Ketone-Based Metabolic Therapy: Is Increased NAD+ a Primary Mechanism?

The researchers tested whether a ketogenic diet (KD) increases the ratio of NAD+ to NADH (the oxidized vs. reduced form of nicotinamide adenine dinucleotide) in the brain. They compared rats fed a standard chow diet (control) against rats fed a 6:1 ratio ketogenic diet (6 parts fat to 1 part protein plus carbohydrates) for either 2 days or 3 weeks. The primary outcome was the NAD+/NADH ratio in two brain regions: the hippocampus and the frontal cortex. Secondary outcomes included blood levels of the ketone body β-hydroxybutyrate (β-OHB). The paper is primarily a perspective article that presents original pilot data alongside a review of existing literature, proposing a mechanistic hypothesis.

20 adult male Sprague-Dawley rats, aged 9–14 weeks

Housed at Trinity College (Hartford, CT, USA)

No prior interventions, healthy, normal weight

No human subjects were studied in the original data; the paper reviews human studies from other published work (e.g., epileptic patients, Alzheimer's patients, Parkinson's patients)

**Blood ketones:** Trunk serum was collected after sacrifice and analyzed for β-hydroxybutyrate (β-OHB) concentration, reported in mmol/L

**Brain NAD+/NADH ratio:** Hippocampus and frontal cortex tissue were dissected after sacrifice. NAD+ and NADH were quantified using a commercial enzymatic cycling assay (NAD/NADH Quantification Kit, BioVision). The ratio was calculated as NAD+ divided by NADH

**Diet composition:** Standard chow (Purina 5001) vs. ketogenic diet (#F3666, Bio-Serv) with a 6:1 fat-to-(protein+carbohydrate) ratio

**Duration:** Two time points — 2 days and 3 weeks of diet exposure

**Study design:** This is a perspective article that includes original pilot data from a controlled laboratory experiment. The authors fed rats either a ketogenic diet or a standard chow diet for two durations (2 days or 3 weeks), then measured blood ketones and brain NAD ratios post-mortem.

**Randomisation and blinding:** The paper does not explicitly state whether rats were randomly assigned to diet groups, nor whether the researchers performing the NAD assays were blinded to diet condition. This is a significant methodological weakness. In animal studies, lack of blinding can introduce bias in tissue processing, assay timing, and data interpretation.

**Sample sizes:** The control group (chow diet) had n=8 rats. The 2-day KD group had n=3 rats. The 3-week KD group had n=4 rats. These are very small sample sizes, particularly for the KD groups. With only 3–4 animals per treatment group, the statistical power to detect true effects is low, and the risk of false positives (or false negatives) is high. The authors do not report a power analysis.

**Duration:** Two days is extremely short for a dietary intervention. While it shows rapid metabolic changes, it cannot speak to long-term adaptation, tolerance, or sustained effects. Three weeks is moderate for a rat study (rat lifespan is ~2–3 years), but still short relative to human interventions that might last months or years.

**Statistical approach:** The authors report p-values from what appears to be a t-test or ANOVA (P < 0.0001 for blood ketones; P < 0.005 for hippocampal NAD ratio). They do not report effect sizes, confidence intervals, or correction for multiple comparisons (they tested two brain regions at two time points, which is four comparisons).

**What this design can prove:** The design can demonstrate that a ketogenic diet causes a measurable increase in blood ketones and a change in hippocampal NAD+/NADH ratio in healthy adult male rats. Because it is a controlled experiment with a defined intervention and comparison group, it can establish causation — the diet caused the metabolic changes.

**What this design cannot prove:** It cannot prove that increased NAD+ is the *primary* mechanism behind all KD benefits. The study only measures one outcome (NAD ratio) in one species (rats) in one sex (male) at two time points. It does not test whether blocking the NAD increase would eliminate KD benefits, nor does it compare NAD changes to other proposed mechanisms (e.g., adenosine, mitochondrial biogenesis, reduced oxidative stress). The small sample sizes mean the results may not be reproducible. The lack of blinding and randomisation details raises concerns about bias. The study cannot speak to human effects, long-term outcomes, or clinical relevance.

**Major methodological weaknesses:**

Very small sample sizes (n=3–4 per KD group)

No explicit randomisation or blinding described

Only male rats studied (sex differences in metabolism are well-documented)

Only two brain regions examined (hippocampus and frontal cortex)

No measurement of other proposed mechanisms to compare effect sizes

No dose-response testing (only one KD formulation: 6:1 ratio)

No measurement of NAD+ levels in other tissues (liver, muscle, blood)

The paper is a perspective, not a full experimental report — the data are preliminary

**Blood ketones (β-OHB):** KD induced a significant increase in blood β-hydroxybutyrate at both 2 days and 3 weeks compared to control diet (P < 0.0001). This confirms the diet successfully induced ketosis.

**Hippocampal NAD+/NADH ratio:** After 3 weeks of KD, the NAD+/NADH ratio was significantly increased in the hippocampus compared to control diet (P < 0.005). The increase was already detectable at 2 days, though the paper does not report the 2-day statistical comparison separately for hippocampus.

**Cortical NAD+/NADH ratio:** No significant difference was detected in the frontal cortex between KD and control groups at either time point. This indicates the effect is region-specific, not global.

**Persistence:** The elevated NAD+/NADH ratio in hippocampus was maintained from 2 days through 3 weeks, suggesting an early and sustained metabolic shift.

**Proposed mechanism:** The authors calculate that glucose metabolism reduces 111 molecules of NAD+ per 1000 molecules of ATP produced, while ketone body metabolism reduces only 41 molecules of NAD+ per 1000 ATP. This ~63% reduction in NAD+ consumption is the biochemical basis for their hypothesis.

The paper does not report the actual numerical values of the NAD+/NADH ratios (mean ± SD) in the text or figure legends — only the p-values and the graphical representation in Figure 2. Based on the figure, the hippocampal NAD+/NADH ratio appears to increase by approximately 40–60% after 3 weeks of KD compared to control. Blood ketone levels increased from near-zero (~0.1–0.2 mmol/L) in controls to approximately 1.5–2.0 mmol/L in KD-fed rats — a 10- to 20-fold increase. To put this in context: in humans, nutritional ketosis typically produces blood ketone levels of 0.5–3.0 mmol/L, while fasting can produce levels up to 5–8 mmol/L. The 40–60% increase in hippocampal NAD+/NADH ratio is modest but potentially meaningful, as NAD+ is a substrate for enzymes like sirtuins and PARPs that are sensitive to small changes in NAD+ availability.

**What the authors acknowledge:**

The data are preliminary and from a small sample

The study only examined two brain regions

The mechanism is proposed, not proven

Further studies are needed to test causality (e.g., blocking NAD+ increase and seeing if KD benefits disappear)

**What a critical reader would note:**

**Sample size:** n=3–4 per KD group is extremely small. With such small groups, a single outlier can drive the result. The authors do not report individual data points or show that the data are normally distributed.

**Sex bias:** Only male rats were used. Female rats have different metabolic responses to ketogenic diets, including different ketone production rates and different hormonal interactions. The results may not generalize to females.

**No blinding or randomisation details:** Without these, the study is at high risk of experimenter bias. For example, if the researcher knew which rats were on KD, they might handle tissue samples differently or process assays at different times.

**No correction for multiple comparisons:** Testing two brain regions at two time points means four comparisons. Using p < 0.05 without correction inflates the chance of a false positive to ~18% (1 – 0.95^4).

**No dose-response:** Only one KD formulation (6:1 ratio) was tested. Different ratios (e.g., 3:1, 4:1) or different types of ketogenic diets (e.g., MCT-based, modified Atkins) might produce different NAD responses.

**No measurement of other mechanisms:** The paper proposes NAD+ as a primary mechanism but does not measure adenosine levels, mitochondrial biogenesis markers, oxidative stress, or sirtuin activity in the same animals. Without these comparisons, it is impossible to know whether NAD+ changes are larger, smaller, or correlated with other known mechanisms.

**No human data:** The original experiment is in rats. Human metabolism differs in important ways, including different ketone utilization rates, different brain size and structure, and different dietary compliance challenges.

**Industry funding:** The paper does not report funding sources. However, the authors have published extensively on ketogenic diets and may have conflicts of interest (e.g., consulting for companies that produce ketogenic products).

**Publication bias:** The paper is a perspective article, not a registered clinical trial or pre-registered animal study. Positive results are more likely to be published than null results. The lack of effect in the frontal cortex is reported but downplayed.

**Short duration:** Three weeks is a short intervention. Long-term effects (months to years) on NAD+ metabolism are unknown. Some studies suggest that prolonged ketosis may have different effects on NAD+ than short-term ketosis.

**No functional outcomes:** The study does not measure whether the NAD+ increase correlates with any functional benefit (e.g., seizure resistance, cognitive performance, longevity). It is purely a biochemical measurement.

For someone running their own n=1 experiment:

### What to test

**Intervention:** A ketogenic diet (high fat, very low carbohydrate, adequate protein). The classic KD uses a 4:1 or 3:1 ratio of fat to protein+carbohydrates by weight. For a self-experiment, a modified Atkins diet (MAD) or medium-chain triglyceride (MCT) oil supplementation may be more practical. Start with 20–50g net carbohydrates per day and increase fat intake to 70–80% of calories.

**Dose:** Aim for blood ketone levels of 0.5–3.0 mmol/L (β-hydroxybutyrate). This typically requires restricting carbohydrates to <50g/day and consuming 150–200g fat/day for a 2000-calorie diet. MCT oil (1–3 tablespoons/day) can help achieve ketosis more quickly.

**Comparator:** A baseline period of at least 1–2 weeks on your normal diet before starting the KD. Alternatively, compare days when you are in ketosis vs. days when you are not (e.g., after a carbohydrate refeed).

### Minimum meaningful duration

**For NAD+ changes:** Based on this study, changes in NAD+/NADH ratio occur within 2 days and persist at 3 weeks. A minimum of 2 weeks is needed to see stable metabolic adaptation. For functional outcomes (cognition, energy, inflammation), 4–8 weeks is more realistic.

**For seizure control (if applicable):** Clinical studies show that some patients become seizure-free within days, while others require weeks. A minimum of 4 weeks is recommended before evaluating efficacy.

**For longevity or anti-aging effects:** These would require months to years and are not practical for a self-experiment. Focus on short-term biomarkers instead.

### What to measure

**Primary metric:** Blood β-hydroxybutyrate levels (using a ketone meter, e.g., Keto-Mojo or Precision Xtra). Measure daily at the same time (fasting morning is best). Target: 0.5–3.0 mmol/L.

**Secondary metrics:**

- Blood glucose (to confirm low glucose, typically 70–90 mg/dL in ketosis)

- Subjective energy levels (daily 1–10 scale)

- Cognitive performance (e.g., reaction time, working memory using apps like BrainHQ or Cambridge Brain Sciences)

- Inflammatory markers (if accessible: hs-CRP, IL-6)

- Sleep quality (using a wearable like Oura Ring or Fitbit)

- Body weight and body composition (if weight loss is a goal)

**Optional advanced metrics:**

- NAD+ levels (commercial NAD+ test kits are available, e.g., from Jinfiniti or DoNotAge, but reliability varies)

- Sirtuin activity (not practical for home testing)

- Mitochondrial function (e.g., using a VO2 max test or lactate threshold test)

### Key confounds to control for

**Carbohydrate intake:** Even small amounts of hidden carbs (e.g., in sauces, medications, supplements) can knock you out of ketosis. Track everything using an app like Cronometer or MyFitnessPal.

**Protein intake:** Too much protein can be converted to glucose via gluconeogenesis and reduce ketone levels. Keep protein moderate (15–25% of calories).

**Calorie intake:** A ketogenic diet is not automatically calorie-restricted. If you eat ad libitum, you may gain or lose weight depending on your baseline. Track calories if weight stability is important.

**Hydration and electrolytes:** Ketosis causes water and electrolyte loss (sodium, potassium, magnesium). Supplement with 3–5g sodium, 2–4g potassium, and 300–500mg magnesium daily to avoid "keto flu" (headache, fatigue, cramps).

**Exercise:** Exercise increases ketone utilization and can lower blood ketone levels temporarily. Measure ketones at the same time each day (fasting morning) to control for this.

**Sleep:** Poor sleep increases cortisol and can affect ketone production and NAD+ metabolism. Track sleep quality and duration.

**Stress:** Chronic stress elevates cortisol, which can increase blood glucose and reduce ketone levels. Use a daily stress rating (1–10).

**Menstrual cycle (for women):** Ketone levels fluctuate with the menstrual cycle. Track cycle phase and compare same-phase data points.

**Medications:** Some medications (e.g., SGLT2 inhibitors for diabetes, certain diuretics) can affect ketone levels or increase the risk of ketoacidosis. Consult a doctor before starting a KD if you take any medications.

**Alcohol:** Alcohol is metabolized before fat and can temporarily halt ketosis. Avoid or track carefully.

### What a positive result would look like

**Biochemical:** Blood β-hydroxybutyrate consistently >0.5 mmol/L (ideally 1.0–3.0 mmol/L) after 3–7 days of carbohydrate restriction. Blood glucose consistently <90 mg/dL.

**NAD+ proxy:** While you cannot easily measure brain NAD+, you might see improvements in biomarkers associated with NAD+ activity: better sleep quality (deeper sleep, fewer awakenings), improved cognitive function (faster reaction time, better memory recall), reduced inflammation (lower hs-CRP if tested), and improved energy stability (no afternoon crashes).

**Functional:** Subjective improvements in mental clarity, sustained energy, reduced brain fog, better mood stability, and (if applicable) reduced seizure frequency or severity.

**What a negative result looks like:** No change in blood ketones after 7–10 days (suggests carbohydrate intake is too high or protein intake is too high). No improvement in cognition or