Over the past 24 hours I have seen someone arguing we need to reduce the intake of Deuterium and someone else trying to sell a product because it has Deuterium replacing Protium (ie Hydrogen without any neutrons).
I don’t think either warrant being linked to. It is, however, interesting to see this. I think it probably does make some difference at some stage. One of the arguments particularly is that shoving a neutron and a proton through the proton pump in the mitochondria causes a problem. That would not necessarily be surprising, but I have not tried to do the maths to see if the mass of the proton matters that much in the circumstances compared to the charge.
I thought I would ask chatGPT and the answer is perhaps more interesting than what is being said elsewhere:
Yes. There is real research, but it is much stronger for Complex IV / cytochrome c oxidase and model proton pumps than for a clean head-to-head “deuterons through the whole mitochondrial ETC versus protons” experiment.
The basic finding is: deuterium can pass through proton-transfer systems because D⁺ has the same charge as H⁺, but it usually moves more slowly and changes hydrogen-bond/proton-transfer kinetics. This is measured as a deuterium kinetic isotope effect: rates in H₂O are compared with rates in D₂O, or mixed H₂O/D₂O.
Best direct evidence: cytochrome c oxidase / Complex IV
Cytochrome c oxidase is the terminal ETC enzyme and a redox-driven proton pump. Several studies have used D₂O to probe proton-pumping steps.
A key 2008 study, “Deuterium isotope effect of proton pumping in cytochrome c oxidase,” examined pH dependence and deuterium isotope effects in wild-type and mutant oxidases. It concluded that some reaction steps are controlled by proton-transfer events, including pumped-proton movement through parts of the protein involving Glu286 and Arg481. In their description, the P³→F³ transition was governed by direct proton transfer to the catalytic site, while the F³→O⁴ transition involved coupled electron transfer and pumped-proton transfer with distinct isotope effects. (DIVA Portal)
An earlier 2000 paper used a proton-inventory approach on Rhodobacter sphaeroides cytochrome c oxidase. It found a large kinetic deuterium isotope effect of about 7 in the F→O step, which is coupled to proton pumping. Importantly, the analysis suggested that despite a long proton-transfer path, the rate-limiting isotope-sensitive step was localized to a single protonatable site, proposed to be Glu286 in the D-pathway. (PubMed)
There is also work on bacterial ba₃ cytochrome c oxidase from Thermus thermophilus, where investigators measured pH dependence and deuterium isotope effects to understand its lower proton-pumping stoichiometry compared with more mitochondrial-like oxidases. (PubMed)
Whole mitochondria / oxidative phosphorylation
Older mitochondrial experiments also exist. A 1982 study on heavy water and mitochondrial respiration/oxidative phosphorylation reported that D₂O altered both respiration and oxidative phosphorylation. The authors interpreted the data as showing both “solvent effects” and “isotope effects”; they suggested the isotope effect was primarily responsible for uncoupling of oxidative phosphorylation, whereas inhibition of ADP-stimulated state 3 respiration may reflect broader solvent effects of heavy water. (Springer Link)
A still older 1966 paper, “Observations on the effect of D₂O on energy-linked reactions of the mitochondrion,” found that D₂O inhibited mitochondrial respiration and energy-linked transhydrogenation; uncouplers reduced some of the inhibition. The authors inferred that D₂O affected processes involving generation, transfer, or utilization of the energy-linked intermediate rather than simply inhibiting ordinary electron flow. (PubMed)
These older studies are useful, but they used high D₂O substitution and predate modern ETC structural/kinetic resolution. They do not prove that physiological/natural-abundance deuterium strongly impairs mitochondria.
Complex I and Complex III
For Complex I, the literature is mostly about proton-coupled electron transfer, long-range conformational coupling, quinone reduction, and four-proton translocation per NADH, but I did not find the same kind of clean experimental H₂O-vs-D₂O “deuteron pumping” literature as for cytochrome c oxidase. A recent review summarizes that Complex I transfers four protons per NADH oxidation and that its mechanism is still being refined through structural, biochemical, kinetic, and molecular dynamics studies. (MDPI)
For Complex III / cytochrome bc₁, there is theoretical and computational work on proton-coupled electron transfer. One 2021 study modeled the primary PCET reaction in bc₁ and found deuterium kinetic isotope effects in the expected range for proton transfer; related analysis estimated KIE values around 3–4 for hydrogen/deuterium transfer in that setting. (PMC)
ATP synthase
ATP synthase is not an ETC proton pump, but it is the main machine that lets protons back through to make ATP. There is a 2007 paper, “Biological effects of deuteronation: ATP synthase as an example,” arguing that deuteration could interfere with ATP synthase and proton-coupled electron transport, potentially increasing electron leakage and ROS. However, that paper is more mechanistic/speculative than a direct demonstration of deuterons jamming mitochondrial ATP synthase at natural abundance. (PMC)
Practical interpretation
So the answer is:
Yes, there is research showing that replacing H with D slows or alters proton-transfer steps in ETC-related proton pumps, especially cytochrome c oxidase. The strongest effects are seen in defined proton-transfer reactions, with KIEs ranging from modest values to about 7 in particular Complex IV steps.
But:
No, I would not say there is strong evidence that ordinary biological deuterium levels cause a major ETC bottleneck in normal mitochondria. Most direct effects are measured in high-D₂O systems or purified enzymes, and D₂O changes water properties as well as isotope mass. The clean conclusion is kinetic: deuterons are chemically usable but slower in proton-coupled transfer chains, especially where tunnelling, hydrogen-bond rearrangement, or protonatable gating residues are rate-limiting.