| |
Impact of glycemic status on longitudinal cognitive performance in men with and without HIV infection
|
| |
| |
Download the PDF here
"Overall, 900 MLWH and 1149 men without HIV were included. MLWH had significantly more person-visits with impaired fasting glucose (52.1 vs. 47.9%) and controlled diabetes mellitus (58.2 vs. 41.8%) than men without HIV (P < 0.05). Compared with men with normal glucose, men with diabetes mellitus had significantly poorer performance on psychomotor speed, executive function and verbal learning (all P < 0.05). There was no difference in cognition by HIV aerostats.......Our data support the hypothesis that abnormalities in glucose metabolism are associated with neuropsychological dysfunction among MLWH and without HIV. Greater cognitive impairment was correlated with poor glycemic control and longer duration of T2DM. In the current study, T2DM was associated with cognitive performance on all the domains. IFG [impaired fasting glucose] affected psychomotor speed/attention but had a less profound impact on cognition compared with T2DM. Of the nine cognitive tests, the TMT and SDMT appeared to be affected most, demonstrating an increased risk of psychomotor inefficiency, attention deficits and visuospatial tracking problems. Moreover, our results indicate a significant association between longer T2DM duration and cognitive dysfunction, suggesting that optimization of glycemic control and management of vascular disease risk factors may improve cognitive performance.
In PLWH, the prevalence of T2DM is up to 14% [10], compared with 9.4% in the US population [27] and associations between glycemic status and cognition have also been reported [28]. For example, in a prospective cohort enriched with older HIV-infected individuals, diabetes was associated with a five-fold increased odds of HIV-associated dementia compared with people without diabetes and this association remained significant after multivariable adjustment [29]. In the Women's Interagency HIV Study, greater insulin resistance was associated with poorer performance on Symbol Digit Modalities Test (SDMT), on Stroop Color-Naming and on Stroop Interference in unadjusted analyses [30]
Among men with diabetes mellitus, disease duration was further categorized as the percentage of time with diabetes mellitus being uncontrolled over the study period (0, 0 to <30, 30 to <70, 70 to <100, 100%), according to quintiles. The impact of glycemic status on cognition was predominantly seen among men who had uncontrolled T2DM more than 70% of the time (Table 4). Specifically, this effect is equivalent to 16.5 years of aging on psychomotor speed and attention (tested by TMT-A, Coef. = -0.726, P = 0.001) as well as 13.4 years of aging on executive function (tested by TMT-B, Coef. = -0.676, P = 0.003)."
--------------------------------
AIDS Aug 24 2018 - Yang, Jingyana,b; Jacobson, Lisa P.c; Becker, James T.d; Levine, Andrewe; Martin, Eileen M.f; Munro, Cynthia A.g; Palella, Frank J.h; Lake, Jordan E.i; Sacktor, Ned C.g; Brown, Todd T.j
Objectives: To determine the relationship between glycemic status and cognitive performance in men living with HIV (MLWH) and without HIV infection.
Design: A prospective HIV/AIDS cohort study in four US cities between 1999 and 2016.
Methods: Glycemic status was categorized as normal glucose, impaired fasting glucose, controlled diabetes mellitus and uncontrolled diabetes mellitus at each semiannual visit. Cognitive performance was evaluated using nine neuropsychological tests which measure attention, constructional ability, verbal learning, executive functioning, memory and psychomotor speed. Linear mixed models were used to assess the association between glycemic status and cognition.
Results: Overall, 900 MLWH and 1149 men without HIV were included. MLWH had significantly more person-visits with impaired fasting glucose (52.1 vs. 47.9%) and controlled diabetes mellitus (58.2 vs. 41.8%) than men without HIV (P < 0.05). Compared with men with normal glucose, men with diabetes mellitus had significantly poorer performance on psychomotor speed, executive function and verbal learning (all P < 0.05). There was no difference in cognition by HIV serostatus. The largest effect was observed in individuals with uncontrolled diabetes mellitus throughout the study period, equivalent to 16.5 and 13.4 years of aging on psychomotor speed and executive function, respectively, the effect of which remained significant after adjusting for HIV-related risk factors. Lower CD4+ nadir was also associated with worse cognitive performance.
Moreover, men with nadir CD4+ T-cell counts below 200 cells/μl (Coef. = -0.157, P = 0.006) had significantly worse cognitive performance on TMT-A than men with counts above 500 cells/μl. However, neither having undetectable HIV-1 RNA nor length of cART showed a significant relationship with cognition.
Among men with diabetes mellitus, disease duration was further categorized as the percentage of time with diabetes mellitusbeing uncontrolled over the study period (0, 0 to <30, 30 to <70, 70 to <100, 100%), according to quintiles. The impact of glycemic status on cognition was predominantly seen among men who had uncontrolled T2DM more than 70% of the time (Table 4). Specifically, this effect is equivalent to 16.5 years of aging on psychomotor speed and attention (tested by TMT-A, Coef. = -0.726, P = 0.001) as well as 13.4 years of aging on executive function (tested by TMT-B, Coef. = -0.676, P = 0.003).
Conclusion: Abnormalities in glucose metabolism were more common among MLWH than men without HIV and were related to impaired cognitive performance. Metabolic status, along with advanced age and previous immunosuppression, may be important predictors of cognition in the modern antiretroviral therapy era.
Introduction
The use of combination antiretroviral therapy (cART) has transformed HIV infection into a chronic illness and has markedly extended survival among people living with HIV (PLWH), often into old age [1]. However, the effect of cART on HIV-associated cognitive impairment is modest [2-4]. The prevalence of cognitive dysfunction among PLWH continues to be high [2,5-7], and the cause is not clear. Hypotheses include, but are not limited to, immune system integrity and inadequate control of persistent central nervous system HIV viral replication [8,9]. In addition, chronic non-AIDS comorbidities such as impaired glucose regulation, which is common among PLWH [10], may also play an important role in the pathogenesis of cognitive dysfunction [11].
In the general population, type 2 diabetes mellitus (T2DM) has been associated with impairment in numerous cognitive domains [12-17]. T2DM predominantly affects hippocampal-based declarative memory performance and attention, but psychomotor speed and executive functioning are also affected [13,18]. The magnitude of cognitive decline depends largely on the duration of T2DM, glycemic control and antidiabetic agent use [19-25]. It has been reported that a 1% higher glycated hemoglobin (HbA1c) value is associated with significantly lower performance on psychomotor speed executive functioning and memory [26].
In PLWH, the prevalence of T2DM is up to 14% [10], compared with 9.4% in the US population [27] and associations between glycemic status and cognition have also been reported [28]. For example, in a prospective cohort enriched with older HIV-infected individuals, diabetes was associated with a five-fold increased odds of HIV-associated dementia compared with people without diabetes and this association remained significant after multivariable adjustment [29]. In the Women's Interagency HIV Study, greater insulin resistance was associated with poorer performance on Symbol Digit Modalities Test (SDMT), on Stroop Color-Naming and on Stroop Interference in unadjusted analyses [30]. Similar results were found using adjusted models in a more recent study of the same cohort [31]. In addition, the authors reported that women with T2DM had worse performance on measure of psychomotor speed and manual dexterity testing than women without T2DM [31].
Despite evidence suggesting that glycemic status is associated with cognition in PLWH and without HIV, previous cross-sectional studies were limited in their ability to investigate longitudinal glycemic status with regard to trajectories of cognitive function over time [29,30,32]. In some cases, assessment of cognitive domains was not comprehensive [33,34] and glycemic status was not adequately characterized [30,31]. The purpose of the current study was to determine the longitudinal relationship between glycemic status and cognition using data from Multicenter AIDS Cohort Study (MACS), one of the largest ongoing cohorts of the natural and treated history of HIV among gay/bisexual men. The MACS has followed a neuropsychological subcohort for over 29 years, allowing for determination of the neurological and neurocognitive consequences of HIV over time [35]. We hypothesized that: glycemic status is associated with cognitive performance among men living with HIV (MLWH) and without HIV; the magnitude of this effect varies by cognitive domains; the duration of T2DM is associated with accelerated cognitive decline; and HIV-specific risk factors affect cognition in MLWH.
Discussion
Our data support the hypothesis that abnormalities in glucose metabolism are associated with neuropsychological dysfunction among MLWH and without HIV. Greater cognitive impairment was correlated with poor glycemic control and longer duration of T2DM. In the current study, T2DM was associated with cognitive performance on all the domains. IFG affected psychomotor speed/attention but had a less profound impact on cognitioncompared with T2DM. Of the nine cognitive tests, the TMT and SDMT appeared to be affected most, demonstrating an increased risk of psychomotor inefficiency, attention deficits and visuospatial tracking problems. Moreover, our results indicate a significant association between longer T2DM duration and cognitive dysfunction, suggesting that optimization of glycemic control and management of vascular disease risk factors may improve cognitive performance.
Although the affected cognitive domains differ across studies, the magnitude of observed impairment reported in people without HIV in general support our findings [50-56].
Specifically, a systematic review reported that memory, processing speed and cognitive flexibility were most consistently affected by diabetes across studies [57]. Among PLWH, increasing insulin resistance negatively affected attention and recognition among women after multivariable adjustment [31].
We also found that duration of T2DM was a strong predictor of cognitive performance, which is consistent with previous studies [20,21,58,59]. A population-based case-control study reported that for participants with a duration of diabetes mellitus for 10 years or longer, their odds of having mild cognitive impairment were significantly elevated, 1.8 fold greater, compared with persons with duration of diabetes mellitus less than 10 years [59]. Similarly, community-dwelling elderly individuals in the highest HbA1c tertile were found to have significantly poorer overall cognition, semantic categorization and executive functioning, whereas such an association was not revealed in the lowest HbA1c tertile, which highlighted the importance of glycemic control in reducing the risk of cognitive decline [22].
In the current study, IFG appeared to be less related than T2DM to cognition, as only mild impairment was observed on psychomotor speed among people with IFG, which has been seen in other studies [60,61]. However, associations reported between IFG and cognition have varied in the literature [62-65]. A pooled cohort analysis that included 8447 individuals from the general population reported that neither elevated fasting glucose levels nor insulin resistance was associated with executive function and memory in older people without a history of diabetes mellitus, whereas individuals with diabetes mellitus were noted to have cognitive decline [66]. In contrast, an analysis from a 4-year study of postmenopausal women reported that IFG increased the risk of developing cognitive dysfunction approximately two fold [67]. Hence, future studies are warranted to investigate whether there are thresholds for effects of dysglycemia on cognition.
We found that HIV stage of disease-related factors such as CD4+ nadir were more informative than HIV serostatus regarding cognition in the cART era. A prospective cohort study that consisted of 1525 PLWH demonstrated that the odds of neurocognitive impairment were reduced by 10% for every 5-U increase in square-root CD4+ nadir [68]. Heaton et al. [69] compared the incidence and characteristics of HIV-associated neurocognitive disorders before and during the cART era and found that low nadir CD4+ was associated with neurocognitive dysfunction in both eras, whereas duration of HIV infection and viral suppression predicted impairment only before the cART era, a finding corroborative of our data. These findings support the importance of early cART initiation for disease avoidance, which may help preserve cognition in the long term.
Impaired glucose regulation plays a crucial role in the pathogenesis of cognitive decline, along with vascular risk factors, socioeconomic and lifestyle factors and depression. Several studies have demonstrated that prediabetes mellitus stage leads to a higher incidence of cognitive impairment [70,71] or dementia [58,72] and T2DM increases the incidence of vascular [73-75] and Alzheimer dementia [75-77]. Increased amyloid processing in the setting of hyperglycemia has been proposed as an underlying mechanism [78]. Other potential mechanisms include increased formation of advanced glycation end products, diacylglycerol activation of protein kinase C and increased glucose shunting in the hexosamine pathway of hyperglycemia, which have been shown to alter function in other organs, but whether similar effects occur in the brain is unclear [79-81].
Multifactorial treatment approaches that target shared causes of diverse vascular disease risk factors, together with lifestyle changes (i.e. weight control, nutrition education) [82,83], may be necessary to prevent cognitive decline. A review has recommended that lifestyle intervention in middle-aged persons should be implemented along with medication for glycemic control to prevent cognitive decline [84]. For persons with a long-standing T2DM, maintaining good glycemic control may modulate the effect of diabetes mellitus on cognition.
To our knowledge, this report represents the largest and most comprehensive evaluation to date of the relationship between glycemic status and cognitive function among MLWH and without HIV. Strengths of our study are its prospective design and long follow-up period with repeated measurement of cognitive function using a sensitive test battery and repeated measures of covariates that allowed us to adjust for a broad array of potential confounders. Our study had several limitations. First, the MACS cohort includes only male participants; additional studies are needed to determine if similar results are observed in women with and without HIV. As multiple, inter-related factors contribute to cognitive performance, understanding the causal pathway of T2DM to cognitive impairment warrants future studies. Further, survivorship bias may have existed in the cohort; some MACS participants have been studied for more than 20 years, whereas those who suffered from more severe HIV-related comorbidities may have been more likely to have dropped out. Lastly, the prevalence of T2DM was high in our cohort, possibly due to extensive ART treatment experience, especially exposure to toxic older ART drugs, and more prolonged untreated HIV infection prior to ART initiation. The generalizability of our results to other cohorts with less treatment experience and shorter HIV infection duration is unclear.
In conclusion, abnormal glucose regulation is associated with worse cognitive performance among both MLWH and without HIV. Glucose-mediated processes and duration of diabetic disease appear to be crucial, and therefore optimization of glycemic control appears to be an important goal for the maintenance of cognitive performance, along with control of other chronic vascular disease risk factors and lifestyle changes. The optimal timing of interventions aimed at improvements in glycemic control cannot be ascertained with precision from these data, but such interventions may be more effective when initiated in midlife than at a more advanced age.
|
|
| |
| |
|
|
|
