Three Cases of HIV Transmission to Infants Through Food Pre-Chewed by HIV-Positive Caregivers

One of the oddest presentations at CROI 2008 involved three case reports of HIV transmission to babies when their HIV-infected caregivers pre-chewed the babies' food. I sat down with Dr. Kenneth Dominguez, an epidemiologist at the CDC [U.S. Centers for Disease Control and Prevention], and the two physicians who made the report: Dr. Aditya Gaur, of St. Jude Children's Research Hospital in Memphis, Tenn. and Dr. Charles Mitchell, of the Leonard M. Miller School of Medicine at the University of Miami.

Ken Dominguez, M.D.
Ken Dominguez, M.D.

Bonnie Goldman: Could you summarize your study?

Kenneth Dominguez: Sure. Essentially, what we've uncovered is that there were three cases of children that we believe were HIV infected through the process of eating pre-masticated food from their caregiver.1

Bonnie Goldman: Could you describe what that means?

Kenneth Dominguez: Sure. Essentially it means pre-chewed food. What happens is caregivers sometimes pre-chew food for their infants, for whatever reason, just to make it more palatable for the children.

Aditya Gaur et al. CROI 2008; abstract 613b. Reprinted with permission.
Click to enlarge

Bonnie Goldman: How is it delivered? Is it put into a bowl and then spoon fed? Or is it kissed? I was reading that it could be done that way as well.

Kenneth Dominguez: I think at this point we don't have a lot of information about the actual procedure. That's something we need to investigate further. That's something the CDC is very interested in looking into.

Bonnie Goldman: It turns out this has been mentioned for H-pylori [Helicobacter pylori] risk as well?2

Kenneth Dominguez: Yes, there have been a number of other organisms that have been implicated to be spread through pre-chewing the food -- a lot of case studies. But in terms of the work that CDC's been working on is these three cases of HIV infection. That's what we're really focusing on in terms of our prevention message.

Bonnie Goldman: Can you discuss the three case studies?

Kenneth Dominguez: The three case studies: Two of the cases occurred in the 1990s in Miami at Dr. Mitchell's institution at the University of Miami. One of the cases was born to an HIV-infected mother. The child was born HIV uninfected -- when you do the testing early on. Then the child later seroconverted. So the question comes up: Why does the child then seroconvert? Why did the child become HIV infected after they were born uninfected?

After doing a very thorough investigation, asking about all the possible routes of transmission -- asking about sexual abuse, needle sticks, blood transfusions -- all those other things were negative.

What they did find out was that the caregiver had pre-masticated food for the child. That was really the only risk factor they could come up with.

The second case in Miami, which also occurred in the 90s, was an HIV-uninfected child born to a mother who was HIV uninfected. The mother was uninfected to begin with. This child was fed pre-masticated food by the great-aunt, who was an HIV-infected relative. That was unbeknownst to the mother. Occasionally the child would go over to spend time with the great-aunt, and she was pre-masticating food for the child.

Charles Mitchell, M.D.
Charles Mitchell, M.D.

Charles Mitchell: The child actually lived with the great-aunt between the ages of nine to 14 months with the mother.

Bonnie Goldman: What's the duration of the feeding? Is it a long duration in general?

Kenneth Dominguez: In terms of the exposure that we've seen?

Bonnie Goldman: Yes.

Kenneth Dominguez: In two of the cases we know that there was a long duration. In one of them we don't really have a time period. She just said that she pre-masticated but she didn't remember the exact time period that it occurred. But we know that in two of the cases there were factors that would increase likelihood of transmission, and that would be blood in the mouth of the mother.

Aditya Gaur et al. CROI 2008; abstract 613b. Reprinted with permission.
Click to enlarge

Two of the caregivers -- there was a report that there was poor oral hygiene and bleeding gums. There was likely to have been blood mixed with the saliva, which provides biologic plausibility for transmission through feeding the kids through this method. In the third case, we didn't have information about the oral hygiene of the mother.

That's really important information. One of the things we want to emphasize is that it's not necessarily just the act of pre-masticating, but very likely it's the blood in the saliva that then gets into the child through -- if the child is teething, the oral mucosa is somehow compromised. Either they're teething, or they'll have some sort of other inflammation in the mouth and that allows the virus to get into the child.

Bonnie Goldman: Isn't that a rather rare occurrence, that someone would have bleeding gums and the child is teething, and there's enough of the blood to get into the food that's being pre-masticated?

Kenneth Dominguez: We do think it is a rare situation. The fact is that we've had three cases in a decade that have been reported. But they also would be very difficult cases to think about because people don't usually ask about this in the first place. But the fact that it did occur -- I think we have something to learn from these three cases.

They're trying to tell us something: Please look into this further. Describe really what's going on. How prevalent is this behavior in the general population? Also in international settings, where oral hygiene may not even be as good as here in the United States, and there's a higher level of HIV infection. That coupled with poor oral hygiene and a lot higher HIV rates could mean this could be an even more important issue in the developing world.

Bonnie Goldman: Could you tell me a little about the ethnicity of the patients?

Charles Mitchell: The two in Miami were both black.

Bonnie Goldman: Are they from particular countries in Africa? Are they recent immigrants?

Charles Mitchell: One family was African American, the other family was African Caribbean.

Bonnie Goldman: Were they born in the United States?

Charles Mitchell: I don't know about that in regards to the mother of the second case, the African Caribbean. But as far as I know she had lived in this country for a period of years. The other mother and the child were both born in the United States in Miami.

Bonnie Goldman: It's surprising that it wasn't a group of recent immigrants, because it seems this practice of pre-mastication is very prevalent in Asia, in Africa -- in West Africa.

Charles Mitchell: That may be a misimpression. Do you want to elaborate on that?

Kenneth Dominguez: I think when you talk to people anecdotally, a lot of the clinicians really haven't heard too much about this practice. You look in the literature, you see these case reports of pre-chewing food being associated with certain pathogens. There really isn't a lot of information about this particular behavior. Aditya uncovered an article you may want to mention.

Aditya Gaur, M.D.
Aditya Gaur, M.D.

Aditya Gaur: Sure. In reviewing the literature that has been found, [there is] a publication in the mid-1980s from the midwest United States where they asked 60-plus women, the majority African American, about the practice of pre-mastication. It was significant. Our report references the exact number -- I don't know if you remember -- but it was more than 60 percent that had either pre-masticated or they knew about the practice of pre-mastication.

Then about eight years later, there was a report from Texas which talked about case studies of patients where they saw this practice and related it to some other infectious disease pathogen.

I think those, plus ours, and then there's an FDA [U.S. Food and Drug Administration]/CDC survey under analysis right now, indicate that it's not a non-existent practice in the United States, despite availability of baby foods and blenders. That's one of our main messages: to bring this to the attention of providers in the United States so they can ask the question. Because sometimes what you don't ask, you don't know, and patients may just assume this is normal.

Charles Mitchell: Since these cases have come up, actually, in our screening clinic -- which is the clinic in which babies born to HIV-infected mothers are being tested to see if they themselves [are infected] -- our staff are finding cases where the mother will admit to either knowing about the practice or will admit that they have done the practice, and it's not just limited to African Americans.

Kenneth Dominguez: That's one of the most important messages: We want clients to know that this practice may have additional risk of transmission, particularly if you're HIV-infected, you really should not be pre-masticating for your child, because of these three cases that we've uncovered. I think it raises the index of suspicion, that this is something that could be happening out there and we want folks to be aware of this.

Bonnie Goldman: It's interesting that the cases were very far apart, and that just right now you're going to present these data at this conference. There were no dates in the abstract, so I thought this had just happened. Could you talk about that?

Kenneth Dominguez: In terms of the timing of the cases? The first two cases occurred in the mid-1990s. One of the reasons we've waited this long to report those cases is that -- one other thing I didn't mention earlier was that we're able to do something called phylogenetic studies, where we do a DNA fingerprinting to see whether the viruses that are in the caregiver, or the person who transmitted to the child, and the child are the same virus. In one of the two Miami cases there was a match. In the second case, the great-aunt had actually passed away before they could get blood samples from her.

Charles Mitchell: Five minutes before the child was brought to our attention.

Aditya Gaur et al. CROI 2008; abstract 613b. Reprinted with permission.
Click to enlarge

Dr. Dominguez: In that case, there was no way that we could actually get a direct link there. But we did look at the great-aunt's sexual partner, the male sexual partner, and there was not a genetic match with the sexual partner. At that point, we didn't feel comfortable making a big announcement at the time since we only had one match with one of the cases.

When Dr. Gaur reported this third case in 2004, we then did a very thorough investigation and we also contacted the Division of Nutrition at CDC, who then began conducting this national survey -- they were in the process of doing this infant feeding practices survey, and they added a question about pre-mastication. That went on until just last year. That was completed and we started getting the preliminary results back. We started thinking: It's happening here in the U.S. This is something that's real; it's not a figment of our imaginations. We felt it was something that needed to get out to the public.

Bonnie Goldman: What is the CDC plan? I understand the CDC is going to begin a research agenda?

Kenneth Dominguez: Yes. We're in the process of thinking both on the domestic front and on the international front. What are the next steps in terms of trying to describe the practice of pre-mastication in general. I think we need to know more about who's doing it, why people are doing it, how often they're doing it. What are some of the other conditions that are associated at the time they're pre-masticating? How many of them actually have bleeding gums? What is their knowledge of the practice? Why are they doing it? All those sorts of questions we have.

Also, in an international setting, it's even that much more important, because there are so many more infected women out there in the international setting, developing countries. We think the conditions are ripe for transmission if there's poor oral hygiene, and there's a lot of HIV-infected women, and there aren't a whole lot of feeding alternatives in terms of canned baby foods and blenders. The likelihood is that a lot of them are practicing this sort of practice.

Bonnie Goldman: It strikes me, though, that the likelihood that they're practicing this and we haven't seen any cases or -- those countries are well surveyed in terms of how people are getting HIV ...

Kenneth Dominguez: We think that, probably, the fact that, generally, if you're born to an HIV-infected mother, people will make the assumption later on if they seroconvert that, oh, it was perinatal transmission, it just didn't, for some reason, show up initially.

Bonnie Goldman: Since they're mostly breastfeeding --

Kenneth Dominguez: Right. Breastfeeding really clouds the picture. In that sort of picture it's really hard to tease out what it was. And if you're not even thinking about [pre-mastication] in the first place, why would they even consider that?

Charles Mitchell: We've had, I wouldn't say a large number, but over the years in Miami as we continue to follow children born to mothers with HIV, we've had a smattering of cases where the child will be tested very early -- similar to Dr. Gaur's finding at St. Jude's -- where the initial testing was negative and then at approximately four or six or seven months of age, will turn out to be positive. We've just attributed it in the past to just missing -- with the viral load being so low initially that it was not detectable. Now I have to wonder if this has a role in that type of occurrence.

Bonnie Goldman: Is it just two hospitals that are aware of this? Or have you had meetings with other people in New York or other states?

Charles Mitchell: Not yet.

Bonnie Goldman: So this is the initiation of bringing this to people's attention and making it part of the counseling for women after they have a baby? Because now they're just warned not to breastfeed. It's made kind of simple. But this is a new idea.

Kenneth Dominguez: It's just another little piece of the pie that's likely contributing somehow and I think people need to be made aware of that.

Bonnie Goldman: Are these children who were infected through the pre-chewing of food still alive?

Charles Mitchell: Of the two in Miami, one died in 1996, and that was right before the availability of HAART therapy. The other one is now -- he is 16 and he is under care and when he is compliant with his medication he does fine.

[All laugh.]

Bonnie Goldman: And your patient, Dr. Gaur?

Aditya Gaur: Our patient is doing very well. She is engaged in care with us.

Bonnie Goldman: How old is she now?

Aditya Gaur: Four years.

Bonnie Goldman: She was born in 2004?

Aditya Gaur: I think one of the things we'd like to share with readers is just the part about investigating more about this practice. If clinicians note similar instances like what Dr. Mitchell mentioned, where the infection is not easily explainable, then they may want to use the health department resources or contacting the CDC to bring to attention other suspicious cases which could help do a more thorough investigation on a larger number of cases.

Kenneth Dominguez: That's usually how these cases come to our attention. They'll contact the local state health department and say, "We need some help in trying to figure [this] out. Could you please send someone to come in and help interview the client and ask about all these other potential modes of transmission and do a very thorough job of trying to rule those other things out, before you can say it's --

Bonnie Goldman: Is that how you became involved with this?

Kenneth Dominguez: Yes, that's how we became involved. The division of surveillance has a group that looks at children with non-identified risk. The lab also helped out in terms of doing the DNA analysis.

Bonnie Goldman: Do people know where to contact you at the CDC? Is it a particular department?

Kenneth Dominguez: Right now we haven't released a particular number for that. We're planning to publish an article. It's going through clearance right now. Hopefully as a part of that we'll be able to give people a number to contact. We're still getting clearance on that.

Bonnie Goldman: About this specific phenomenon?

Kenneth Dominguez: Yes.


  1. Gaur A, Dominguez K, Kalish M, Rivera-Hernandez D, Donohoe M, Mitchell C. Practice of offering a child pre-masticated food: An unrecognized possible risk factor for HIV transmission. In: Program and abstracts of the 15th Conference on Retroviruses and Opportunistic Infections; February 3-6, 2008; Boston, Mass. Abstract 613b.
    View poster:
  2. Mégraud F. Transmission of Helicobacter pylori: Faecal-oral versus oral-oral route. Aliment Pharmacol Ther. 1995;9(Suppl 2):85-91.

Download the audio file