The effect of baseline prediabetes on the relation between lipoprotein (a) and incident type 2 diabetes and outcome: observations from the UK-Biobank

Abstract Background A inverse association between plasma Lp(a) and risk of incident type 2 diabetes mellitus (T2DM) is known. Lp(a) concentration is genetically determined. Lower Lp(a) itself is not causally related to higher risk of incident T2DM. Purpose To see if baseline prediabetes (preDM) affects the association between Lp(a) and incident T2DM and outcome. Methods UK-Biobank cohort whose Lp(a) is available; known and undiagnosed T2DM (from baseline HbA1c) excluded. Major adverse cardiovascular event (MACE) - death, non-fatal myocardial infarction (MI), revascularisation or ischaemic stroke after entry obtained from Death Register and hospital episode statistics data. Using baseline HbA1c subjects divided into normal (Grp1), preDM (HbA1c≥39 mmol/mol) subdivided into mild (39-42) (Grp2) and severe (≥42) (Grp3). Time to incident T2DM and MACE in Lp(a) quartiles analysed by Kaplan-Meier method. Cox regression used to assess risk of incident T2DM/MACE in Lp(a) quartiles. Result 333753 entrants (83892-Q1, 82997-Q2, 83446-Q3, 83438-Q4); 45436 (13.61%) had preDM; 288337 (86.39%)-Grp1, 33618 (10.07%)-Grp2, 11818 (3.54%)-Grp3. T2DM developed in 15.3%, 13.9%, 12.9%, 12.8% patients with (p0.0001) and 1.72%, 1.7%, 1.7%, 1.6% without preDM (p=0.342) in Lp(a) Q1-4 respectively over 11.9 years follow up. In preDM patients, Lp(a) was lower in those who develop T2DM than in those who did not (20.33 nmol/l, IQR 9.04-60.27 vs 23.42 nmol/l, IQR 10.10-65.85, p0.0001). In non-preDM, Lp(a) was similar in patients who did or did not develop T2DM (p=0.067). Incident T2DM was inversely associated with Lp(a) quartiles (Fig1) in whole cohort (A) and in preDM (C) but not non-preDM patients i.e. Grp 1 (B). In preDM subgroups the inverse relation was seen only in Grp 3, (Fig1C2). After full adjustment, T2DM was lower in Q4 in whole cohort (HR 0.92, 95% CI 0.87-0.97, p=0.004) and in preDM (HR 0.89, 95% CI 0.82-0.96, p=0.002) and in Q3 (HR 0.88, 95% CI 0.80- 0.98, p=0.017) and Q4 (HR 0.82, 95% CI 0.74-0.91, p0.001) in Grp 3. MACE increased with Lp(a) quartiles in the whole cohort (Fig2A), in patients without (2B) and with (2C) preDM. Within the preDM subjects the inverse relation is seen only in those with highest risk of incident T2DM i.e. Grp3 (2C2). For every SD increase in the Lp(a), adjusted MACE increases by 6.43% (p0.0001). After full adjustment, compared to Q1, risk of MACE was higher in Q3 (HR 1.03, 95% CI 1.00-1.07, p=0.043) and Q4 (HR 1.15, 95% CI 1.12- 1.19, p0.0001) in whole cohort, in Q3 (HR 1.04, 95% CI 1.01-1.08, p=0.025) and Q4 (HR 1.16, 95% CI 1.11-1.20, p=0.0001) in Grp1 and only Q4 in Grp2 (HR 1.15, 95% CI 1.06-1.25, p=0.001). There was no increase in MACE at higher quartiles in Grp3 (Q3: HR 1.05, 95% CI 0.93-1.20, p=0.425; Q4: HR 1.13, 95% CI 0.99-1.28. p=0.064). Conclusion Incident T2DM was inversely associated with Lp(a) quartiles only in preDM. MACE does not increase with Lp(a) quartiles in preDM.Fig 1 Fig 2