In total, 15,940 diagnoses of 18,534 study participants were eligible for the analysis of HIV-ICs in the time period 1999 until 2023 (Table 2). Of those, 14,174 diagnoses were included from the ClinSurv-HIV Study and 1,766 diagnoses from the HIV-1 Seroconverter Study (see Supplemental Fig. 1 for an overview of the excluded patients). The sociodemographic characteristics of eligible study participants overall and stratified by cohort study are shown in Table 2. The majority of participants in both cohort studies were male (81%) and the median age at the beginning of the observation was 36 years (interquartile range [IQR]: 29–44 years). Overall, 56% reported being MSM as the likely route of HIV-transmission, followed by 16% who reported heterosexual contact as the likely HIV-transmission route (Table 2). In the HIV-1 Seroconverter study, we observed a higher percentage of male study participants (94%) with the majority reported being MSM as their likely route of HIV-transmission compared to the ClinSurv-HIV Study. The age difference between the two cohort studies was small.

The median observation time was 9 months (interquartile range (IQR): 6–39 months) and differed between the participants of the two cohort studies. While ClinSurv-HIV participants included in this analysis had a median observation time of 8 months (IQR: 6–32 months), HIV-1 Seroconverter participants had a median observation time of 24 months (IQR: 8–58 months). The median observation time did not differ between study participants with or without records of HIV-ICs and was 9 months (IQR participants with HIV-IC: 6–42 months; IQR participants without HIV-IC: 6–31 months). Overall, 51% of participants were reported with any of the selected HIV-ICs, while 29% of participants had one diagnosed HIV-IC, 13% had two HIV-ICs and 5.5% three HIV-ICs. The remaining 3.2% of participants had 4–13 HIV-ICs.

In the following paragraphs the prevalence, incidence of the respective diagnoses will be presented in the disease categories STIs, viral hepatitis, HIV-associated and AIDS-defining diagnoses.

The prevalence of STIs among participants of the HIV-1 Seroconverter study, namely chlamydial infection, anogenital (venereal) warts, gonococcal infection, diagnoses with papillomavirus as the cause of disease and syphilis, ranged from 0.17% for diagnoses with papillomavirus as the cause of disease to 7.8% for syphilis. Similarly, the lowest incidence rate was found for papillomavirus as the cause of diseases with 1.0 per 1,000 person-years and the highest incidence for syphilis with 34 per 1,000 person-years. Prevalence and incidence among HIV-1 Seroconverter cohort participants for the respective infections is shown in Table 3. There were no diagnoses for trichomoniasis reported for the study participants, hence trichomoniasis is not included in the subsequent tables and figures.

Considering the HIV-transmission risk of participants, there was no association with diagnosis of anogenital (veneral) warts, chlamydial infection or papillomavirus as the cause of diseases. For gonococcal infection and syphilis an association with HIV-transmission risk was found (p-value: 0.017 and 0.018, respectively). Of participants indicating MSM-contact as their HIV-transmission risk 5.0% and 8.5% tested positive for gonococcal infection or syphilis, respectively (see supplemental Table 2 for details).

The incidence rates for the STI diagnoses were determined for four different time periods, starting in 2008 as described in the methods. Incidence rates for papillomavirus were not determined due to sparse data. Starting from the first time period (2008–2010) an increasing incidence rate can be seen for all four STI analysed until the time period 2015–2018. This increase over time was confirmed by a test for trend (nogenital warts: p-value: 0.004; hlamydial infection: p-value: < 0.001; onococcal infection: p-value: < 0.001; syphilis: p-value: < 0.001). The last time period has wider confidence intervals due to the decreasing number of PYs contributed, making the interpretation of the incidence rates less reliable. An overview of the incidence rate per time period and STI is shown in Fig. 1.

For the analysis of viral hepatitis, diagnoses of acute and chronic hepatitis of the types A, B, C, D, E and with unknown hepatitis type were included. Data from both cohort studies were analysed, from 1999 until 2018 for ClinSurv-HIV participants and until 2023 for HIV-1 Seroconverter participants. The analysis does not differentiate between acute and chronic hepatitis. The prevalence found for the overall study period ranged from 0.01% for hepatitis E to 11% for hepatitis B and incidence rate per 1.000 PYs ranged from 0.03 for hepatitis E to 18 for hepatitis B. Detailed information on prevalence and incidence are shown in Table 4. There were no diagnoses for hepatitis D reported.

For viral hepatitis diagnoses an association with the HIV-transmission risk of participants was found for hepatitis A, B, C and viral hepatitis with unknown type (p-value < 0.001, respectively). The highest proportion of participants with hepatitis A diagnoses was found among participants reporting coming from a high-prevalence country for HIV (19%) as their likely HIV-transmission risk, followed by those reporting IDU (11%). For hepatitis B, 20% of participants reporting IDU and 20% of participant coming from a high-prevalence country for HIV had a hepatitis B diagnosis. For hepatitis C and viral hepatitis with unknown type, 40% and 6.5% of participants with the HIV-transmission risk IDU had a respective HIV-IC diagnosis (see supplemental Table 2 for details).

The incidence rate by time period (1999–2023) for viral hepatitis A, B, C and diagnoses with unknown type are shown in Fig. 2. Trends in diagnoses of hepatitis E were not analysed due to sparse data. There were no diagnoses found for hepatitis A and viral hepatitis (type unknown) in the time period 2019–2023. The incidence rates for hepatitis A and B decreased compared to the first time period, while the confidence intervals for hepatitis C and diagnoses with unknown type remained slightly overlapping. As mentioned for the incidence rates of STIs, the interpretation of incidence during the last time period is challenging due to the wide confidence intervals caused by the very small amount of PYs contributed. Despite these overlapping confidence intervals the test for trend showed a decreasing incidence rate for hepatitis A-C as well as hepatitis cases with unknown type (hepatitis A: p-value: < 0.001; hepatitis B: p-value: < 0.001; hepatitis C: p-value: 0.02; viral hepatitis (type unknown): p-value: < 0.001).

Incidence rates of viral hepatitis diagnoses per 1.000 person-years among ClinSurv-HIV and HIV-1 Seroconverter cohort participants by time period 1999–2005, 2006–2010, 2011–2014, 2015–2018 and 2019–2023

We analysed the following HIV-associated diagnoses, according to the category B of the CDC-classification for HIV infections: dysplasia of the cervix uteri, herpes zoster, Hodgkin lymphoma, malignant neoplasm of anus and anal canal, malignant neoplasm of bronchus and lung, oral hairy leukoplakia, and seborrheic dermatitis. In addition, we included infectious mononucleosis to this category, as the symptoms can be very similar to those during acute HIV-infection. Again, diagnoses among participants of both cohorts were analysed from 1999–2023. No diagnoses were found for malignant neoplasm of anus and anal canal and malignant neoplasm of bronchus and lung. The lowest prevalence and incidence rates were found for Hodgkin lymphoma with 0.26% and 0.80 per 1.000 PYs, respectively. The highest prevalence and incidence rate was found for herpes zoster with 5.2% and 22 per 1.000 PYs. An overview of the prevalence and incidence rates found for the diagnoses in this category is shown in Table 5.

Among participants with any of the ICs considered as HIV-associated, an association with the HIV-transmission risk was found for the following diagnoses: herpes zoster (p-value: 0.02), oral hairy leucoplakia (p-value: 0.001) and seborrheic dermatitis (p-value: < 0.001). For herpes zoster, participants reporting heterosexual transmission or MSM had the highest diagnosis prevalence with 5.9% and 5.4%, respectively. For both diagnoses, oral hairy leucoplakia and seborrheic dermatitis the groups “other” and “unknown” had the highest prevalence of the respective diagnosis, while the prevalence ranged from 2.2%—2.8% and 2.1%—2.7%, respectively, for participants reporting being MSM, heterosexual transmission or IDU as their likely HIV-transmission route (see supplemental Table 2).

The incidence rates by time period (1999–2023) and per HIV-associated diagnoses are shown in Fig. 3. There were no diagnoses found for dysplasia of cervix uteri and oral hairy leukoplakia in the time period 2019–2023. Based on the test for trend decreasing incidence rates over time can be seen for oral hairy leukoplakia (p-value: < 0.001), dysplasia of cervix uteri (p-value: 0.04) and infectious mononucleosis (p-value: 0.002). There was no change of incidence confirmed over time for herpes zoster and seborrheic dermatitis.

Incidence rates of HIV-associated diagnoses per 1.000 person-years among ClinSurv-HIV and HIV-1 Seroconverter cohort participants by time period 1999–2005, 2006–2010, 2011–2014, 2015–2018 and 2019–2023

AIDS-defining diagnoses were analysed from 1999–2023. The respective diagnoses used are listed in Table 1. The highest prevalence and incidence rate found in this group of diagnoses was for candidiasis with a prevalence of 13% and an incidence rate of 30 per 1,000 PYs. Diagnosis of malignant neoplasm of cervix uteri showed the lowest prevalence with 0.2% and lowest incidence rate with 1.0 per 1,000 PYs (see Table 6).

For the AIDS-defining diagnoses an association was found between diagnosis and HIV-transmission risk for herpes simplex infection (p-value: 0.016), candidiasis (p-value: < 0.001), Kaposi sarcoma (p-value: < 0.001), non-Hodgkin lymphoma (p-value: 0.005), pneumonia (p-value: < 0.001), pneumocystis (p-value: < 0.001) and tuberculosis (p-value: < 0.001; see supplemental Table 2). Highest prevalence for herpes simplex infections was found among participants reporting heterosexual transmission (4.0%) and MSM (3.8%), for candidiasis highest prevalence was found for heterosexual transmission (15%) and IDU (14%). For Kaposi sarcoma diagnosis, people reporting MSM as their likely HIV-transmission route had the highest diagnosis prevalence (3.3%), for pneumonia diagnosis it was IDU (7.1%). The HIV-transmission route with the highest diagnosis prevalence of pneumocystis were people reporting heterosexual transmission (5.8%) and for tuberculosis the highest diagnosis prevalence was found among participant reporting coming from a country of high HIV-prevalence (6.7%).

The incidence rates by time period (1999–2023) and per AIDS-associated diagnoses are shown in Fig. 4. There were no diagnoses found for herpes simplex infections, pneumocystosis, pneumonia and tuberculosis in the time period 2019–2023. Decreasing incidence rates over time, based on the test for trend, are observed for candidiasis (p-value: < 0.001), Kaposi sarcoma (p-value: 0.03), non-Hodgkin lymphoma (p-value: 0.03), pneumonia (p-value: < 0.001) and tuberculosis (p-value: 0.02). The incidence of abnormal weight loss increased over time (p-value: < 0.001).

Incidence rates of AIDS-defining diagnoses per 1.000 person-years among ClinSurv-HIV and HIV-1 Seroconverter cohort participants by time period 1999–2005, 2006–2010, 2011–2014, 2015–2018 and 2019–2023

To characterize the immune status, CD4 counts were analysed as a proxy. We chose the time wise closest CD4-count to the respective diagnosis (until ± 30 days) and categorised the CD4-counts into the three categories < 350 cells/µl blood, >  = 350—< 500 cells/µl blood, >  = 500 cells/µl blood. The number of diagnoses with available CD4 counts is shown in Fig. 5, while the total number of respective diagnoses can be found as the prevalence in the respective tables per disease category above. The proportion of CD4 counts is therefore based on the total number of CD4 counts available per diagnosis. Among STI diagnoses more than half of the diagnoses occurred when participants had CD4 counts of >  = 500 cells/µl, while only 5.9–22% had CD4 counts of < 350 cells/µl blood (Fig. 5). This differs for viral hepatitis diagnoses. Here 25–34% of diagnoses occurred when participants had CD4 counts of >  = 500 cells/µl blood around their hepatitis diagnosis and 42–55% CD4 counts of < 350 cells/µl. The CD4 counts at diagnosis further decrease if measured around the diagnoses of AIDS-defining conditions, such as Kaposi sarcoma, where 87% of participants have a CD4 count of < 350 cells/µl blood or pneumocystosis with 96% of participants with a CD4 count of < 350 cells/µl blood. For a detailed overview per HIV-IC, see Fig. 5.

Proportion of CD4-cell count/µl blood category for diagnoses of the respective HIV-IC with available data. HIV-ICs with less than 10 available observations were excluded