Blog

An additional six million new HIV infections and four million AIDS-related deaths could occur between 2025 and 2029 if US-supported HIV treatment and prevention services collapse, according to UNAIDS.

“This is not just a funding gap. It’s a ticking time bomb,” said UNAIDS Executive Director Winnie Byanyima at the launch of the organisation’s 2025 global AIDS update on Thursday. 

“We have seen services vanish overnight. Health workers have been sent home. And people – especially children and key populations – are being pushed out of care.” 

“Key populations” refer to people most vulnerable to HIV infection, including sex workers, men to have sex with men, people who inject drugs and young women.

Some of the immediate effects of the US withdrawal of funds since Donald Trump assumed the presidency in January include the closure of health facilities, healthworker job losses, and disrupted treatment, testing and prevention services.

The US President’s Emergency Plan for AIDS Relief (PEPFAR) had committed $4.3 billion in bilateral support in 2025 and “those services were stopped overnight when the US government shifted its foreign assistance strategies,” notes the UNAIDS report. 

PEPFAR had supported HIV testing for 84.1 million people and HIV treatment for 20.6 million people.

“Disruptions are being felt across the HIV response and pose a huge risk of increased mortality, a surge of new HIV infections, and the development of resistance to the most commonly used treatment regimens.”

In Mozambique, for example, over 30,000 health personnel have lost their jobs.

UNAIDS itself faces huge job losses, and is reducing its Geneva head office staff from 127 to a mere 19 employees, according to a report this week by Geneva Solutions.

The UN agency’s restructuring plan will cut staff by 54% globally, leaving 280 staff worldwide. 

HIV prevention programmes hit hard

External funding financed almost 80% of HIV prevention in sub-Saharan Africa, 66% in the Caribbean and 60% in the Middle East and North Africa, according to UNAIDS.

PEPFAR alone reached 2.3 million adolescent girls and young women with comprehensive HIV prevention services in 2024 and enabled 2.5 million people to use pre-exposure prophylaxis (PrEP). Many of these programmes have now stopped completely, according to UNAIDS. 

PrEP involves taking medication to prevent HIV infection and is usually taken by people at high risk of infection, and PEPFAR funded over 90% of PrEP initiations globally in 2024.

“Countries are reporting limited availability of PrEP and reduced activities to prevent new HIV acquisitions, including among adolescent girls and young women,” said UNAIDS.

In Nigeria, budget cuts have reduced PrEP initiation from 40,000 to 6000 people per month.

At the end of 2024, just before a sudden collapse in funding, new HIV infections had been reduced by 40% and AIDS-related deaths by 56% since 2010, Byanyima notes in the report. Countries had also reduced the annual number of children acquiring HIV from their mothers by 62% to 120,000 since 2010.

However, prevention efforts were already flatlining before the withdrawal of US aid. In 2024, there were 1.3 million new infections, which was almost the same as the year before.

“Over 210,000 girls and young women aged 15 to 24 acquired HIV in 2024 – an average of 570 new infections every day,” according to the UNAIDS report.

In 2024, 630,000 people died from AIDS-related causes, 61% of them in sub-Saharan Africa. “Community-led services, which are vital to reaching marginalised populations, are being defunded at alarming rates,” said UNAIDS.

“In early 2025, over 60% of women-led HIV organisations surveyed had lost funding or were forced to suspend services.”

Domestic budgets inadequate

Only 25 of the 60 low- and middle-income countries included in the report have increased their domestic budgets for HIV in 2026. The average increase amounts to 8%, approximately $180 million in additional domestic resources. 

“This is promising, but not sufficient to replace the scale of international funding in countries that are heavily reliant,” UNAIDS notes.

“It is important for donors to recognize that the option of increasing domestic HIV funding is not immediately or equally available to all countries,” UNAIDS notes.

“Combinations of debt distress, slow economic growth and underperforming tax systems leave many countries, notably in sub-Saharan Africa, with limited fiscal space to increase their domestic funding for HIV. “

It cites the recent International Conference on Financing for Development in Seville in Spain, as offering a way forward with “calls for debt relief, international tax cooperation and reform of international financial institutions”.

These measure would provide “the first steps towards a new economic settlement that can give countries the fiscal space needed to invest in the global HIV response”, UNAIDS notes.

“Urgent action and revived solidarity are needed to sustain the progress made and prevent a resurgence of HIV.”

Image Credits: UNAIDS.

Trends in SGLT-2 inhibitor prescriptions by purpose

Supplementary Fig. 2 illustrates the trend of the average monthly prescription volume by year, categorized according to the purpose of SGLT-2 inhibitor prescriptions. Of the 6787 patients analyzed, 5805 (85.5%) were prescribed SGLT-2 inhibitors for DM, 550 (8.1%) for CKD, and 432 (6.4%) for HF. Prescriptions have been rapidly increasing since 2018, with the number of prescriptions in 2023 being approximately 7.4 times higher than in 2018. Additionally, both CKD and HF prescriptions exhibited a marked rise from 2020 onward.

Baseline characteristics and prevalence of erythrocytosis

Among the 6787 patients included in this retrospective study at Bundang Seoul National University Hospital, 1145 (16.9%) developed erythrocytosis after initiating SGLT-2 inhibitor therapy, while 5642 (83.1%) did not. The median follow-up was 530 days (IQR, 277–981 days) for the overall cohort and 773 days (IQR, 445–1305 days) for the erythrocytosis subgroup. Table 1 summarizes the baseline characteristics of the study population, stratified by the presence of erythrocytosis. Patients with erythrocytosis were significantly younger, with only 11.2% being ≥ 70 years of age compared to 37.5% in the non-erythrocytosis group (p < 0.001). Males were more prevalent in the erythrocytosis group (88.9% vs. 60.0%, p < 0.001). A higher proportion of patients with erythrocytosis had a BMI ≥ 25 kg/m² (63.9% vs. 48.1%, p < 0.001). Regarding comorbidities, DM was more frequent in patients with erythrocytosis (89.9% vs. 85.4%; p < 0.001). In contrast, comorbidities including hypertension (HTN), dyslipidemia (DL), HF, CKD, coronary artery disease (CAD), cerebrovascular accident (CVD), peripheral artery disease (PAD), and chronic obstructive pulmonary disease (COPD), were more prevalent in the non-erythrocytosis group (all p < 0.001).

Current smoking (40.0% vs. 15.7%) and alcohol consumption (35.2% vs. 19.5%) were more prevalent in the erythrocytosis group (both p < 0.001). SGLT-2 inhibitor type did not differ (p = 0.590), with dapagliflozin most common (57.2%). Antiplatelet use was less frequent in the erythrocytosis group (36.3% vs. 41.9%; p = 0.001), while anticoagulation was similar (12.1% vs. 11.4%, p = 0.516). Baseline erythrocytosis was significantly more common in the erythrocytosis group than in the non-erythrocytosis group (26.6% vs. 1.2%; p < 0.001). In the analysis excluding patients with baseline erythrocytosis (n = 372), the incidence of erythrocytosis was 13.1% (n = 840/6415).

Multivariable analysis of risk factors for erythrocytosis

To address the potential impact of baseline erythrocytosis, a sensitivity analysis was performed by excluding patients with baseline erythrocytosis (n = 372). The remaining cohort (n = 6415) was analyzed for erythrocytosis risk factors. Multivariable analysis identified several risk factors for erythrocytosis (Table 2). Male sex (OR 3.24, 95% CI 2.47–4.26), BMI ≥ 25 kg/m² (OR 1.97, 95% CI 1.63–2.39), and current smoking (OR 2.41, 95% CI 1.96–2.96) were strongly associated with increased risk (all p < 0.001). Conversely, age ≥ 70 years (OR 0.46, 95% CI 0.35–0.59), HTN (OR 0.71, 95% CI 0.58–0.86), DL (OR 0.67, 95% CI 0.55–0.81), and CKD (OR 0.51, 95% CI 0.40–0.66) were associated with reduced risk. PAD, and COPD also showed protective effects (Table 2).

Effect of SGLT-2 inhibitors on hematologic parameters

Figure 1 illustrates the changes in Hb and Hct from baseline to peak levels in the total cohort. The median Hb increase for the entire cohort was 1.0 g/dL (IQR, 0.4–1.8 g/dL, p < 0.001), and the median Hct increase was 3.5% (IQR, 1.4–5.7%, p < 0.001). The median time to peak Hb was 210 days (IQR, 109–434 days) following SGLT-2 inhibitor initiation. Compared to baseline Hb and Hct, peak Hb and Hct showed a statistically significant increase, indicating a significant hematologic response to SGLT-2 inhibitor therapy, consistent with the known erythropoietic effects of these agents. Figure 2 shows that the change in Hb levels significantly differs based on the purpose of treatment (p < 0.05). CKD patients showed a median Hb increase of 0.9 (IQR, 0.3–1.6), while DM patients had a median increase of 1.0 (IQR, 0.4–1.7). HF patients exhibited the highest median Hb increase at 1.3 (IQR, 0.5–2.1). This indicates that HF patients experienced the most significant rise in Hb levels, whereas CKD patients showed the lowest increase. Patients without baseline erythrocytosis exhibited significantly greater Hb changes compared to those with erythrocytosis (p < 0.05). The median Hb change was 1.1 (IQR, 0.4–1.8) in patients without erythrocytosis, whereas it was 0.3 (IQR, -0.4–0.8) in those with erythrocytosis (Fig. 3). Table 3 summarizes the hematologic parameters at baseline and peak levels for the total cohort (n = 6787), stratified by sex. Significant sex differences were observed in both baseline and peak Hb and Hct levels, with males consistently exhibiting higher values than females.

Among patients with erythrocytosis (n = 1145), the median hemoglobin increase was 1.4 g/dL (IQR, 0.8–2.2), and the median hematocrit increase was 4.7% (IQR, 2.7–6.8). The median time to peak hemoglobin was 361 days (IQR, 161–634). Serial Hb changes were assessed over 12 months in 769 patients receiving SGLT-2 inhibitors with at least 12 months of Hb follow-up data (Fig. 4). Hb levels rose rapidly within the first 3 months, followed by a more gradual increase, maintaining an overall upward trend over the 12-month period.

Of the 6787 patients, 4756 (70.1%) continued treatment, 944 (13.9%) were lost to follow-up or referred out, and 1087 (16.0%) discontinued treatment. Among the 983 patients with known reasons for discontinuation (excluding 104 patients with unknown reasons), the most common were poor DM control (34.7%), side effects other than erythrocytosis (32.2%), and diabetes medication reduction (14.3%), with erythrocytosis accounting for only 5 (0.5%) discontinuations.

Among patients who developed erythrocytosis while using SGLT-2 inhibitors, hematology consultations occurred in 18 (1.6%) patients, with median serum erythropoietin (EPO) levels of 12.5 mIU/mL (range 7.1–17.61 mIU/mL, reference 2.59–18.50 mIU/mL). Of these, 10 (55.6%) underwent JAK2 mutation testing, 4 (22.2%) had bone marrow examinations, 2 (11.1%) received phlebotomy, and 2 (11.1%) were prescribed aspirin. Four patients discontinued SGLT-2 inhibitors: two due to erythrocytosis (one resolved spontaneously, and one improved with phlebotomy) and two due to well-controlled diabetes (Supplementary Fig. 3).

Thrombosis risk factors

Among the 6787 patients, 0.5% (33 patients) developed thrombosis during the treatment period with SGLT-2 inhibitors (Table 4). HF was more common in the thrombosis group (33.3% vs. 19.5%, p = 0.046), as were antiplatelet use (63.6% vs. 40.8%, p = 0.008), anticoagulant use (36.4% vs. 11.4%, p < 0.001), and baseline erythrocytosis (15.2% vs. 5.4%, p = 0.032). Other factors, including age ≥ 70 years, sex, BMI, eGFR, most comorbidities (e.g., DM, HTN. CAD), smoking, and alcohol use, showed no significant differences between groups.

Antiplatelet agent use (OR 3.57, 95% CI 1.60–7.97, p = 0.002), anticoagulant use (OR 5.93, 95% CI 2.60–13.57, p < 0.001), and baseline erythrocytosis (OR 3.75, 95% CI 1.41–9.96, p = 0.008) were significantly linked to an increased risk of thrombosis in multivariable analysis (Table 5).

Among 33 patients, five patients had erythrocytosis at the time of the thrombosis event and had experienced at least one episode of erythrocytosis within the preceding six months. Notably, all five patients exhibited arterial thrombosis without any venous thrombosis events. However, these five patients also had underlying conditions (atrial fibrillation, severe coronary calcification, large artery atherosclerosis) associated with their thrombosis events, making it unlikely that the thrombosis events were directly related to erythrocytosis. Thrombotic events occurred in 1.8% (n = 21) of the erythrocytosis group, compared with 0.2% (n = 12) in the non-erythrocytosis group (p < 0.001). Within the erythrocytosis group, peak Hb levels were stratified into tertiles: low (16.1–16.9 g/dL), medium (16.9–17.5 g/dL), and high (17.5–22.0 g/dL). Thrombosis rates were 1.4% (n = 6), 2.2% (n = 8), and 2.0% (n = 7), respectively (p = 0.649), suggesting no clear dose–response relationship between Hb levels and thrombosis.

People at higher risk of HIV, such as gay men and people who inject drugs, are facing record levels of criminalisation worldwide, according to UNAids.

For the first time since the joint UN programme on HIV/Aids began reporting on punitive laws a decade ago, the number of countries criminalising same-sex sexual activity and gender expression has increased.

In the past year, Mali has made homosexuality a criminal offence, where the law previously only banned “public indecency”, and has also criminalised transgender people. Trinidad and Tobago’s court of appeal has overturned a landmark 2018 ruling that decriminalised consensual same-sex relations, reinstating the colonial-era ban. In Uganda, the 2023 Anti-Homosexuality Act has “intensified the proscription of same-sex relations”, and Ghana has moved in a similar direction with the reintroduction of legislation that would increase sentences for gay sex.

The crackdown on gay rights comes as the fight against HIV/Aids has been hit by abrupt US funding cuts, which have combined with “unprecedented” humanitarian challenges and climate crisis shocks to jeopardise hopes of ending the global epidemic this decade, UNAids said.

Several groups of people, known as “key populations”, are more likely to be infected with HIV. They include sex workers, gay men and other men who have sex with men, people who inject drugs, transgender people, and those in prisons and other enclosed settings.

In 2025, only eight of 193 countries did not criminalise any of those groups or behaviours, or criminalise non-disclosure of HIV status, exposure or transmission, according to the report.

The number of people infected by HIV or dying from Aids-related causes in 2024 was the lowest for more than 30 years, according to the UNAids annual report, at 1.3 million and 630,000 respectively.

Progress was uneven – ranging from a 56% fall in infections since 2010 in sub-Saharan Africa to a 94% increase in the Middle East and North Africa. But coupled with scientific advances – such as twice-yearly drugs to prevent infection – the world had the “means and momentum” to end Aids as a public health threat by 2030, an internationally agreed goal, it said.

However, that has been “seriously jeopardised” in the early months of this year after sweeping US aid cuts that could undo decades of progress. In January, Donald Trump cut funding that had underpinned much of the global HIV response almost overnight.

The report highlights HIV-prevention services as an area of concern, with many particularly reliant on donor funding. The reported number of people receiving preventive drugs in Nigeria in November 2024 was approximately 43,000. By April 2025, that number had fallen to below 6,000.

Activists say access to prevention will be a particular issue for key populations, who may not be able to access mainstream healthcare due to factors such as stigma or fear of prosecution, but relied on donor-funded community clinics that have now closed.

Key populations were “always left behind”, said Dr Beatriz Grinsztejn, president of the International Aids Society (IAS).

The report is being released before an IAS conference next week in Kigali, Rwanda, where researchers will share data on the impact of cuts.

Modelling by Bristol University calculated that a one-year halt in US funding for preventive drugs in key populations in sub-Saharan Africa would mean roughly 700,000 people no longer used them, and lead to about 10,000 extra cases of HIV over the next five years.

UNAids modelling suggests that without any replacement for funding from US Pepfar (president’s emergency plan for Aids relief), an additional 4m deaths and 6m new infections could be expected globally by 2029.

However, Winnie Byanyima, executive director of UNAids, said 25 of the 60 low- and middle-income countries included in the report had found ways to increase HIV spending from domestic resources to 2026. “This is the future of the HIV response – nationally owned and led, sustainable, inclusive and multisectoral,” she said.