The epidemiology of the Streptococcus agalactiae bacterium (group B Streptococcus [GBS]) in Southeast Asia differs from that traditionally seen in the literature. Investigations after the foodborne outbreak of GBS sepsis in 2015 in Singapore concluded that GBS sepsis is primarily a foodborne infection of adults in parts of Southeast Asia (1).

The Singapore outbreak was associated with consumption of raw freshwater fish (2,3). The GBS were serotype III sequence type 283 (ST283) and behaved more aggressively in adults than other GBS; adults without comorbidities made up 22% of the ST283 bacteremia cases but only 2% of non-ST283 GBS cases (4). Subsequent studies showed that although ST283 is almost absent from the rest of the world, ST283 disease is widespread around Southeast Asia in humans and tilapia; human data from Laos and Thailand showed ST283 accounted for 76% (Laos) and 73% (Thailand) of invasive GBS collected during 2000–2017 (1,5). Aquaculture data from Malaysia, Thailand, and Vietnam show that ST283 accounted for 12%–100% of all GBS isolated from streptococcosis in farmed tilapia during 2003–2018 (1,6). Whole-genome analysis shows ST283 from humans and tilapia are 1 clone (1,7). Because consumption of raw freshwater fish is common in affected Southeast Asia countries (T. Barkham, unpub. data), raw tilapia could plausibly be the main source of human ST283, although this theory has not been studied.

This fishborne epidemiology is reflected in a report of 2 sisters in their 50s without comorbidities who returned to Laos after visiting friends and relatives in the United States. Both sisters became ill with ST283 bacteremia a day after a meal that included Mekong fish salad, traditionally made with raw freshwater fish (8). However, in case–control studies in Singapore in 2015, which reported statistically strong associations between eating raw freshwater fish and ST283 bacteremia, 21/40 participants in a retrospective study (3) and 2/9 participants in a prospective study (2) denied eating raw freshwater fish. This finding could be because of poor recollection or varying definitions of raw, as in the case of dishes prepared without heat but regarded as being no longer raw (e.g., raw fish fileted or ground and mixed with lemon juice or other sauces). Nevertheless, the possibility of interhuman transmission or that other foods are a source also deserves study. In addition, when authorities investigated a surge of 18 ST283 bacteremia cases in Singapore in July 2020, no affected persons admitted eating raw freshwater fish. Selling raw freshwater fish as a ready-to-eat food was illegal at that time in Singapore, so this denial was not surprising, but it might also suggest an alternative source or vehicle.

Studies of vaginal and rectal carriage in women in Southeast Asia have not reported sequence types, because their focus has been capsular serotypes for use in developing vaccines. Two small sequencing studies failed to find ST283; 1 looked at stool samples from 82 food handlers and fishmongers in Singapore during the 2015 outbreak (4), and the other looked at an opportunistic collection of 38 vaginal GBS samples isolated from women with colpitis in Hanoi, Vietnam, in 2016 (1).

Clinical evidence for carriage and human-to-human transmission of this bacterial strain is provided by several cases of ST283 neonatal early onset disease in Hong Kong (5) and Laos (1); at least 2 from Laos were isolated on the infants’ first day of life, which suggests maternal carriage. We sought to verify carriage. Because ST283 is known to account for 73% of invasive GBS in humans in parts of Thailand (1), we assessed ST283 carriage in a population in Thailand that consumes raw freshwater fish, as verified by the widespread occurrence of Opisthorchis viverrini, a liver fluke infection acquired by eating raw freshwater fish.