Streptococcus agalactiae [Group B Streptococcus (GBS)] is a commensal of the gastrointestinal tract and vagina and has been estimated to colonize the vagina in 10–35% of pregnant women (1). GBS is a well-known agent in meningitis and sepsis in newborns, and in recent years, an increasing incidence of GBS infections among the elderly, mainly bacteremia and meningitis, has been observed in the industrialized part of the world (2–4). In newborns, GBS infections are classified as early-onset disease (EOD; age 0–6 days) or late-onset disease (LOD; age 7–90 days). EOD infections are considered to be caused by vertical transmission from the mother, whereas LOD infections are mainly due to horizontal transmission from surrounding caregivers including the mother (5, 6).

Based on capsular polysaccharide antigens, 10 different serotypes of GBS have been identified, designated as Ia, Ib, and II–IX (7). The serotype distribution of colonizing GBS varies depending on the geographical region (8). A similar variation is also observed regarding invasive GBS serotypes, although five serotypes (Ia, Ib, II, III, and V) were predominant worldwide during the period 2000–2017 (9).

None of the previous studies on the epidemiology of invasive GBS infections in Denmark (6, 10–12) provided detailed information on all age groups. In these Danish studies, serotypes Ia, Ib, II, III, and V were predominant, as also observed in the global study by Madrid et al. (9). In 2001 and 2002, data from Denmark showed the presence of invasive serotype VIII isolates, which until then were observed virtually only in Japan (12). Since then, all 10 GBS serotypes have been identified in invasive infections in Denmark, although with greatly varying prevalence (13).

The aim of this study was to present the serotype distribution and the antibiotic susceptibility of submitted invasive GBS isolates from all age groups including EOD and LOD during the period 2005–2018 in Denmark.

During the study period, isolates from 1,875 unique cases of invasive GBS infection were received, 59 of which were from patients with meningitis (3.1%; 95% CI, 2.44–4.04).

The number of meningitis cases in each age group was 11 (EOD), 29 (LOD), 1 (91 days −4 years), 1 (5–19 years), 10 (20–64 years), 6 (65–74 years), and 1 (75+ years). Because of the very low number, no specific statistical evaluation was conducted for the meningitis group (Table 1, Supplementary Table 1).

Streptococcus agalactiae is a great problem for newborns, causing EOD and LOD (2, 3). In general, two different programs for reducing the incidence of EOD have been introduced for pregnant women, the risk-based approach and the culture-based screening approach (1). Whereas, the culture-based screening has been used in the USA, the risk-based approach has been used throughout the period of this study in Denmark and is still in use (1, 6). After the introduction of programs for GBS prevention in pregnant women in some countries, a pronounced reduction has been observed, in particular regarding EOD (2, 17). However, the incidence of EOD is still high in developing countries (18).

During 1984–2002, the EOD incidence (per 1,000 live births) in Denmark ranged from 0.1 to 0.6 and the LOD incidence (per 1,000 live births) ranged from 0.0 to 0.2 (6). Ballard et al. presented data from parts of Denmark for the period 2000–2010, showing an EOD incidence of 0.18 (both Copenhagen City, Copenhagen County, and Northern Denmark) per 1,000 live births and an LOD incidence at 0.07 (Copenhagen City, Copenhagen County) and at 0.18 (Northern Denmark) per 1,000 live births (2). In the present study, the EOD and LOD incidences were nearly identical and without significant changes over the years (Figure 1, Table 2). Similar EOD and LOD incidences have also been observed in other developed countries, such as Australia, Sweden, Finland, United Kingdom, Canada, and the USA (2, 19–22). From 2005 to 2011 in Denmark, the total GBS incidence per 100,000 was around 2.0 (11). Incidence data from parts of Denmark (Copenhagen City, Copenhagen County, and Northern Denmark) in the period from 2000 to 2010 were in the range of 1.9–2.4 per 100,000 (2). In the present study, we found a similar total incidence from 2005 to 2018 (2.39 per 100,000), whereas there was a significantly increasing trend from to 1.57 per 100,000 in 2005 to 3.48 in 2018 (Figure 2, Table 2). The total incidence found in this study is comparable with the incidence detected in other developed countries, such as Iceland, Canada, Sweden, Finland, and the United Kingdom (2, 3, 19). In addition, the increase in the total GBS incidence and particular among the elderly observed in this study is in line with findings in Iceland, Finland, Norway, England and Wales, Canada, and other countries (2, 3, 19, 21, 22).

The observed increase in total GBS incidence in different countries has generally been explained by an increase of GBS infections in the elderly (2, 3, 19, 23). It has been suggested that the increase among the elderly in the developed countries can be due to an aging population with increasing prevalence of comorbidities (2, 22). This seems also to be the situation in Denmark where an increase in the general population of both age groups 65–74 and 75+ years is observed together with an increased percentage of GBS infected population of the same age groups (Supplementary Figure 1).

We found a significantly higher incidence among males 65–74 years old and 75+ years than among females in the same age groups (Table 2, Figure 1). This difference by gender was also observed by Ballard et al. (2); however, it was not observed in the study by Lamagni et al. (22). An explanation for this difference in gender-related GBS infections in Denmark cannot be established with the available data.

The serotype distribution in this study is very similar to the serotype distribution observed in other studies from developed countries (19, 21), in which it was also observed that serotypes III, Ia, and Ib were the most common serotypes in the age groups below 5 years of age; serotypes VI, VII, VIII, and IX were very rare among the age groups EOD, LOD, and 91 days −4 years; and for patients 5 years and older, all serotypes were observed, although with different frequencies. In this study, the most common serotypes in patients aged 20 years and above were serotypes III, Ia, and V (Figure 2). These serotypes were also the most prevalent serotypes among patients aged 65 years and above, which were also observed in Norway and Canada (19, 21), although they both found that serotype V was more common than serotype III. In the USA and Canada, an emergence of serotype IV has been observed in recent years in all age groups, and in some regions, serotype IV has been the second most common cause of invasive GBS in adults (20, 24). In Denmark, serotype IV was observed each year from 2005 to 2018; however, it was not observed as a predominant serotype in any of the age groups, and it constituted only 4.6% of the isolates (n = 87) detected in 2005–2018. Furthermore, we did not observe any significantly increasing trend for serotype IV in the period (Table 3, Figure 2). Serotype VIII was observed for the first time in Denmark in 2001 (12) and constituted 2.5% of the isolates detected from 2005 to 2018 with a significantly increasing trend (P < 0.05) (Table 3). This serotype was rather rare among patients younger than 20 years (Figure 2). Serotype VIII seems to be more common in Denmark compared with many other Western countries, and the prevalence resembles more the findings in Taiwan (25). This serotype is very seldomly observed in other European countries and Canada (19, 21, 26). We have no explanation for the higher prevalence in Denmark. Epidemiological studies generally find that GBS NT isolates comprise 5–10% of the detected invasive GBS isolates (3, 19, 22). In this study, the numbers of GBS NT isolates constituted 9.8% of the total number of isolates (Table 3).

The target group for GBS vaccination is generally considered to be pregnant women, although other groups, such as the elderly in general can be considered (18, 27). Currently, several capsular polysaccharide-based GBS vaccines are under development (18, 27). One of these vaccines [the hexavalent capsular polysaccharide conjugate vaccine (GBS6)] covers serotypes Ia, Ib, II, III, IV, and V (27). An evaluation of the predicted vaccine coverage for all Danish invasive GBS isolates for 2018 shows that the vaccine will cover around 77% (Table 3). The coverage for both EOD and LOD isolates will be around 89%. Although a hexavalent GBS vaccine thus theoretically could reduce EOD and LOD dramatically and reduce the overall GBS incidence by 77%, there are many reasons to assume that this might not happen in reality. For example, the GBS6 vaccine is only developed with focus on pregnant woman and not on all age groups (27). As the GBS6 vaccine covers six out of 10 known GBS serotypes, four serotypes are not included in the vaccine. As observed in this study (Table 3), there has been an increase in serotype VIII in recent years in Denmark. Furthermore, as previously mentioned, generally 5–10% of GBS isolates are NT [Table 3; (3, 19, 22)], which potentially could include unique unknown GBS serotypes. With the potential risk of non-vaccinated GBS carriers, increase in GBS serotypes not included in the GBS vaccine, and identification of possible new GBS serotypes, there is a great risk that the GBS vaccine efficacy will not meet the theoretical estimate. Although the serotype coverage of GBS vaccines and pneumococcal vaccines is very different, there is a possibility of observing serotype replacement in the GBS distribution as observed after introduction of pneumococcal vaccines (28).

Erythromycin and clindamycin are in general use in many countries, and resistance to these antibiotics is therefore monitored closely (19, 26). For erythromycin resistance, we observed an overall significantly increasing trend during the period 2005–2018, although with large fluctuations, whereas clindamycin resistance showed a non-significantly increasing trend, also with large fluctuations (Table 2, Figure 3). Moreover, there was a sharp increase in resistance rate of both antibiotics until 2009–2010, followed by a sharp decrease until 2012. These changes are very similar to changes in 2010 and 2011, respectively, reported by Martins et al. (26) and Alhhazmi et al. (19). We have not been able to identify any cause for these changes in Denmark, and the studies by Martins et al. (26) and Alhhazmi et al. (19) do not present an explanation to this observation either. During the period 2014–2017, a decrease in the prevalence of erythromycin resistance was observed, and clindamycin resistance seems to have reached a steady level (Figure 3). This development is also similar to what has been reported from Portugal (26).

The increasing prevalence of resistance toward erythromycin and clindamycin during 2012–2015 (Figure 3) is to some extent coinciding with an increasing proportion of the GBS patients being 75 years or older (Supplementary Figure 1). The developed countries are seeing an aging population with increasing burden of comorbidities requiring greater use of antibiotics (Supplementary Figure 1, Supplementary Table 2) (2, 22), thus possibly promoting development of resistance.

A weakness of this study is that it is based on voluntary submission of invasive GBS from DCMs. Previous studies in Denmark have estimated that the voluntary system results in the submission of around 58% of all invasive GBS isolates (11, 14). In the study by Ekelund et al. (14), they contacted 10 of 15 clinical microbiological departments and obtained data on their GBS positive cultures. Based on their data, it was estimated that we received 58% of all invasive isolates. The data presented in this study may therefore underestimate the incidence of GBS infections in Denmark, although a study by Ballard et al. (2) from parts of Denmark found similar GBS incidence data. Ballard et al. (2) studied invasive GBS infections in Copenhagen City, Copenhagen County, and Northern Denmark in the period from 2000 to 2010 and found an EOD incidence of 0.18 per 1,000 live births and an LOD incidence at 0.07–0.18 per 1,000 live births. In the same study, the total incidence was in the range from 1.9 to 2.4 per 100,000. These numbers correspond to the incidence found in the present study where 0.17 cases of EOD per 1,000 live births and 0.14 cases of LOD per 1,000 live births were found. The total incidence in our study during 2005–2018 was in the range 1.5–3.5 per 100,000. Also, the observed increases in GBS incidence in the elderly and some of the fluctuations observed for both antibiotic susceptibility and serotype distributions cannot be due to a sudden increase in submission of GBS isolates. We therefore believe that our data provide a representative picture of the GBS incidence in Denmark. Another limitation of the study is that we do not have data on penicillin susceptibility before 2012. However, because we did not find any penicillin non-susceptible isolates after 2012, we do not consider the absence of penicillin susceptibility data for 2005–2011 as a problem for providing a status of GBS penicillin susceptibility in Denmark.

A strength of this study is that it is nationwide and that all hospital departments are represented.

In conclusion, this study presents the GBS incidence for all age groups within the last 12 years in Denmark. We found that although EOD and LOD GBS incidences continue to be very low, we have seen high incidences showing a significantly increasing trend in GBS infections in the elderly, as also observed in other countries. Finally, we have observed an increase in serotype VIII disease in Denmark in recent years.

All datasets generated for this study are included in the article/Supplementary Material.

H-CS designed the study, analyzed the data, and drafted the manuscript. SH analyzed and reviewed the data, contributed to the manuscript, and critically revised the manuscript. Both authors have approved the final manuscript.