Integrase Inhibitors Suppress HIV Faster but Bring Metabolic Trade-offs in Real-World Chinese Cohort

People starting HIV treatment with modern integrase strand transfer inhibitor (INSTI)-based regimens achieve documented virological suppression at a higher rate and experience far fewer episodes of high-level viremia than those starting older non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimens, according to a province-wide cohort study conducted in Shandong, China. The trade-off, the researchers report in BMC Infectious Diseases, is a distinct metabolic and laboratory signature within the first year of therapy: greater weight gain, rising cholesterol, elevated bilirubin, and early declines in kidney function estimates, offset by a lower risk of liver enzyme elevation.

The analysis drew on de-identified records from the Shandong Province subset of China’s National Free Antiretroviral Treatment Program, capturing 7,474 adults aged 18 or older who began their first antiretroviral therapy between January 1, 2023 and December 31, 2025. Of these, 5,046 initiated NNRTI-based regimens and 2,428 initiated INSTI-based regimens. Because the data came from routine care rather than a controlled trial, the study reflects how these drug classes perform when prescribed across an entire province’s treatment network, including patients who might be excluded from clinical trials.

Virologically, the INSTI group fared better on both primary measures. After adjustment for baseline differences, INSTI-based initial therapy was associated with a higher rate of documented virological suppression, with an adjusted incidence rate ratio of 1.09 (95% confidence interval 1.07–1.12). More strikingly, the risk of documented high-level viremia between 6 and 24 months after starting treatment was reduced by more than three-quarters, with an adjusted risk ratio of 0.23 (95% CI 0.15–0.35). High-level viremia serves as an early warning signal for treatment failure and emerging drug resistance, so a 77% relative reduction carries substantial clinical weight.

The metabolic picture was more complicated. At approximately 12 months on therapy, patients on INSTI-based regimens showed higher risks of a 10% or greater increase in body mass index, incident elevated total cholesterol, incident total bilirubin elevation, and a 10% or greater decline in creatinine-based estimated glomerular filtration rate. The one favorable laboratory finding was a lower risk of incident alanine aminotransferase elevation, suggesting less liver enzyme stress with the integrase-based approach.

Continuous trajectories of body mass index, total cholesterol, alanine aminotransferase, total bilirubin, and kidney function tracked the same directions as the binary outcomes, reinforcing that these were not statistical artifacts of arbitrary thresholds. Linear mixed-effects models confirmed that the INSTI group’s weight and cholesterol curves climbed more steeply over follow-up, while their bilirubin and kidney function curves moved in less favorable directions, and their liver enzyme trajectories rose more gently.

Statistically, the team deployed a battery of methods suited to the data structure: Poisson regression with time offset for suppression rates, modified Poisson regression for common binary outcomes, Firth penalized logistic regression for rare events where conventional models fail, and linear mixed-effects models for continuous laboratory trajectories. This combination allowed the investigators to extract signal from a large, heterogeneous real-world dataset where missing laboratory values and uneven follow-up complicate inference.

An important caveat shapes how the results should be read. The authors emphasize that the NRTI backbone composition differed substantially between the two strategy groups, meaning the findings should be interpreted as comparisons of contemporary treatment strategies rather than as independent effects of the integrase inhibitors themselves. In routine Chinese care, INSTI regimens and NNRTI regimens are typically paired with different nucleoside analog combinations, so the metabolic signals may partly reflect backbone drugs rather than the integrase class alone.

INSTI-based regimens have expanded rapidly in China as initial therapy, following the pattern seen in high-income countries where integrase inhibitors dominate first-line guidelines because of their potency, high barrier to resistance, and tolerability. This study offers some of the first province-scale real-world evidence on how that transition plays out in a middle-income setting with a free national treatment program, where laboratory monitoring intensity and patient demographics differ markedly from the trials that shaped international guidelines.

The authors conclude that INSTI-based initial therapy delivers favorable early virological outcomes alongside distinct metabolic and laboratory patterns in routine HIV care in Shandong. They argue the findings support integrated post-initiation monitoring of weight, lipid levels, bilirubin, and creatinine-based renal function indicators as standard practice, rather than waiting for symptoms or late complications to emerge.

For clinicians managing the growing population of people on long-term antiretroviral therapy, the study underscores that virological success and metabolic safety must be weighed together when selecting initial regimens. As integrase inhibitors continue to displace NNRTIs worldwide, real-world cohorts like this one will remain essential for detecting the slower-moving laboratory signals that randomized trials, with their shorter horizons and selected populations, often cannot capture.

Subject of Research: Virological effectiveness and metabolic outcomes of INSTI-based versus NNRTI-based initial antiretroviral therapy in Shandong, China

Article Title: Virological effectiveness and metabolic/laboratory outcome profile of INSTI-based versus NNRTI-based initial ART in Shandong, China: a province-wide real-world cohort study

Article References: Yuan, H., Zhu, X., Zhang, N., Han, L., Liao, M., Hao, L., Sun, X., Li, Y., Li, L., Jiao, K., Zhu, M., Wang, G., & Ma, W. (2026). Virological effectiveness and metabolic/laboratory outcome profile of INSTI-based versus NNRTI-based initial ART in Shandong, China: a province-wide real-world cohort study. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14464-3

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14464-3

Keywords: HIV, antiretroviral therapy, integrase strand transfer inhibitor, virological suppression, metabolic outcomes, real-world evidence, China, Virological, effectiveness, metabolic, laboratory, outcome