There is a Facebook/WhatsApp message that has gone viral since yesterday, reporting on a patient at the National University Hospital with Group B Streptococcus septic arthritis and linking this with a rise in other such cases in Singapore, as well as consumption of raw fish from various porridge stalls in hawker centres. The message has even made it to Reddit, which has been in the news recently for other reasons.

This raw fish is not the same as the Chinese New Year “yu sheng” in Singapore, which is basically eaten only during the festive season. Rather, as the popular doctor-food blogger ieatishootipost explained, it is a traditional side dish/comfort food that has been around for a long time. Initially, the fish used for this dish is the seawater wolf herring, or ikan parang as it is known locally. Recently (and I don’t know when – though it may be important to know this), there has been a switch to the farmed freshwater bighead carp (or “song fish” colloquially) – the head of this fish (usually steamed) is also a popular dish in local hawker centres. Probably hundreds of people eat this raw fish on a daily basis island-wide.

Back to Group B Streptococcus (GBS). Since the early part of this year, infectious diseases physicians and other clinicians in various local hospitals (both public and private) have noticed cases of severe invasive GBS infections in non-pregnant, healthy young-to-middle-aged adults. These include infections where the bacteria had infected the joints (septic arthritis), blood (bacteraemia), heart valves (endocarditis), and brain (meningitis). As mentioned in an earlier post, GBS tends to cause such invasive disease only in neonates, the elderly, and people who are otherwise immunosuppressed (i.e. patients with diseases or who are taking drugs that compromise their immune systems, such as uncontrolled diabetes mellitus, cancer, taking steroids, etc.). Invasive disease can happen in healthy young-middle-aged adults, but it is rare, as one can also see from this CDC information page.

Just to be clear, we are not talking about dozens of severe GBS infections in healthy adults in Singapore every day. The “outbreak” would have been noticed far sooner in that case. Probably just a handful of cases every 1-2 weeks, at the most, if discussions among colleagues are accurate. To put this in context, there are probably more than 3,000 cases of GBS – both colonisation (mostly women during pregnancy screening) and infection (both mild and severe) – in Singapore each year, and it has not been straightforward to document a rise in GBS cases based on laboratory data alone. What is unusual is the patient profile – healthy individuals rather than older and/or more immunosuppressed individuals.

What about the fish? It turned out that on direct questioning, several of the later patients (the earlier ones were not asked – no one tried to make a link then) were found to have eaten raw fish from a porridge stall – not just one single stall, but quite a number of porridge stalls were mentioned – within two weeks before the onset of their illness. This link is biologically plausible, as can be seen from the previous post again. A history of fish ingestion was shown to be associated with increased GBS acquisition in a U.S. university dorm, and the GBS clone causing invasive disease in non-pregnant adults in Hong Kong in the 1990’s to 2000’s had a molecular signature that is identical to that found in Asian fish. But this is by no means proof, and indeed there are several problems with this hypothesis that have to be overcome:

• GBS has not been found or reported in bighead carp previously.
• GBS also causes invasive, fatal disease in fish – in fact, it is usually detected only as a result of an investigation into unexplained fish dying off. It has been difficult to find reports where GBS only colonises fish. Fish with severe infections would probably not have been sold in Singapore hawker stalls.
• People have been eating raw fish in Singapore – prepared virtually the same way – for decades, and have had few issues (certainly none observed with GBS until now).
• Far more people eat raw fish locally and do not come down with invasive GBS disease (the attack rate would be very, very low if the association is indeed true).

It is important not to jump the gun, and the issue is also complex because a significant portion of porridge sellers’ income probably comes from selling raw fish. Livelihoods may be affected. And of course, correlation does not imply causation.

Can this be a false association? Very possible indeed. Because the questioning of the patients was not systematic, it is possible that the high proportion of cases with a history of raw fish consumption could be a statistical anomaly. Or else there were other associations with food or lifestyle that were not uncovered, simply because the right questions were not asked at that time, and some of these other associations are possibly also associated with consumption of raw fish. There is a famous example often used for teaching bias – a New England Journal publication in 1981 describing a link between coffee drinking and pancreatic cancer. But this association could not be found in repeat studies, and the likely explanation were that the controls (patients without cancers) were recruited by the same doctors that treated patients with pancreatic cancers, and many had been advised not to drink coffee.

It is important to let the relevant authorities – MOH, NEA and AVA – have some space and time to conduct proper and rigorous investigations in order to determine the extent and source of this rise in GBS invasive disease among healthy non-pregnant adults in Singapore. There is considerable room for research as well, such as whether we are dealing with a new, virulent clone of GBS; and better epidemiological and clinical profiling of GBS in Singapore.