corticosteroids for patients who demonstrate a significant bronchospastic component.
Despite these limitations, the implications of these findings may elucidate better treatments and prognosis of overweight infants hospitalized with bronchiolitis, and even other respiratory diseases. Furthermore, it adds to the existing knowledge of negative health outcomes associated with overweight in infancy. Future, larger, multisite studies are needed to confirm that overweight infants experience more severe disease and to better understand the respiratory pathophysiology in this population. It would also be helpful to analyze the data with more covariates including race, ethnicity, and insurance. Additionally, prospective research comparing different treatment measures for overweight infants, such as HFNC, may help guide new patient-centered standards of care.
Acknowledgments
We thank Diana Barany and Jody Barbeau for their contributions.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLAIMER: The authors have indicated they have no conflicts of interest relevant to this article to disclose.
Dr Madion was the principal investigator, designed the study, collected the data and interpreted it, drafted the initial version of the manuscript, and revised the manuscript; Drs Bauer, Parakininkas, and McFadden, and Ms Karls assisted in study design, interpreted the data, and assisted in manuscript revisions; Dr Pan assisted in study design, performed the statistical analysis, interpreted the data, and assisted in manuscript revisions; Dr Liljestrom was a mentor throughout project, assisted in study design, interpreted the data, and assisted in manuscript revisions; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work. Dr Madion’s current affiliations is East Tennessee Children’s Hospital.
Comments