Several factors were associated with fatigue. Consistent with previous studies, depressive mood showed a strong relationship with fatigue, and depressive mood was also a predictor of future fatigue status. Fatigue and depression share common symptomatology and potentially common CNS pathways which inevitably lead to an overlap of these two conditions.
We also observed that younger and more educated patients tend to report more fatigue symptoms than older and less-educated patients. One possible explanation is that younger and more educated subjects have higher functional expectations than older and less-educated subjects and therefore may be more affected by fatigue symptoms.
While fatigued and non-fatigued subjects did not differ at baseline in terms of cognitive performance, baseline performance on the Stroop test, which is designed to assess frontal lobe functions, was a predictor of fatigue. Furthermore, patients with higher MSK staging reported more concurrent fatigue during the study period. However, it should be noted that MSK staging and overall cognitive performance as assessed by NPZ summary scores differ as MSK incorporates both cognitive and functional evaluations. In this regard, the association between fatigue and MSK staging is consistent with the association between decline in functional activity, as measured by the Karnofsky score, and fatigue. Karnofsky score also predicted future fatigue status.
Another potential cause for fatigue might be use of antiretroviral drugs. In a large cohort of HIV-negative and HIV-positive women, Silverberg et al. (2004)
showed that more HIV-infected women reported fatigue compared to HIV negative women, but there was no substantial difference in fatigue between HIV-infected HAART-naïve and HAART-stable patients. However, there may be a mechanistic link with at least one class of antiretroviral drugs, nucleoside reverse transcriptase inhibitors (NRTIs). NRTIs have been associated with mitochondrial toxicities as indicated by the occurrence of lactic acidosis, hepatic steatosis, pancreatitis, myopathy, and neuropathy (Carr and Cooper 2000
; Dalakas et al. 2001
; Brinkman et al. 1999
; Kakuda et al. 1999
). Protease inhibitors can also be associated with mitochondrial abnormalities (Kim et al. 2007
). Increased fatigue is a typical presentation of mitochondrial disorders and may be present even when there is minimal clinical evidence in the affected organ (Cote et al. 2002
; Delgado et al. 2001
; Carr et al. 2000
). NRTIs are known to inhibit mitochondrial DNA polymerase-γ (pol-γ) leading to depletion of mitochondrial DNA (mtDNA). Interestingly, decreased mtDNA is also present in antiretroviral-naïve subjects (Cote et al. 2002
) suggesting that mitochondrial toxicity is associated with HIV infection itself.
In this study, the proportion of subjects using NRTIs was very high, and did not differ between fatigued and non-fatigued groups.
Fatigue has also been associated with cytokine dysregulation. For example, some investigations in cancer and multiple sclerosis (MS) have implicated pro-inflammatory cytokines such as IL-1, IL-6, and TNF-α in the pathogenesis of fatigue (Collado-Hidalgo et al. 2006
; Kos et al. 2008
; Heesen et al. 2006
). Cytokine dysregulation is a common feature of HIV infection and its associated neurological complications (McArthur et al. 2005
; Schifitto et al. 2005
; Jones and Power 2006
). However, patients on a stable antiretroviral regimen usually have only mild evidence of increased products of immune activation (McArthur et al. 2005
; Clifford et al. 2002
). We did not assess cytokines but patients were on a stable antiretroviral regimen and plasma and CSF HIV viral load, and CD4 count did not differ among fatigued and non-fatigued patients.
There is mounting evidence in other neurological disorders, such as MS, that disruption of the striatal–cortical or striatal–thalamic–cortical circuitry will predispose patients to fatigue (Chaudhuri and Behan 2000
). A recent positron-emission tomography (PET) imaging study (Roelcke et al. 1997
), using 18
F-fluorodeoxyglucose, revealed significant metabolic alteration in the lateral and medial prefrontal cortex, in the premotor cortex, putamen, and in the right supplementary motor area of fatigued MS patients. There were also metabolic changes in the white matter extending from the rostral putamen toward the lateral head of the caudate nucleus. These findings are supported by a recent fMRI study (Deluca et al. 2008
Additional support for basal ganglia dysregulation in fatigue states comes from fatigue associated with interferon-alpha treatment (Capuron et al. 2007
Our findings that markers of energy metabolism (Cr concentration) and glutaminergic transmission (lower concentration of Glx) in the basal ganglia are associated with fatigue are in line with the above observations in other diseases. The Cr peak observed on 1
H MRS reflects the sum of creatine plus phosphocreatine, and the Glx peak represents the sum of glutamate plus glutamine. Since the high-energy phosphate metabolism, as well as the synthesis of glutamate via the TCA cycle, involves the mitochondria, reductions in the Cr and Glx concentrations suggest lower cellular energy levels in the basal ganglia of subjects with fatigue. One prior study found lower concentrations of basal ganglia Cr in a group of subjects with, but not in those without HIV-associated dementia (Chang et al. 2002
), but the effect of fatigue was not evaluated specifically.
In summary, this study emphasizes that fatigue is common and persistent in HIV-positive subjects, and is associated with decreased functioning. A neuronal circuitry that involves striatal–cortical pathways may play an important role in HIV-associated fatigue, and may be amenable to therapeutic intervention. In this regard, therapeutic advances directed at fatigue may also benefit cognition and mood. A recent trial of modafinil that ameliorated fatigue and also improved mood and cognitive performance in HIV-infected individuals is consistent with this concept. (Rabkin et al. 2004
; McElhiney et al. 2009
; Rabkin et al. 2010