Normal LDL-C, elevated ApoB: how often Ultrahuman's cohort carries the discordance the new guidelines target

LDL cholesterol measures how much cholesterol rides inside atherogenic particles. Apolipoprotein B counts the particles directly, since every LDL, IDL, VLDL, and Lp(a) particle carries exactly one ApoB on its surface. Two people with identical LDL-C can host very different particle counts, and the higher-count individual sits at materially higher cardiovascular risk. A 432,092-adult UK Biobank analysis found ApoB the stronger predictor of cardiovascular disease in men (Kelly et al., 2026). In UK Biobank participants with type 2 diabetes, ApoB reclassified the largest share of statin users whose LDL-C had already fallen to or below 100 mg/dL (Lyu et al., 2026). A March 2026 simulation argued an ApoB goal beats LDL-C and non-HDL-C targets in 65 percent of primary-prevention scenarios (Luebbe et al., 2026).

We pulled Blood Vision panels between August 2025 and April 2026 carrying both LDL-C and ApoB cleanly parsed from the lab report, giving 3,087 members. Median age 37 (IQR 31 to 44), 1,993 men and 1,038 women. Cut-points followed the ACC/AHA 2018 framework and the JAMA 2026 simulation: LDL-C below 130 mg/dL, ApoB below 90 mg/dL.

Quadrant prevalence:

• Concordant-low (LDL-C < 130, ApoB < 90): 1,228 members, 40 percent
• Concordant-high (both elevated): 1,001 members, 32 percent
• Hidden-ApoB (LDL-C < 130, ApoB ≥ 90): 799 members, 26 percent
• Discordant-LDL (LDL-C ≥ 130, ApoB < 90): 59 members, 2 percent

Of the 858 panels with discordant LDL-C and ApoB, 799 (93 percent) showed elevated ApoB with normal LDL-C; 59 (7 percent) showed the reverse.

‍

A hidden-ApoB share near a quarter of an engaged-tracker cohort is consistent with the reasoning behind the recent guideline shift. The discordance is common enough that an LDL-only screen misses a non-trivial fraction of the people the new framing is built to catch. Among the 1,817 members with at least 10 nights of ring-recorded resting heart rate in the 30 days before their draw, the hidden-ApoB group ran a median overnight RHR of 57.0 bpm against 56.5 bpm in the concordant-low group, a 0.5 bpm gap whose bootstrap 95 percent confidence interval brushes zero. The direction matches what the literature predicts; the magnitude in cross-sectional data sits below the pre-specified 2 bpm threshold and is reported as directional only.

The 2026 guidance assumes the LDL-versus-ApoB discordance is common enough that an LDL-only screen misses meaningful risk. In 26 percent of this cohort, it did.