That some aging-associated phenotypes may be reversible is an emerging theme in contemporary aging research. Gomes et al. report that age-associated oxidative phosphorylation (OXPHOS) defects in murine skeletal muscle are biphasic. In the first phase, OXPHOS is decreased because of reduced expression of mitochondrially encoded genes. Treatment of moderately old mice (first-phase OXPHOS defects) with nicotinamide adenine dinucleotide (NAD⁺) precursor nicotinamide mononucleotide (NMN) for 1 week restores oxidative phosphorylation activity and other markers of mitochondrial function in skeletal muscle. However, muscle strength is not restored. In very old animals (second-phase OXPHOS defects), expression of OXPHOS genes from both the nucleus and mitochondria is reduced and mitochondrial DNA integrity is diminished. Gomes et al. propose a model linking decreased NAD⁺ to loss of nuclear SIRT1 activity to stabilization of the hypoxia-associated transcription factor hypoxia-inducible factor 1-alpha (HIF-1a). HIF-1a promotes an hypoxic-like (Warburg effect) state in the cell. The HIF-1a protein interacts with c-Myc, decreasing c-Myc-regulated transcription of the key mitochondrial regulator mitochondrial transcription factor A (TFAM). Low levels of TFAM lead to first-phase OXPHOS dysfunction. The transition to irreversible phase 2 dysfunction remains to be characterized, but may be related to increased reactive oxygen species (ROS) production. This model suggests that intervention in mitochondrial aging may be possible using appropriate NAD⁺ precursors such as nicotinamide riboside. Restoring NAD⁺ levels may be beneficial throughout the organism. For example, aging-associated disturbances in circadian rhythm are linked to diminished SIRT1 activity, and loss of hematopoietic stem cell function to reduced SIRT3. Work to elucidate other biphasic aging mechanisms is strongly encouraged.