The Effects of Anesthesia on Cognition and Brain Structure Following Early Childhood Surgery

A Review of:

Backeljauw, B., Holland, S. K., Altaye, M., & Loepke, a. W. (2015). Cognition and Brain Structure Following Early Childhood Surgery With Anesthesia. Pediatrics, 136(1). doi:10.1542/peds.2014-3526

by Chase Grosse


Purpose of the research:

Research in a variety of animal models have demonstrated that exposure to anesthetic used in human surgeries result in widespread neuron death, alteration in dendritic structure and long-term learning and memory impairments for the animal (Backeljauw, Holland, Altaye, & Loepke, 2015). Due to these findings in animal models it is plausible that exposure to anesthetic in childhood, a sensitive period for neurodevelopment, may affect human in a similar manner. Thus, childhood exposure to anesthesia may result in lifelong deficits of memory and cognition.

Previous works to show the connection between early childhood exposure to anesthesia, and cognition and structural changes have resulted in inconclusive findings. Several studies have found an association with learning, language, or behavioral deficits while others found no results (Backeljauw et al., 2015). The authors found no current data regarding the structural effects of early childhood anesthetic exposure. Yet the available data on animal models have shown neurostructural, neurobehavioral, and cognitive changes. The authors believe the discrepancy between animal models and within human studies are due to the way outcomes were measured in previous studies, e.g., diverse assessment measures including group testing, individually administered tests, and school records. Additionally, previous studies did not adjust for a variety of confounds, e.g., cognitive performance, gender, socioeconomic status (SES), anesthetic exposure duration, and handedness. Thus, the current study set out to address issues with previous research to determine whether surgical anesthesia used in otherwise healthy children before the age of four had an effect on cognition, language, and brain structure.

How they did it:

Volunteers of an existing language development study with a history of surgery with anesthesia before their fourth birthday were matched to unexposed control subjects for age, gender, handedness, and SES. Researchers reviewed and quantified anesthesia exposure and excluded inoperative events, i.e., hypotension, bradycardia, or hypoxemia. They compared data between groups from individually administered assessments for language development and intelligences between exposed and control subjects. The assessment battery included the Oral and Written Language Scales (OWLS) and the Wechsler Intelligence Scale for Children – Third Edition (WISC-III) or the Wechsler Adult Intelligence Scale – Third Edition (WAIS-III) as appropriate for age. MRIs were used to compare brain structure between groups. Anesthesia records were analyzed by a pediatric anesthesiologist to extract relevant data, i.e., peripheral oxygen saturation, systolic blood pressure, heart rate, and durations meeting age-appropriate Pediatric Advanced Life Support criteria for hypotension and bradycardia. Data was quantified into minimum alveolar concentration over time (MAC-hour) (Backeljauw et al., 2015).

The subjects were drawn from an existing MRI database that included 5 to 18 year old volunteers. Exclusion criteria involved histories of neurological or psychological illness, head trauma with loss of consciousness, previous or current use of psychostimulants medication, learning disability, birth before 38 weeks, and abnormalities observed during a clinical neurologic examination. Subjects were included in the exposure group if they had any documented surgery with anesthesia on their medical record before the age of four. Control subjects lacking evidence of anesthesia exposure in their medical records were matched to exposure subjects according to gender, handedness, age, and SES.


Group comparisons of neurocognitive assessment data were analyzed with paired t-tests followed by false discovery rate error correction for multiple comparisons. Structural effects were analyzed using voxel-based morphometry. Researchers identified regions of interest based on previous studies that suggested vulnerability in animal models, i.e., the thalamus and retrosplenial cortex (Backeljauw et al., 2015). They then examined potential interactions between neurocognitive performance and regional brain volume as a function of group, neurocognitive assessment scores and the interaction effect between the two.

What they found:

                Regardless of group, mean neurocognitive assessment scores were within population norms. However, the exposure group consistently scored lower than the control group in all subtest on both neurocognitive assessments. Significant results were found for the performance IQ on the WISC-III and WAIS-III, and the listening comprehension subtest on the OWLS. No group differences were found in brain volume for the thalamus or the retrosplenial cortex as found in animal models. The researchers then searched for and found potential structural correlates to the lower neurocognitive assessment scores. They found decreased gray matter volume in the anterior cerebellum, parts of the frontal lobe, in the lingual gyrus and the occipital lobes compared to controls. They found increased volume in regions of the rolandic operculum, right frontal lobe, and Brodmann area. It is hypothesized that the relatively lower IQ performance is associated with these volume increases. A final analysis found an interaction between lower listening scores on the OWLS and decreased volume in parts of the right lingual gyrus, the occipital lobes, temporal lobes, and the parahippocampal gyrus compared to unexposed controls. The researchers found no associations between white matter volume and neurocognitive assessment scores.

What does it mean?

                Previous studies in this area of study had inconclusive results and were plagued with confounds (Backeljauw et al., 2015). The current study sought to address these issues. Regardless of exposure, all subjects scored within population norms; however, exposed subjects consistently scored lower on individually administered neurocognitive assessments. Thus, the researchers concluded that individually administered assessments are more sensitive to detecting potential deficit and predicting neurostructural changes after anesthetic exposure. IQ in exposed subjects averaged 5-6 points lower than controls. The researchers indicate the socioeconomic potential of the finding; including the potential label a 5-6 point difference in IQ and estimated lifetime earning potential (Backeljauw et al., 2015).

The authors note that the findings cannot be causally linked due to any number of factors including postoperative issues (Backeljauw et al., 2015). Further they acknowledged although there are certainly concerns involving the use of anesthetics in surgery and their effects on a developing brain; many early life procedures are directed towards treating life threating and/or to prevent serious or life threating conditions (Backeljauw et al., 2015). Such surgeries can neither be avoided nor postponed. Thus, in such cases the benefits may well outweigh the potential risk.