Tuesday, March 1, 2016

I’ll trade you my coffee table for some breast milk . . .buying milk off Craigslist

Recently I was approached by one of our local news channels about a project on online milk buying. The reporter had purchased several samples of human milk online, and I was wondering if I would be willing to be the scientific adviser for the project.

I’ll admit something – I went into this with a motive. I really felt like I was there to protect the milk for poor interpretation, panic arising at the mere presence of bacteria in the samples. Because there are bacteria in milk. These bacteria range from harmless skin bacteria to important bacteria that will colonize the infant’s gut and provide long term benefits to the infant (Marin et al., 2009; Thompson et al., 2012). 

The reporter had 3 samples, after making dozens of inquiries. Two samples were shipped, one arrived seemingly frozen and the other a leaking mess. The third sample was picked up from the mother. The leaking mess was thrown away, and the remaining two samples were tested.

Sample 1: Total bacterial count of 700,000 CFU/gram. This is 7x what is allowed by the dairy industry for pasteurization, and 70x what is allowed by the Human Milk Banking Association of America (HMABA). This included a substantial number of coliform bacteria. Now while many coliforms are not themselves harmful, they are generally used as a measure of bacterial contamination.  

Sample 2: Total bacterial count of <10 CFU/gram. No coliforms.

Compared to the literature, these numbers are not surprising. Earlier, full sized studies of human milk purchased online has found very similar results – Keim et al., (2013) reported that 74% of 102 samples they purchased online had microbial loads >24 CFU/g; comparable data for samples from milk banks are much lower. A second study, purchasing milk online in Canada, also found high levels of bacterial contamination for purchased milk (St. Onge et al., 2015). 
Figure 1: Figure 2 from Keim et al., (2013) showing the distribution of bacteria counts in milk purchased online. Sample 1 would have been placed in the 10^5 group for total bacteria. Image: doi: 10.1542/peds.2013-1687

What seems to be the biggest risk factor for microbial growth in milk? Keim et al., (2013) reported that the amount of time the milk spent in transit was the biggest predictor of microbial growth. Geraghty et al., (2013) reported shipping conditions broke down as follows: 89% of milk samples arrived at a temperature above -20C/-4F (hard frozen) and of those, 45% arrived above 4C/39.2F (or warmer than your fridge). We know milk spoils, and human milk is no exception to the Milk Spoils rule. 

Figure 2: Milk shipping is important, and just because samples look frozen does not mean they are! Photo:Lansinoh.com
What were some other factors? Keim et al (2013) noted that “Information sellers conveyed in their advertisements about their health and behaviors were poor indicators of milk quality.” Donor claims as well (sellers claiming they were milk bank donors) and large volumes were also risk factors. It should be a general red flag that someone claiming to be a milk bank donor but selling thousands of ounces of milk online has something else going on . . .either the bank turned down the milk or the person is not associated with a milk bank despite claims otherwise. 

Keim et al., (2013) also found that the age of milk – that is the interval between when it was expressed and when it was shipped – was also a predictor of bacterial load. They hypothesized that anti-microbial properties of the milk may decline with storage time – and that assumes immediate and appropriate storage. One other potential point of risk may be at the pump itself. There have been several neonatal intensive care unit based outbreaks of bacterial infections linked to improperly cleaned pumps, and one estimate suggests that nearly 1/3 of all pumps may be improperly cleaned and potentially harbor bacteria. 

Figure 3: PSA – maybe give your pump a good scrub down today, just in case. This style of pump is especially problematic because the rubber bulb is extremely hospitable to bacteria. Photo: Etsy
Overall, I was really surprised by this. I expected the Craigslist sample to be a little dodgy, but I genuinely expected that the samples would be safe. Reading the literature, seeing our results, and digging into the online ads has really changed my mind. Buying breast milk online, especially from complete strangers is a risky game. I mean, I wouldn’t buy a dozen cookies off of Craigslist, and the same goes for milk. Both the FDA and the AAP recommend against such practices – and for evidence based reasons as shown in the Pediatrics papers and others. Things in the ads that made me curious: large volumes of milk with a young baby, packages of milk that look too uniform, poor organization of the freezer.

Figure 4: Two things I would never personally buy on Craigslist. Photo: http://wholelifestylenutrition.com
What about milk sharing? As Aunchalee Palmquist has previously reported (see her blog here) there is something different about milk sharing - it falls more under the auspices of a gift economy, and most studies show that peer to peer milk sharing does not include the same risks of bacterial contamination as milk sold AND shipped. In recent work, Palmquist report that of 867 mothers who completed their survey 100% did not participate in anonymous donation but knew or screened their donors. Almost 30% of mothers only shared with families or close friends, and a similar number (27%) of mothers had serological screening for their donors. This really highlights that we are talking about two very different practices here - milk sharing and milk selling - with possible different underlying motivations and practices.


I had completely forgotten a post on milk selling over at mammalssuck -  worth checking out here!

Geraghty SR, McNamara KA, Dillon CE, Hogan JS, Kwiek JJ, Keim SA. (2013) Buying human milk via the internet: just a click away. Breastfeed Med. 2013 Dec;8(6):474-8. doi: 10.1089/bfm.2013.0048.

Keim SA, Hogan JS, McNamara KA, Gudimetla V, Dillon CE, Kwiek JJ, Geraghty SR. (2013) Microbial contamination of human milk purchased via the Internet. Pediatrics. 132(5):e1227-35. doi: 10.1542/peds.2013-1687.

Keim SA, McNamara KA, Jayadeva CM, Braun AC, Dillon CE, Geraghty SR. (2014) Breast milk sharing via the internet: the practice and health and safety considerations. Matern Child Health J. 18(6):1471-9. doi: 10.1007/s10995-013-1387-6.

Marín ML, Arroyo R, Jiménez E, Gómez A, Fernández L, Rodríguez JM. (2009) Cold storage of human milk: effect on its bacterial composition. J Pediatr Gastroenterol Nutr. 49(3):343-8. doi: 10.1097/MPG.0b013e31818cf53d.

Palmquist AE, Doehler K. 2015 Human milk sharing practices in the U.S. Matern Child Nutr.  doi: 10.1111/mcn.12221. [Epub ahead of print]

St-Onge M, Chaudhry S, Koren G. (2015) Donated breast milk stored in banks versus breast milk purchased online. Can Fam Physician. 2015 Feb;61(2):143-6.
Thompson AL. (2012) Developmental origins of obesity: early feeding environments, infant growth, and the intestinal microbiome. Am J Hum Biol. 24(3):350-60. doi: 10.1002/ajhb.22254.