Background:  Although some caregivers are known to pre-masticate (pre-chew) food for infants, usually during the weaning period, HIV transmission has not been linked to this practice. We describe 3 cases of HIV transmission in 2 cities in the United States over a decade, possibly related to this practice.

Methods:  A thorough investigation to rule out alternative modes of transmission was conducted by primary care staff and Health Department investigators. Blood samples from the cases and suspected sources were sent to the Centers for Disease Control and Prevention (CDC) for HIV DNA extraction, amplification, sequencing, and genetic and phylogenetic analysis using the C2V3C3 or gp41 region of env and the p17 coding region of gag. Phylogenetic comparisons of virus from cases and suspected available sources were performed.

Results:  Case 1 (Memphis) presented at age 9 months with fever, failure to thrive, jaundice, and had a positive HIV RNA polymerase chain reaction (PCR) analyses preceded by 3 negative RNA PCR prior to age 4 months. Case 1’s HIV-infected mother reported pre-masticating the child’s food starting at age 4 months. After 2 previous negative HIV serology tests at ages 21 and 22 months, Case 2 (Miami) tested positive for HIV (by enzyme-linked immunoassay [ELISA] and Western blot) at age 39 months during a work-up for recurrent submandibular lymphadenitis. Case 2’s HIV-infected mother reported offering the child pre-masticated food intermittently. Case 3 (Miami) tested positive for HIV (ELISA and Western blot) at age 15 months during a work-up for recurrent diarrhea and otitis. Case 3’s HIV-uninfected mother reported that the child’s HIV-infected aunt fed the child pre-masticated food between ages 9 and 14 months. Case 1’s mother and Case 3’s aunt were reported to have bleeding gums while pre-masticating. A thorough investigation of these cases did not reveal other modes of HIV transmission (i.e. breast feeding, percutaneous injuries, or sexual abuse). Phylogenetic analysis showed strong relatedness of the virus from the infant and mother in Cases 1 and 2 and ruled out a household contact (aunt’s partner) as a source of infection in Case 3. Case 3’s aunt died before blood could be collected for this investigation.

Conclusions:  While all 3 cases raise the index of suspicion, 2 of the cases provide compelling evidence linking pre-mastication to HIV transmission. This route of transmission of HIV has not to our knowledge been previously reported. The risk of infant HIV transmission associated with pre-mastication of food deserves further investigation and has important global implications.