Article Text
PDF
Abstract
Background With the ageing of people living with HIV/AIDS (PLWHA), the prevalence of chronic comorbidities, especially hyperglycaemia, is increasing among elderly PLWHA. Antiretroviral therapy (ART) is associated with fasting plasma glucose (FPG) levels. This study aimed to investigate both short-term and long-term FPG characteristics and trends across different ART regimens in elderly Chinese PLWHA.
Methods This retrospective cohort study, based on hospital treatment information, classified ART regimens as this retrospective cohort study used hospital treatment data. ART regimens are classified into three categories: non-nucleoside reverse transcriptase inhibitors (NNRTIs) based, protease inhibitors (PIs) based and integrase strand transfer inhibitor (INSTIs) based. Propensity score matching was applied to control for confounding factors. Follow-up FPG characteristics were then described, and a generalised linear mixed model was employed to estimate FPG trends under different regimens within 1-year and 5-year periods following ART initiation.
Results Participants had an average age of 58.28 years, with 75.02% male. FPG increased following ART initiation, with the most significant rise within 1 year of ART, followed by stabilisation. The FPG increase within 1 year was slower in the PIs-based group compared with the NNRTIs-based group (β=−0.08, 95% CI −0.15 to –0.01), while there was a higher prevalence of diabetes within 5 years of ART (31.55% vs 22.33%, standardised difference=0.357). The FPG increase within 1 year of ART did not differ between NNRTIs-based and INSTIs-based groups (β=−0.01, 95% CI −0.20, 0.18).
Conclusion Our study highlights that elderly Chinese PLWHA experience an increase in FPG levels, particularly during the first year of ART, with variations observed across different ART regimens. The higher long-term prevalence of diabetes in the PIs-based regimen group emphasises the need for tailored glucose management strategies. Routine glucose monitoring and proactive management are crucial for preventing and controlling diabetes in this population, particularly given the long-term metabolic risks associated with ART.
- HIV & AIDS
- China
- Medicine
- DIABETES & ENDOCRINOLOGY
Data availability statement
Data are available on reasonable request. The data that support the findings of this study are available from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
It is the first study to explore both short-term and long-term fasting plasma glucose (FPG) trends following antiretroviral therapy (ART) initiation in elderly Chinese people living with HIV/AIDS (PLWHA).
It explicitly analyses the impact of three common ART regimens, with a particular focus on the integrase strand transfer inhibitors-based regimen, on FPG trends.
The study uses propensity score matching to control for baseline confounders, which is a distinctive methodology compared with other adjustment methods, enhancing its applicability to real-world clinical settings.
The diagnostic criteria for diabetes in the study were relatively narrow, relying solely on FPG measurements.
The study exclusively focused on PLWHA without including a reference group of elderly individuals without HIV/AIDS.
Introduction
The ageing of people living with HIV/AIDS (PLWHA) has become increasingly severe worldwide in recent years. According to data from the Joint United Nations Programme on HIV and AIDS, the proportion of global PLWHA over 50 years old (elderly PLWHA) increased from 8% in 2000 to 16% in 2019.1 This ageing trend can be attributed to the decreased mortality rates and increased life expectancy among PLWHA, owing to the widespread use of highly effective antiretroviral therapy (ART).2 Additionally, the number of newly diagnosed elderly PLWHA has risen significantly in recent years, reflecting a broader demographic ageing trend in several countries, including China.3 Ageing contributes to a higher prevalence of chronic disease comorbidities among elderly PLWHA, with hyperglycaemia being one of the most common conditions.4 Insulin resistance tends to increase with age due to various mechanisms, including mitochondrial dysfunction, increased intramyocellular lipids and inflammation. This reduction in insulin sensitivity impairs glucose metabolism, heightening the risk of type 2 diabetes in older adults.5
PLWHA undergoing ART are at a higher risk of developing hyperglycaemia compared with HIV-naïve individuals. A study by Zhang C et al6 reported a diabetes prevalence of 2.62 per 100 person-years among 1415 PLWHA in a Chinese multicentre clinical trial, which was significantly higher than the prevalence observed in HIV-negative Chinese adults of the same age range (0.43–1.55 per 100 person-years). Similarly, a study involving 4000 elderly PLWHA in the USA found a higher prevalence of diabetes (28.8%) compared with their HIV-negative counterparts (24.2%).7 Although infectious diseases can exacerbate insulin resistance and unveil underlying beta cell deficiency, PLWHA commonly experience weight loss, muscle wasting and micronutrient deficiencies, which contribute to malnutrition.8 Malnutrition can further increase insulin resistance through mechanisms such as mitochondrial dysfunction in beta cells and altered metabolic programming.9 10
With increased access to ART, HIV is increasingly recognised as a chronic, manageable condition. Consequently, there has been a rise in overweight and obesity among HIV patients.11 The increased adiposity associated with obesity disrupts metabolic processes, leading to dysregulated adipokine secretion, chronic low-grade inflammation, endoplasmic reticulum (ER) stress and mitochondrial dysfunction.12 These factors collectively impair insulin signalling in key metabolic tissues, including adipose tissue, muscle and liver, ultimately resulting in systemic insulin resistance.13 Additionally, certain therapies for infectious diseases can contribute to the development of insulin resistance and dysglycaemia. Various ART regimens have distinct mechanisms of action on fasting plasma glucose (FPG). Non-nucleoside reverse transcriptase inhibitors (NNRTIs), which are currently the predominant first-line treatment regimens, may induce diabetes through mechanisms involving steatosis, insulin resistance and mitochondrial dysfunction.14 Protease inhibitors (PIs) can lead to insulin resistance by impacting insulin signalling pathways.15 Among the more recent ART regimens, integrase strand transfer inhibitors (INSTIs) exhibit a more favourable metabolic profile compared with older regimens. However, they also influence glucose transport via glucose transporter type 4 (GLUT-4), potentially leading to insulin resistance.16
Previous research has indicated that PIs significantly elevate the risk of diabetes compared with NNRTIs.17 18 However, the effects of INSTIs on diabetes risk remain inconsistent. A study conducted in South Africa found that patients on PIs were at least 21 times more likely to develop diabetes than those on NNRTIs.19 Nevertheless, data from observational studies on the association between INSTI use and diabetes mellitus have produced conflicting results. A large North American cohort demonstrated a 22% higher risk of diabetes in patients on INSTIs and PIs compared with those on NNRTIs,20 whereas findings from a sizeable French cohort did not identify any such association.21
The onset of drug-induced diabetes and hyperglycaemia is typically characterised by a delayed occurrence.22 A meta-analysis reported a significant increase in FPG levels 18 months after the initiation of ART.23 In China, a study found that PLWHA receiving NNRTIs-based regimens (TDF+3TC+EFV and AZT+3TC+EFV) experienced a significant rise in FPG at 12 months from the initiation of treatment.24 However, there is a paucity of studies investigating the effects of long-term ART treatment on blood glucose levels.
In summary, there is inconsistency in the impact of INSTIs on blood glucose levels, and there is a notable absence of studies investigating the long-term effects of different ART regimens on blood glucose levels. Furthermore, there is a shortage of analyses focusing on glycaemic trends following ART among elderly PLWHA, whose changes in glucose levels may be more pronounced. To address these knowledge gaps, this retrospective cohort study aims to delineate trends in FPG among elderly PLWHA after the initiation of ART, both in the short term (within 1 year of ART) and long term (within 5 years of ART). Furthermore, the study seeks to analyse the impact of different ART regimens (NNRTIs based, PIs based and INSTIs based) on FPG trends. Based on the findings from our study, we intend to propose practical measures for the prevention and management of hyperglycaemia among elderly PLWHA.
Methods
Study design and participants
This retrospective cohort study encompassed elderly PLWHA who initiated ART between January 2010 and December 2019 at Guangzhou Eighth People’s Hospital. Baseline and follow-up information for these individuals were sourced from the Chinese National Infectious Disease Surveillance System. As the inaugural ART centre in Guangdong Province, China, Guangzhou Eighth People’s Hospital has provided ART services to a substantial number, exceeding 21 000 PLWHA, since its establishment in 2004.25 In 2010, the Project Management Programme for AIDS Prevention and Control in China introduced standardised management protocols for PLWHA on ART, leading to more systematic and comprehensive information on ART post-2010.26 The scale of antiviral treatment has continued to grow over the years. In 2016, a new policy to expand treatment was implemented, and all infected people can receive free treatment after diagnosis.27 PLWHA on ART necessitate long-term and regular follow-up. In China, individuals initiating ART are scheduled for follow-up visits at 1, 2, 3, 6 and 9 months after ART initiation, followed by 3-month intervals thereafter. These follow-up visits involve clinical assessments and laboratory tests.28
The inclusion criteria were (1) elderly PLWHA aged over 50 years, (2) ART initiation at this hospital after 2010, (3) initiation of ART with regimens including NNRTIs-based, PIs-based and INSTIs-based regimens and (4) attendance at least one follow-up clinic visit for antiretroviral medication and health examination after initiating ART. The exclusion criteria were (1) abnormal FPG (≥6.1 mmol/L) at ART initiation, (2) lack of at least one follow-up FPG examination before regimen adjustment.
A total of 18 479 PLWHA initiated ART in Guangzhou Eighth People’s Hospital from 2010 to 2019. Among them, there were 2442 elderly PLWHA who met the inclusion criteria and 937 patients who met the exclusion criteria. A total of 1505 patients were finally included in our study. The flow chart of inclusion is shown in online supplemental figure S1. A baseline comparison of patients excluded and included within 1 year is shown in online supplemental table S4.
Supplemental material
Supplemental material
The data for our study were sourced from the electronic medical database following deidentification, ensuring the absence of patient-identifiable information.
Measurement
Initial ART regimens
In our study, the initial ART regimens were categorised into three following the guidelines outlined in the Chinese National Manual on Free HIV Antiviral Therapy.28 The regimes were classified as NNRTIs-based, PIs-based and INSTIs-based regimens. NNRTIs-based regimens included nevirapine (NVP) or efavirenz (EFV) combined with lamivudine (3TC) and tenofovir disoproxil fumarate (TDF) or zidovudine (AZT) in our study. PIs-based regimens comprised those containing lopinavir (LPV) (no patients in our study used PIs other than LPV). INSTIs-based regimens involved regimens containing dolutegravir (DTG) or raltegravir (RAL). During the follow-up period, some PLWHA undergoing ART experienced adjustments to their treatment regimens. The frequency and the total number of regimen adjustments during follow-up were calculated based on regimen records at each follow-up visit.
FPG characteristics
FPG levels were measured at baseline and each follow-up visit. To minimise the potential influence of unexamined factors, such as glucose-lowering medication and dietary glycaemic control, was minimised, patients with abnormal FPG levels (≥6.1 mmol/L) at the initiation of ART were excluded from the analysis. According to the American Diabetes Association criteria, FPG levels were categorised as normal (FPG<6.1 mmol/L), fasting glucose tolerant (6.1 mmol/L≤FPG<7.0 mmol/L) and diabetic (FPG≥7.0 mmol/L).29
Baseline characteristics
Baseline characteristics included sociodemographic data, HIV/AIDS-related status and comorbidities-related status. Sociodemographic data include current address, year of ART initiation, age, gender and marital status. HIV/AIDS-related status includes route of HIV transmission, CD4 cell counts (unit) and CD4/CD8. Comorbidities-related status includes body mass index (BMI), hyperlipidaemia status and coinfections of hepatitis B, hepatitis C and tuberculosis.
Follow-up cohort
The cohort period for this study spanned from January 2010 to December 2019. Patients who initiated ART during this period and met the inclusion and exclusion criteria were included in the analysis. For each patient, follow-up data within the first 12 months were used to assess the short-term effects of ART drugs on FPG, while data from up to 60 months of follow-up were used to analyse the long-term effects. The follow-up time for each patient was calculated as the difference between the last follow-up time and the baseline time, measured in months. For example, if a patient’s total follow-up period exceeded 1 year, their follow-up time for the short-term effect analysis was capped at 12 months. The average follow-up time for both short-term and long-term effect analyses was defined as the mean follow-up time across all patients in the respective periods. Patients were withdrawn from the follow-up cohort when their initial ART was adjusted, and only data collected prior to the regimen adjustment were included in the statistical analysis.
Matching
We used propensity score matching (PSM)30 with nearest neighbour algorithms to control for baseline confounders among patients in different initial ART regimen groups. Given that patient health status may vary over the follow-up period, matching was performed separately for the 12-month and 60-month follow-up datasets.
Propensity scores (PSs)31 were calculated using logistic regression, incorporating follow-up intervals (in months) and all baseline covariates into the PS model. The covariates included current address, year of ART initiation, age, gender, marital status, HIV transmission route, CD4 cell counts, CD4/CD8 ratio, BMI, hyperlipidaemia status, hepatitis B, hepatitis C and tuberculosis infection. Following PSM, the balance of covariates between matched groups was assessed using standardised difference (SDiff).32 As recommended in a review on PSM with a small sample size, an SDiff value less than 0.25 was considered indicative of balance.33 The characteristics of the pre-PSM and post-PSM samples were described according to different initial ART regimen groups. Continuous variables were reported as means and SDs, while categorical variables were reported as percentages.
Separate matching was conducted due to concerns that matching all three regimen groups simultaneously could result in a substantial loss of sample size. The primary focus of this study was on the NNRTIs-based regimen, which served as the reference group for comparison.
Statistical analysis
Since ART initiation, the mean value of follow-up FPG (measured in mmol/L) and the frequency of abnormal occurrences were calculated based on FPG measurements obtained during regular follow-up visits. A linear mixed model, estimated with restricted maximum likelihood, was used to investigate FPG trends within both the first year and 5 years since ART initiation. Random coefficient models were used under the assumption that individual patients have distinct baseline FPG levels and varying rates of FPG increase. Follow-up time (in months) was treated as the low level, coded as 0, 3, 6, 9 and 12 months (or 1, 2, 3, 4 and 5 years) to represent the longitudinal duration, while individuals were considered the high level. In the analysis within 5 years of ART, a quadratic term for time (ie, time2 was included, acknowledging that long-term FPG trends do not consistently increase linearly and tend to level off over time.31 For the first-year analysis, fixed effects included antiviral regimen, time and the interaction between antiviral regimen and time. For the 5-year analysis, the model additionally incorporated time2 and the interaction between the antiviral regimen and time.2 Further adjustments were made for important glycaemic influence factors among baseline covariates (model 1) and all baseline covariates (model 2). The estimated mean FPG values at 0, 3, 6, 9 and 12 months (or 1, 2, 3, 4 and 5 years) were graphically displayed, with the trend over time (per month/year) for each ART regimen calculated. Results are presented as regression coefficients (β) with 95% CIs. For baseline data, multiple imputation by chained equations was used to impute missing values. No imputation was performed for follow-up data. R V.3.4.5 was used for data analysis.
Patient and public involvement
None.
Results
Baseline and follow-up characteristics of before-and-after matching sample
Among the 1505 participants, 69.37% initiated ART with NNRTIs-based regimens, 13.82% with PIs-based regimens and 16.81% with INSTIs-based regimens. Statistically significant differences were observed in sociodemographic, AIDS-related and chronic disease-related characteristics among the three treatment groups (online supplemental table S1). Compared with the NNRTIs-based group, individuals in the PIs-based and INSTIs-based groups predominantly initiated ART after 2016 (56.51%, 67.31%, 93.28%, respectively) and were older (57.14, 60.75, 60.97 years, respectively). The proportion of women was higher in the PIs-based group, while the INSTIs-based group had a higher proportion of men (74.33%, 69.71%, 82.21%, respectively). Regarding AIDS-related conditions, the INSTIs-based group exhibited lower CD4 levels and higher rates of underweight. The average follow-up time within 1 year was 8.64 months, with the INSTIs-based group having the shortest duration (9.36, 8.06 and 6.18 months). The average follow-up time within 5 years was 22.55 months, again with the INSTIs-based group having the shortest duration (26.13, 20.50 and 9.47 months, respectively).
Supplemental material
To address the imbalance in baseline covariates and the follow-up duration among the three regimens, PSM was employed. The distribution of PSs before and after matching is illustrated in online supplemental figures S2 and S3. Following PSM, there were no statistically significant differences in baseline covariates and follow-up duration between the matched groups of NNRTIs-based and PIs-based regimens (205 participants each) and between the matched groups of NNRTIs-based and INSTIs-based regimens (190 participants each) within 1 year of ART, except for a lower proportion of patients with normal lipids at baseline in the PIs-based group compared with the NNRTIs-based group (55.12% vs 61.46%; (SDiff=0.293) (online supplemental table S2). The matching results between the matched groups of NNRTIs-based and PIs-based regimens (206 participants each) and between the matched groups of NNRTIs-based and INSTIs-based regimens (189 participants each) within 5 years of ART were consistent with those within 1 year of ART (online supplemental table S3).
Supplemental material
Additionally, there were no statistically significant differences in regimen adjustment between the matched groups, except for a higher proportion of patients in the INSTIs-based group who underwent regimen adjustments compared with the NNRTIs-based group within 1 year of treatment (47.89% vs 34.74%; SDiff=0.27).
Follow-up fasting glucose characteristics after ART
Table 1 presents the follow-up FPG levels after PSM within 1 year and 5 years of ART, respectively. In the matched groups of NNRTIs-based and PIs-based regimens, the prevalence of diabetes (FPG≥7.00 mmol/L) within 1 year of ART was 21.95% for NNRTIs-based group and 25.85% for PIs-based group. Within 5 years of ART, the prevalence of diabetes in the PIs-based group was higher than that in the NNRTIs-based group (31.55% vs 22.33%, SDiff=0.36). In the matched groups of NNRTIs-based and INSTIs-based regimens, the prevalence of diabetes within 1 year of ART was 24.74% for NNRTIs-based group and 26.32% for INSTIs-based group. Within 5 years of ART, the prevalence of diabetes was 20.40% for NNRTIs-based group and 26.98% for INSTIs-based group. No significant differences were observed in the mean value of follow-up FPG among the matched groups.
Follow-up FPG characteristics after propensity score matching
Fasting glucose trends after ART within 1 year of ART
Table 2 presents the results of multilevel linear regression analysis within the first year of ART. After adjusting for potential confounders, the PIs-based group did not show a higher FPG compared with the NNRTIs-based group (β=−0.04, 95% CI −0.17, 0.10). Time was positively associated with FPG (β=0.19, 95% CI 0.14 to 0.24), and the interaction term between the ART regimen group and time was negatively associated with FPG (β=−0.08, 95% CI −0.15 to –0.01). In the matched groups of NNRTIs-based and INSTIs-based regimens, the INSTIs-based group did not exhibit higher FPG compared with the NNRTIs group within 1 year of ART (β=−0.01, 95% CI −0.20, 0.18). Time was positively associated with FPG (β=0.23, 95% CI 0.16 to 0.31), while the interaction term between the ART regimen group and time was not statistically significant (β=0.06, 95% CI −0.05, 0.17).
Effect of different antiviral regimens on fasting glucose trends (within 1 year)
Figure 1 illustrates the FPG trends among different initial ART regimens within 1 year of ART. In the matched groups of NNRTIs-based and PIs-based regimens, FPG increased at a slower rate in the PIs-based group compared with the NNRTIs-based group. In the matched groups of NNRTIs-based and INSTIs-based regimens, the follow-up FPG trends were similar between the two groups.
The initial antiviral regimen differences in fasting glucose trends (within 1 year). FPG, fasting plasma glucose; INSTIs, integrase strand transfer inhibitor; NNRTIs, non-nucleoside reverse transcriptase inhibitors; PIs, protease inhibitors.
Fasting glucose trends after ART within 5 years of ART
Table 3 presents the results of multilevel linear regression analysis within 5 years of ART. After adjusting for potential confounders, in the matched groups of NNRTIs-based and PIs-based regimens, the PIs-based group did not show a higher FPG compared with the NNRTIs-based group at 5 years of ART (β=−0.04, 95% CI −0.16, 0.10). The time quadratic term was negatively associated with FPG (β=−0.11, 95% CI −0.14 to –0.09), while the interaction term between the ART regimen group and the time quadratic term was not statistically significant (β=0.02, 95% CI −0.01, 0.06). In the matched groups of NNRTIs-based and INSTIs-based regimens, the INSTIs-based group did not exhibit higher FPG than the NNRTIs-based group at 5 years of ART (β=−0.03, 95% CI −0.19, 0.12). The time quadratic term was negatively associated with FPG (β=−0.22, 95% CI −0.31 to –0.13), and the interaction term between the ART regimen group and the time quadratic term was positively associated with FPG (β=0.18, 95% CI 0.06, 0.31).
Effect of different antiviral regimens on fasting glucose trends (within 5 years)
Figure 2 depicts the FPG trends among different initial ART regimens within 5 years of ART. In the matched groups of NNRTIs-based and PIs-based regimens, the follow-up FPG trends were similar, displaying a linear upward trend within 1 year and stabilising at a higher level thereafter. In the matched groups of NNRTIs-based and INSTIs-based regimens, the follow-up FPG trends were similar within 2 years of ART. After 2 years, the FPG in the INSTIs-based group exhibited an unstable increase, with a wide 95% CI.
The initial antiviral regimen differences in fasting glucose trends (within 5 years). FPG, fasting plasma glucose; INSTIs, integrase strand transfer inhibitor; NNRTIs, non-nucleoside reverse transcriptase inhibitors; PIs, protease inhibitors.
Discussion
This study examined FPG trends from the initiation of ART among elderly Chinese PLWHA under three common ART regimens (NNRTIs based, PIs based and INSTIs based). By controlling for baseline confounders through PSM, the findings indicate that FPG increased in all ART regimens, with the most significant rise occurring within 1 year of ART initiation and stabilising thereafter. In PIs-based regimens, FPG increased more slowly than NNRTIs-based regimens during the early stage, but a higher prevalence of diabetes was observed in the long term. No statistically significant differences were found between INSTIs-based and NNRTIs-based regimens regarding diabetes prevalence and the trend of follow-up FPG over time.
The findings suggest that FPG tended to increase within the first year of ART and stabilise at a higher level later among elderly PLWHA. This observation aligns with previous studies indicating an increasing trend in FPG and insulin resistance, with significant changes occurring within the initial year of ART.31 Furthermore, the prevalence of diabetes among elderly PLWHA in this study ranged from 20.40% to 31.55% over 5 years of ART, surpassing the prevalence among the general elderly population in China (15.00%–21.40%).34 This finding is consistent with recent cohort studies of elderly PLWHA in Canada35 and the USA.7 Previous research has identified potential causes of increased FPG in PLWHA, including the effects of HIV infection36 and antiretroviral drugs.4 The sharpest increase in FPG within 1 year of ART could result from an alternate mechanism triggered on ART initiation, which directly inhibits glucose transporters, leading to insulin resistance.37 Besides, elevated FPG may be a marker of immune reconstitution inflammatory syndrome (IRIS) in PLWHA receiving ART,38 as the inflammatory response associated with IRIS can cause metabolic disturbances and increased insulin resistance.39 Redistribution of fat in the early stage of ART may also contribute to increased FPG.40 Consequently, this study suggests that PLWHA on ART, especially elderly individuals, should be closely monitored for FPG levels. FPG and glycosylated haemoglobin (HbA1c) testing should be integrated into their regular follow-up routine.
The study underscores that within 5 years of ART, the prevalence of diabetes in the PIs-based group was higher than that in the NNRTIs-based group. Conversely, the NNRTIs-based group exhibited a faster rise within the first year of ART. Consistent with these findings, Wang 41 found that the diabetes prevalence in PLWHA on a PIs-based regimen was 2.52 times higher (95% CI 1.09 to 5.83) than that in the NNRTIs-based regimen in a cross-sectional survey conducted in Yunan, China. A similar pattern was observed by Paengsai et al42 in Thailand, further supporting the results of the present study. However, a study in Nigeria43 found no statistically significant difference in diabetes prevalence between NNRTIs-based and PIs-based groups (RR 1.10, 95% CI 0.93 to 1.29), although it included only a small number of PLWHA on a PIs-based regimen (29%). The PIs-based group exhibited a slower rise in FPG during the early stages but a higher prevalence of diabetes in the long term. This divergence may be attributed to the distinct mechanisms of glucose metabolic effects between NNRTIs-based and PIs-based regimens. NNRTIs can increase the expression of TNF-α, IL-6 and IL-1β, causing insulin resistance through decreased adiponectin concentrations.44 In contrast, PIs have no acute detrimental effect on insulin resistance but can impair insulin signalling and directly affect glucose uptake with long-term exposure,45 leading to insuline resistance.46 Previous studies have indicated that PI-based regimens are the most common ART associated with insulin resistance.47 48 Consequently, this study emphasises the importance of comprehensive glucose monitoring for PIs users, with a focus on preventing chronic cardiovascular diseases associated with abnormal lipid metabolism and hyperglycaemia.
No statistical differences were observed between the INSTIs-based and NNRTIs-based groups in terms of diabetes prevalence and follow-up FPG. The increasing trend of FPG remained consistent within the first 2 years of ART, while FPG values in the INSTIs-based group showed an unstable increase after 2 years. This instability may be associated with the relatively small number of individuals using INSTIs-based regimens in this dataset, as well as the shorter duration of treatment. In China, patients with INSTIs-based regimens, such as DTG and RAL, often bear the cost themselves. Consequently, many PLWHA on INSTIs-based regimens may switch to NNRTIs-based regimens (provided for free) after stabilisation due to the financial burden. Moreover, INSTIs have only recently been introduced, with approximately 90% of PLWHA on INSTIs-based initial regimens initiating treatment between 2016 and 2019. Few of these patients (less than 34 individuals) were followed up for more than 3 years. As a result, the wide CIs and unstable changes observed in the analysis within 5 years of ART highlight the need for a larger sample size and longer-term follow-up. Research on hyperglycaemia related to INSTIs is limited. Lake JE et al49 found no statistical difference in FPG changes after 24 weeks of follow-up between PLWHA who switched to INSTIs-based regimens and those who continued on NNRTIs-based regimens, consistent with the findings of this study. Conversely, some studies concluded that short-term use of INSTIs does not affect hyperglycaemia compared with other antiretroviral drugs50 and may even be protective against type 1 diabetes.51 However, these conclusions were derived from studies conducted on HIV-negative adults. Bigoloni et al52 found an increase in FPG and a decrease in insulin secretion capacity among PLWHA on INSTIs-based regimens after 3 years of follow-up, but the mechanism of INSTIs’ effect on glucose metabolism remains unclear. Therefore, while an INSTIs-based regimen may pose a similar risk of elevated glucose as other regimens, further investigations with a larger sample size and longer-term follow-up are warranted.
This retrospective cohort study on FPG trends among elderly Chinese PLWHA has several notable strengths. First, it is the first study to explore both short-term and long-term FPG trends following ART initiation in elderly Chinese PLWHA. Furthermore, it specifically analyses the impact of three prevalent ART regimens, with a focus on the INSTIs-based regimen, on FPG trends. Additionally, the study employs PSM as a methodology to control for baseline confounders, which is a distinctive approach in contrast to other adjustment methods, rendering it more applicable to real-world clinical settings.
Despite its strengths, the study has some limitations that need to be acknowledged. First, the diagnostic criteria for diabetes employed in the study were relatively narrow, relying solely on FPG. The lack of additional indicators such as HbA1c, self-reported diagnosis and information on hypoglycaemic medication use53 might lead to an underestimation of diabetes prevalence, and the measurement error might result in non-differential misclassification of FPG. Second, the study exclusively focused on PLWHA without including a control group of normal elderly individuals. This omission makes it challenging to determine whether the observed increase in FPG is attributable to ageing, HIV infection, ART regimens or a combination of these factors. Furthermore, certain potential confounding factors affecting FPG, such as opportunistic infections, smoking, alcohol consumption, exercise and diet, were not considered in the study. The absence of data on some of these variables in the regular follow-up assessments among PLWHA on ART in China suggests that complete adjustment for confounding may not have been achieved. Additionally, the study did not include an assessment of insulin secretion function, which could limit the ability to explain the changes in FPG mechanistically.
Conclusion
To summarise, our study affirms that FPG increases in both groups among elderly Chinese PLWHA, with the most significant rise occurring within the first year of ART and subsequently plateauing. Distinctive effects of different regimens on FPG trends were observed, with the PIs-based group experiencing a slower FPG rise in the early stage but a higher prevalence of diabetes in the long term. No statistical differences were found between the INSTIs-based regimen and NNRTIs-based regimen regarding diabetes prevalence and follow-up FPG trend. Based on these findings, we recommend emphasising the importance of regular glucose monitoring among elderly PLWHA. Additionally, once diabetes is diagnosed, initiating chronic disease management for this population becomes crucial. Furthermore, further research is needed to explore the impact of hyperglycaemia on ART effectiveness and the life expectancy of PLWHA.
Data availability statement
Data are available on reasonable request. The data that support the findings of this study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The study was approved by the ethics committee of the School of Public Health, Sun Yat-sen University (Reference No.: 2019-139). The data for our study were obtained from the electronic medical database after deidentification, without patient-identifiable information. Informed consent from patients was not required.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
MS and JZ contributed equally.
Contributors The guarantor of the study is JG; accepted full responsibility for the finished work and the conduct of the study, had access to the data and controlled the decision to publish. JG and LL had conceived the research questions, assembled the team of collaborators and conducted quality control. MC, XC, CL, HZ, PD and QL coordinated the fieldwork and collected data. MS and JZ conducted the statistical analysis and drafted the manuscript, revised the manuscript and finalised the manuscript. JG, LL and WC revised the manuscript, gave scientific comments. All authors assisted in data collection, data interpretation and gave comments on the intellectual content of the manuscript. All authors read and approved the final manuscript.
Funding This study was supported by the National Science and Technology Major Project of China (grant ID 2018ZX10715004).
Disclaimer The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.