Our objective was to examine the characteristics of preterm and full-term infants < or = 6 weeks old that influence the success of obtaining sufficient sweat for diagnosis of CF, and corresponding sweat chloride concentrations. A retrospective chart review of 119 sweat tests was performed on 103 preterm and full-term infants < or = 6 weeks of age. Bivariate and multivariate regression analyses were used to determine the predictors of successful sweat testing and characteristics influencing sweat chloride concentrations. Adequate amounts of sweat (> or = 75 mg) were obtained for analysis in 73.8% of initial attempts in the infant group. The following characteristics were associated with increased odds of obtaining a quantity not sufficient (QNS) for sweat chloride concentration measurement: African-American race, infant weight < 2,000 g, preterm birth, and postmenstrual age (PMA) < 36 weeks. With a multivariable logistic model, the only significant predictors were African-American race (7.3, 2.4-21.7) and PMA < 36 weeks (17.9, 4.2-75.9). Sweat chloride concentration in non-CF individuals is inversely related to both gestational age and age at testing, and this effect is additive in a linear regression model. In conclusion, sweat collection can be reliably performed in infants > or = 36 weeks postmenstrual age, > 2,000 g, and > 3 days postnatal age. Maturational factors have a mild impact on sweat chloride concentration.
Copyright 2005 Wiley-Liss, Inc.