Cognitive function following diabetic ketoacidosis in young children with type 1 diabetes

Abstract Introduction Young children with type 1 diabetes (T1D) may be at particularly high risk of cognitive decline following diabetic ketoacidosis (DKA). However, studies of cognitive functioning in T1D typically examine school‐age children. The goal of this study was to examine whether a single experience of DKA is associated with lower cognitive functioning in young children. We found that recently diagnosed 3‐ to 5‐year‐olds who experienced one DKA episode, regardless of its severity, exhibited lower IQ scores than those with no DKA exposure. Methods We prospectively enrolled 46 3‐ to 5‐year‐old children, who presented with DKA at the onset of T1D, in a randomized multi‐site clinical trial evaluating intravenous fluid protocols for DKA treatment. DKA was moderate/severe in 22 children and mild in 24 children. Neurocognitive function was assessed once 2–6 months after the DKA episode. A comparison group of 27 children with T1D, but no DKA exposure, was also assessed. Patient groups were matched for age and T1D duration at the time of neurocognitive testing. Results Children who experienced DKA, regardless of its severity, exhibited significantly lower IQ scores than children who did not experience DKA, F(2, 70) = 6.26, p = .003, partial η 2 = .15. This effect persisted after accounting for socioeconomic status and ethnicity. Conclusions A single DKA episode is associated with lower IQ scores soon after exposure to DKA in young children.


| INTRODUC TI ON
Diabetic ketoacidosis (DKA), a common complication of type 1 diabetes (T1D), 1 is associated with cognitive declines, 2-6 even when no obvious neurological symptoms occur during DKA. Young children may be more vulnerable and at greater risk of developing these declines. 7 Previous paediatric studies examining associations between DKA and cognition have focused on broad age ranges, extending beyond early childhood. [2][3][4][5][6] In a small-scale study, DKA was associated with larger memory declines if it occurred before age 7 years. 2 However, children were assessed between 7 and 16 years of age, making it impossible to determine whether declines manifested soon after DKA exposure or emerged later and worsened over time. 7 In a recent study from our group, lower pH at diagnosis of DKA, which was experienced at diabetes onset, was associated with lower intellectual quotient (IQ) 6 in a sample of newly diagnosed 6-to 18-year-old children, suggesting small effects soon after the onset of T1D. However, there were no significant group differences in IQ scores between children who did and did not experience DKA; group differences were restricted to long-term memory. In the current study, we asked whether group differences in IQ scores would be evident in young children tested 2-6 months after DKA at onset of T1D. Such a finding would suggest greater susceptibility to cognitive decline in young children.

| ME THODS
We recruited 3-to 5-year-old children who were newly diagnosed with T1D and exhibited DKA at onset from 12 centres participating in the Pediatric Emergency Care Applied Research Network (PECARN) Fluid Therapies Under Investigation in DKA (FLUID) trial. 8 All of these children were randomized in the trial. We distinguished moderate/severe DKA (pH ≤ 7.19 or serum bicarbonate concentration ≤9 mmol/L) and mild DKA (pH between 7.20 and 7.25, or serum bicarbonate concentration between 10 and 15 mmol/L). 6 Children with recent T1D diagnosis (<2 years), but no DKA episodes based on guardian history and medical review, were recruited from the paediatric diabetes clinics at the participating PECARN centres.
Neurocognitive assessment sessions were delayed or rescheduled if children had either hypoglycaemia (glucose <70 mg/dL) or hyperglycaemia (glucose >350 mg/dL). All participants were English-speaking.
We administered the English version of the Wechsler Preschool and Primary Scale of Intelligence (WPPSI-III), 9 which is validated for children ranging in age from 2 years and 6 months to 7 years and 3 months of age (normed IQ M = 100; SD = 15). Research coordinators trained by a study investigator with a doctoral degree in psychology administered the WPPSI. This measure was administered once between 2-to 6-months after DKA for children with

Abstract
Introduction: Young children with type 1 diabetes (T1D) may be at particularly high risk of cognitive decline following diabetic ketoacidosis (DKA). However, studies of cognitive functioning in T1D typically examine school-age children. The goal of this study was to examine whether a single experience of DKA is associated with lower cognitive functioning in young children. We found that recently diagnosed 3-to 5-year-olds who experienced one DKA episode, regardless of its severity, exhibited lower IQ scores than those with no DKA exposure. Conclusions: A single DKA episode is associated with lower IQ scores soon after exposure to DKA in young children.

K E Y W O R D S
cognitive function, diabetic ketoacidosis, early childhood, intelligence, type 1 diabetes exposure to DKA or soon after recruitment for children with T1D without exposure to DKA. To limit the duration of the assessment and the child's fatigue, we only administered the core subtests of the WPPSI-III corresponding to the minimal number of subsets that yielded a reliable and valid IQ score based on the test manual. 9 The ability of the WPPSI to yield reliable and valid IQ scores from a subset of core subtests has been demonstrated across multiple versions. 9

| Statistical analyses
We tested for differences in frequency distribution of demographics as a function of DKA status (severe/moderate DKA vs mild DKA, vs. no DKA) using chi-squared tests. We tested for differences in IQ scores as a function of DKA status using analysis of variance tests.
We subsequently included socioeconomic status (SES), and ethnicity as covariates and conducted an analysis of covariance. Age was not included in the models because IQ scores are age-normed. We performed the analyses using IBM SPSS Statistics (Version 27).

| DISCUSS ION
Diabetic ketoacidosis has been associated with cognitive decline in children with T1D. [2][3][4][5] Our multi-centre study provides new evidence of a strong and sizeable association between DKA status and IQ score in 3-to 5-year-old children, just a few months after the onset of T1D. Our recruitment and assessment procedures were analogous to those used in a large study with 6-to TA B L E 1 Participant characteristics as a function of DKA status.   N (%)), and analysis of variance F-tests for continuous characteristics (those with mean (SD)).; **p-Values were not calculated due to design-imposed relationships with DKA Status.
18-year-old children, 6 in which no significant group differences in IQ scores were found between children with or without DKA exposure at diabetes onset, whose neurocgnitive function was also assessed 2-to 6-months after T1D diagnosis. Thus, following DKA, IQ declines may occur more quickly and/or be more robust in young children.
We recognize several limitations in the current study. First, the sample size is relatively small, although it is, to our knowledge, the largest study of cognitive functioning in very young children with T1D. Therefore, we were unable to measure the relative contribution of additional clinically relevant factors associated with cognitive functioning above and beyond DKA, such as acute kidney injury, which is a frequent complication of DKA. 14 Second, we cannot exclude the possibility that the group differences in IQ scores reported here may depend, at least in part, on additional factors affecting the probability that children become exposed to DKA in the first place.
For example, lower parental education may limit the ability to recog-

CO N FLI C T O F I NTER E S T S TATEM ENT
None of the authors have any financial arrangements that represent conflicts of interest related to the study.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.