#21 – Tom Dayspring, M.D., FACP, FNLA – Part II of V: Lipid metrics and cholesterol regulation

Dayspring insists this is not a fringe position — citing MESA and re-analyzed Framingham data that make the particle-superior-to-cholesterol finding 'unambiguous.' His frustration is institutional: guidelines and third-party payers have not caught up. Attia notes patients can self-order from LabCorp if their physician declines. Dayspring also emphasizes that when concordant (both high or both normal), either metric suffices; the clinical urgency is in discordant cases.

Dayspring places non-HDL-C in a hierarchy: Total-C → LDL-C → Non-HDL-C → apoB/LDL-P. Each step better approximates true particle burden. Non-HDL-C was the 'new thing' being incorporated into guidelines at the time of this episode, partially because it required no new test.

Dayspring's evolutionary argument is compelling: NPC1L1 prefers cholesterol over phytosterols; ACAT barely esterifies phytosterols; ABCG5/G8 evicts phytosterols before cholesterol. Three independent layers of defense suggest evolution regarded circulating phytosterols as harmful. The extreme case — sitosterolemia (homozygous ABCG5/G8 loss-of-function) — causes massive sterol accumulation and early coronary artery disease, demonstrating pathological potential even at modest elevations.

Dayspring: 'Triglycerides... the formula would be useless. They actually develop the direct LDL-C to give us a metric in people with triglycerides of 800, 1200, 4000.' The Hopkins Martin-Hopkins equation (developed after this episode aired) significantly reduces error in the 150-400 TG range vs Friedewald. Until direct assays or apoB are universally adopted, the Hopkins formula is a meaningful interim improvement.

Dayspring explains that lathosterol — a cholesterol precursor in the Kandutsch-Russell synthetic pathway — can be measured by liquid chromatography/mass spectrometry and serves as a marker of endogenous synthesis rate. 'If it was up, hey, you're over-synthesizing cholesterol.' The bifurcated Bloch and Kandutsch-Russell pathways both terminate in cholesterol but use different penultimate intermediates (desmosterol vs. lathosterol/laosterol). Knowing which branch dominates informs therapeutic selection.

Attia describes noticing during residency that HDL-C plummeted whenever patients entered SIRS. His retrospective explanation: the adrenal glands dramatically upregulate steroid hormone synthesis during physiologic crisis and upregulate LDL receptor expression to supply substrate. This is a reversible, demand-driven reallocation — not chronic lipid pathology. Drawing and treating a lipid panel in this context would be both inaccurate and potentially harmful.

The system's selectivity is a window into evolutionary priorities: NPC1L1 is most eager for cholesterol (essential nutrient), ABCG5/G8 is most eager to expel phytosterols (xenosterols). Stanols (fully saturated sterols produced by gut microbiota from biliary cholesterol) are virtually non-absorbable under normal conditions because they are the lowest-affinity NPC1L1 substrates and are rapidly ABCG5/G8-exported — hence why stanols in food don't accumulate in plasma.

Dayspring explains the particle-cholesterol split mechanistically: small LDL particles carry less cholesterol per particle, so a patient with small-LDL phenotype can have a normal LDL-C while running 2000+ nmol/L LDL-P — roughly double the atherogenic burden implied by the cholesterol number. The MESA and re-examined Framingham data made this 'unambiguously clear.' Attia argues that LDL-P should be on the standard five most important metrics every patient receives, and that patients who cannot get apoB/LDL-P ordered by their physician can access LabCorp directly without a prescription.

Dayspring's hierarchy: total cholesterol is the crudest proxy for apoB burden; LDL-C is a better proxy; non-HDL-C is better still; apoB or LDL-P is the actual measurement. The newer Hopkins-derived LDL-C calculation (replacing Friedewald) also improves accuracy at high TG, but non-HDL-C is immediately available on any standard panel without additional testing.

Dayspring notes: 'They actually develop the direct LDL-C to give us a LDL cholesterol metric and people with triglycerides of eight hundred, twelve hundred, four thousand — we now know what their LDL cholesterol is, whereas the formula would be useless.' The Hopkins-derived new formula tightens accuracy in the 150-400 TG range where Friedewald begins to drift but hasn't failed completely.

Dayspring: 'Data shows that phytosterols on a per molecule basis are probably more atherogenic than cholesterol.' The clinical protocol he advocates: before prescribing phytosterol supplements, measure plasma sitosterol and campesterol. A hyper-absorber — defined by loss-of-function variants in ABCG5/G8 or over-expression of NPC1L1 — will accumulate plant sterols in LDL particles and deliver them to the arterial wall, where they act like cholesterol but may be more pro-inflammatory. The fact that evolution built three sequential barriers (NPC1L1 uptake selectivity, ACAT esterification selectivity, ABCG5/G8 efflux preference) against phytosterols entering the body is Dayspring's strongest evidence that they were never meant to circulate.

Dayspring describes the enterocyte as having two efflux portals: ABCG5/G8 on the luminal face (excretes sterols to stool) and ABCA1 on the basolateral face (loads sterols onto nascent HDL for return to circulation). TICE uses the luminal portal, driven by cholesterol delivered to enterocytes from plasma-side lipoproteins or albumin — not from dietary absorption. The clinical implication is that statins and PCSK9 inhibitors, which lower LDL-C by increasing hepatic uptake, do not necessarily maximize fecal sterol excretion; TICE-targeting strategies could add complementary benefit.

Dayspring endorses the hypothesis as 'brilliant': 'cholesterol is not producing energy, cholesterol cannot be metabolized and produced ATP.' Cholesterol can be converted to bile acids (hepatocytes), steroid hormones (adrenal cortex, gonads), or excreted — those are the only fates. The rigidity of this design allows plasma LDL-C to be driven very low by PCSK9 inhibitors without compromising cellular cholesterol homeostasis, because every cell autonomously synthesizes what it needs via the mevalonate pathway.

Dayspring: 'Cholesterol lipidology in the brain might as well be in another different body — it has nothing to do with what's going on in the rest of your body.' ApoE in the CNS performs the same lipid-trafficking role that apoB100 performs peripherally, but it is the peripheral apoE variant (E2/E3/E4) that determines Alzheimer's risk through mechanisms unrelated to circulating lipid levels. The brain's autonomy explains why individuals with PCSK9 gain-of-function mutations (extremely low circulating LDL) show no cognitive deficits — the brain simply ramps up its own mevalonate pathway activity.

Dayspring explains that apoB100's non-transferability is due to its structural role: it provides the 'structural integrity' of the particle and serves as the ligand for the LDL receptor. Unlike apoC1, apoC2, apoC3, and apoE, which freely exchange between particles, apoB stays on the particle from synthesis to catabolism. The LDL receptor 'lock' reads the apoB 'key' to internalize the particle when cells express it. A particle lacking apoB (e.g., an HDL) has no LDL-receptor ligand and is cleared by entirely different machinery. Because 95% of apoB particles are LDL, plasma apoB is essentially an LDL particle count for clinical purposes — VLDL and IDL contribute the remaining ~5%.