LAUSR

represents only an incremental advance over the LCA approach. It relies heavily on imputation, and itmay be difficult for other groups to replicate this analytic approach, depending on the needed sample size. The population studied does not appear to be racially diverse. A recent study using LCA showed similar wheeze patterns inWhite andAfrican American children; however, the AfricanAmerican childrenweremore likely to be in the PEWgroup (5). A study of 11,000 children showed the incidence rates of asthma amongAfricanAmerican childrenwith no family history of asthmaweremarkedly higher than those of nonHispanicWhite children during the preschool years (11). Thus, how the wheeze classification described heremay changewith amore diverse population remains to be determined. The cohorts also did not perform early-life airway function tests that could have helped identify early-life associationswith wheeze phenotypes. Wheeze phenotypes such as lung function trajectories are at least partially established by perinatal factors, allowing the potential for primary prevention with early-life interventions (12). More precise clustering of wheeze phenotypes as described here has the potential to facilitate these strategies. This paper once again demonstrates the importance of longitudinal, diverse birth cohorts to identify the possible genetic, prenatal, and environmental factors associated with different wheeze phenotypes and their association with lung function trajectories, lung disease, and prevention strategies.