Frailty Is Associated With Insulin Resistance in Chronic Human Immunodeficiency Virus Infection

To the Editor—We read with interest the article by Kelly et al [1] who described frailty as an independent risk factor for chronic diseases and mortality in persons living with human immunodeficiency virus (HIV; PLWH). Specifically, the authors observed that slow gait speed (GS) was associated with diabetes mellitus (DM). We add to their findings that measures of abnormal glucose metabolism prior to DM development can be linked to prefrail and frail conditions.

Data from the Hawaii Aging with HIV-Cardiovascular Disease study [2] were analyzed to assess the association of frailty with insulin resistance (IR). Frailty was characterized using Fried’s frailty phenotype (frail ≥3 criteria; prefrail = 1–2; nonfrail: 0) [3]. To assess IR, the homeostatic model assessment of IR (HOMA-IR) and oral disposition index (DIo) were calculated. DIo is also used to assess β-cell function [4].

Of 73 PLWH in our cohort, median age was 51 years and median CD4 count was 466 cells/mm³. The majority were male (89%) with plasma HIV RNA <50 copies/mL (84%). We found that 42 participants were nonfrail, 26 were prefrail, and 5 were frail. We ran separate analysis of variance models with a Tukey adjustment to assess the association between handgrip strength (HGS), GS, and IR between frailty groups. Frail and prefrail groups demonstrated significantly lower HGS compared with nonfrail participants (Table 1). Frail participants showed significantly slower GS compared with prefrail and nonfrail participants. HOMA-IR was significantly increased in frail participants compared with nonfrail participants. DIo was significantly lower in frail participants compared with nonfrail participants. DIo correlated with GS (rho = –0.245, P = .036) but not with HGS. Given the limited number of frail participants, we combined the prefrail and frail participants and performed sensitivity analyses on the above results to determine which associations still held. In our comparison of frail and nonfrail participants, we found that HGS and DIo remained significantly lower and GS remained significantly higher.

Table 1.

Characteristics of Persons Living With Human Immunodeficiency Virus Stratified by Frailty Score

Characteristic	Nonfrail (n = 42)	Prefrail (n = 26)	Frail (n = 5)
Frail characteristic
Average grip strength, kg (IQR)	41 (36–48)	30 (27–38)a	29 (26–30)a
Average gait speed, seconds (IQR)	3.3 (3.0–3.7)	3.7 (3.5–4.2)a	5.2 (4.5–5.6)a,b
Low physical activity, n (%)c,d	0 (0)	10 (38)	5 (100)
Exhaustion, n (%)e,d	0 (0)	6 (23)	5 (100)
≥10 lb or ≥5% unintentional body weight loss, n (%)d	0 (0)	3 (11)	1 (20)
Glucose metabolism
Insulin sensitivity (IQR)	0.02 (0.01–.03)	0.02 (0.01–.03)	0.008 (0.007–.017)a,b
Homeostatic model assessment of insulin resistance (IQR)	1.44 (0.88–2.14)	1.48 (0.89–1.93)	4.06 (1.78–4.35)a
Oral disposition index (IQR)	2.90 (1.79–3.90)	1.91 (1.45–3.43)	1.35 (1.14–1.61)a

Abbreviation: IQR, interquartile range.

^a P ≤ .05 with nonfrail as the reference variable using Tukey adjustment.

^b P ≤ .05 with prefrail as the reference variable using Tukey adjustment

^cLow physical activity defined as health limiting vigorous activities.

^dProportions of low physical activity, exhaustion, and weight loss were significantly different between groups by exact χ² test.

^eExhaustion defined as having little energy 3 or more days per week.

Frailty reflects reduced mobility related to poor skeletal muscle glucose uptake, suboptimal glycemic control, and IR [5]. Of Fried’s phenotypes, HGS and GS are the only parameters that provide objective measures of frailty. In clinical practice, implementing HGS and GS may also determine the degree of sarcopenia [6]. In older populations, guidelines to improve sarcopenia [7] have suggested specific measures such as prescription of resistance-based physical activity and protein supplementation. Similar interventional methods should be considered for frail and sarcopenic PLWH. The AIDS Clinical Trials Group A5322 study found that the development of slow gait was associated with baseline hemoglobin A1c [8]. A limitation of our study was that not all participants’ frailty measurements were performed at the same time as the IR measurement. Frailty and IR assessment were within 2 years of each other. Additionally, to address confounding by diabetes, participants with DM were removed from the analyses. Despite these limitations, frailty and lower HGS were significantly associated with worsening IR.

In summary, we found that frailty was significantly associated with reduced insulin sensitivity and increased IR. The role of chronic inflammation or persistent immune activation in frailty needs to be further explored as well as the utility of objective measures such as HGS and GS.

Notes

Acknowledgments. The authors thank the clinical and laboratory staff of the Hawaii Center for AIDS, University of Hawaii, and the many patients of the Hawaii Aging with HIV cohort study who made this study possible.

Financial support. This work was supported by National Institutes of Health grants U54 NS43049, R01HL095135, U54MD007584 and U54MD007601.

Potential conflicts of interest. The authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.