Skip to main content

Table 2 Adjusted mean differences in cortical thickness across the prefrontal cortex by HIV exposure groups

From: Association of in utero HIV exposure with child brain structure and language development: a South African birth cohort study

Prefrontal cortex regions

Cortical thickness (mm)

Minimally adjusted modela

Full adjusted modelb

HEU

Mean (SD)

(n = 70)

HU

Mean (SD)

(n = 92)

Mean difference (95% CI)

p-value

Effect size

Mean difference (95% CI)

p-value

Effect size

Superior frontal

3.31 (0.16)

3.30 (0.16)

0.01 (− 0.04 to 0.06)

0.607

0.08 (− 0.23 to 0.39)

0.03 (− 0.03 to 0.08)

0.321

0.16 (− 0.15 to 0.48)

Caudal middle frontal

2.96 (0.16)

2.93 (0.15)

0.02 (− 0.03 to 0.07)

0.351

0.15 (− 0.16 to 0.46)

0.03 (− 0.02 to 0.08)

0.310

0.17 (− 0.14 to 0.48)

Rostral middle frontal

2.97 (0.12)

2.95 (0.13)

0.02 (− 0.02 to 0.06)

0.319

0.16 (− 0.15 to 0.47)

0.03 (− 0.01 to 0.07)

0.102

0.27 (− 0.04 to 0.58)

Medial orbitofrontal

3.21 (0.21)

3.14 (0.19)

0.08 (0.01 to 0.14)

0.016*

0.39 (0.07 to 0.70)

0.09 (0.02 to 0.15)

0.009*

0.44 (0.12 to 0.75)

Lateral orbitofrontal

3.27 (0.14)

3.26 (0.15)

0.01 (− 0.04 to 0.06)

0.648

0.07 (− 0.24 to 0.39)

0.02 (− 0.02 to 0.07)

0.318

0.17 (− 0.15 to 0.48)

Inferior frontal

3.15 (0.13)

3.15 (0.13)

0.00 (− 0.04 to 0.04)

0.923

0.02 (− 0.30 to 0.33)

0.02 (− 0.03 to 0.06)

0.412

0.13 (− 0.18 to 0.45)

Frontal pole

3.54 (0.27)

3.51 (0.30)

0.04 (− 0.05 to 0.13)

0.386

0.14 (− 0.17 to 0.45)

0.06 (− 0.03 to 0.15)

0.212

0.21 (− 0.10 to 0.52)

  1. Multiple linear regression estimates for HIV exposure on cortical thickness. *p < 0.05. aAdjusted for child age and sex. bAdjusted for child age and sex, household income, maternal age, and education. Cortical thickness (mean of left and right hemispheres), mean differences (regression coefficients minimally and fully adjusted in multiple regression models), p-values, and effect sizes are presented. Effect sizes were calculated using Cohen’s d with associated 95% confidence intervals. Residuals were assessed for each model using quantile–quantile plots and scatterplots and were normally distributed. A positive mean difference estimate indicates that HIV exposure is associated with thicker cortex in that region. Post hoc, the regions with a significant difference were assessed bilaterally in fully adjusted models: medial orbitofrontal cortex (left hemisphere) effect size 0.36 (0.04 to 0.67) and (right hemisphere) effect size 0.44 (0.12 to 0.75). Abbreviations: HEU, HIV-exposed and uninfected; HU, HIV-unexposed