‘Can a meaningful pattern be discerned in the large variations in syphilis rates over the last century?’ This is the question addressed by a recent systematic review – Kenyon & Tsoumanis (K&T) – based on published data on ante-natal syphilis prevalence (ASP) from those countries for which that data is available since at least 1951.   This cutoff reduces the number of countries that qualify for inclusion, but allows more recent trends in the late twentieth, and early twenty-first, centuries, to be set against the background of the impact of the introduction of penicillin in the 1950s. A pattern emerges from the data, which K&T then to seek to explain by investigating its association with various potential variables through multivariate analysis: per capita GDP; circumcision practice; health expenditure; efficacy of diagnosis/treatment; geographical region.

The pattern itself is: in most parts of the world, a more or less steep decline following the introduction of penicillin – ultimately, by the 1990s, to below 1%, and by the 2000s, to below (massively below, in many cases) 0.5%; in sub-Saharan Africa alone, a decline plateauing out at around 6% up until the end of the twentieth century, when there is a further decline to just above 1.5%.  A limitation of the study is its concentration on eleven countries for which ASP data is available from before the days of penicillin, with only two of those countries being in sub-Saharan Africa (South Africa and Zimbabwe).  So far as concerns more recent evidence for ASP prevalence, the kind of rates that the authors give for SA and Zimbabwe seem, broadly, to be replicated in other countries of sub-Saharan Africa (Otieno-Nyunya & Kaiser/STIs (Kenya); Makasa & Sandoy/STIs (Zambia); Kirakoya-Samadoulougou & Nagot/STIs (Burkina-Faso);  Ardu-Sarkodie & Peeling/STIs (Ghana), and their rates for the other regions to be replicated in other non-sub-Saharan African settings (Cheng & Cai/STIs (China); Galdava & Domeika/STIs (Georgia) Thirumoorthy & Lim/STIs (Singapore)).

As for the explanation of this pattern, the authors find no association on multivariate analysis with any of their potential variables, save with residence in sub-Saharan Africa.  This is itself an interesting negative finding, and prompts the authors to consider other population-level correlations also included in the evidence reviewed – notably, with prevalence of HIV and HSV-2; ‘the populations that in the 1990s had high prevalences of syphilis and HSV-2 went on to have high HIV prevalences’.  The correlation with prevalence of HSV-2 is of particular interest because it is unlikely that the prevalence of the one infection could have influenced that of the other (see also: Hochberg & Dandona/STIs).  To K&T, it suggests the likely importance of ‘more connected sexual networks’ and ‘greater partner concurrency’ in explaining traditionally – and currently – high relative ASP levels in sub-Saharan Africa.  However, they refer to studies that contest this hypothesis, and emphasize the need for further research to elucidate the factors underpinning difference in syphilis rates – especially given the possibility that the successful use of ART in those countries may be accompanied by the re-establishment of former sexual networks.