Abstract Body

Background:

Isoniazid preventive therapy (IPT) decreases risk of tuberculosis (TB) disease. It is challenging to evaluate the growth impact of infant IPT in observational studies because of confounding by indication – children who received IPT may differ from those who did not. In a recent randomized trial (RCT) we assessed IPT effects on infant growth without a known TB exposure.

Methods:

The infant TB Infection Prevention Study (iTIPS) trial was a non-blinded RCT among HIV-exposed uninfected (HEU) infants in Kenya. Inclusion criteria for the parent RCT were age 6-10 weeks, birthweight >2.5 kg, and gestation >37 weeks. Infants in the IPT arm received 10 mg/kg isoniazid daily for 12 months while the control arm received no intervention; post-trial observational follow-up continued through 24 months of age. We used intent-to-treat linear mixed-effects models to compare growth rates (weight-for-age z-score [WAZ] and height-for-age z-score [HAZ]) between trial arms.

Results:

Among 298 infants, 150 were randomized to IPT, 47.6% were females, median birthweight was 3.4 Kg (inter-quartile range [IQR] 3.0-3.7), and 98.3% were breastfed. During 12-month intervention period and 12-month post-RCT follow-up, WAZ and HAZ declined significantly in all children with more HAZ decline in male infants. There were no growth differences between trial arms, including in sex-stratified analyses. In longitudinal linear analysis, mean WAZ (β=0.04 [95% CI:-0.14, 0.22]), HAZ (β=0.14 [95% CI:-0.06, 0.34]), and WHZ [β=-0.07 [95% CI:-0.26, 0.11]) z-scores were similar between arms as were WAZ and HAZ growth trajectories. Infants in the IPT arm had higher monthly WHZ increase (β to 24 months 0.02 [95% CI:0.01, 0.04]) than the no-IPT arm.

Conclusions:

IPT administered to HEU infants without known TB exposure did not significantly impact growth outcomes in the first two years of life.