Research Article

Hypoglycaemia, Abnormal Lipids, and Cardiovascular Disease among Chinese with Type 2 Diabetes

Table 4

Sensitivity analysis of odds ratio of hypoglycaemia and lipid profiles for cardiovascular disease after exclusion of 408 patients who used lipid lowering drugs.

OR (95% CI) value

Hypoglycaemia for CVD
 Model one<0.0001
  Nonhyperglycemia Reference
  Mild hyperglycemia2.86 (2.04–4.02)
  Severe hyperglycemia7.72 (4.73–12.60)
 Model two<0.0001
  Nonhyperglycemia Reference
  Mild hyperglycemia2.95 (2.02–4.29)
  Severe hyperglycemia6.75 (3.79–12.04)
 Model three<0.0001
  Nonhyperglycemia Reference
  Mild hyperglycemia2.95 (2.02–4.29)
  Severe hyperglycemia6.75 (3.79–12.04)
Lipid profile for CVD
 Model four
  LDL cholesterol, mmol/L1.06 (0.94–1.20)0.3604
  HDL cholesterol, mmol/L0.79 (0.70–0.89)<0.0001
  Triglyceride, mmol/L1.09 (1.02–1.16)0.0165
 Model five
  LDL cholesterol, mmol/L0.93 (0.80–1.09)0.3755
  HDL cholesterol, mmol/L0.91 (0.82–1.03)0.1201
  Triglyceride, mmol/L1.07 (0.99–1.16)0.0739
 Model six
  LDL cholesterol, mmol/L0.91 (0.77–1.07)0.2469
  HDL cholesterol, mmol/L0.88 (0.77–1.00)0.0580
  Triglyceride, mmol/L1.08 (1.00–1.17)0.0593

Model one: not adjusted for other variables.
Model two: adjusted for age, sex, BMI, systolic blood pressure and log-transformed urinary albumin to creatinine ratio, and drug use (statins, other lipid lowering drugs, renin-angiotensin system inhibitors, oral antidiabetes drugs [OADs] only, glucagon-like peptide-1 based treatment, basal insulin based treatment, and premixed insulin based treatment).
Model three: further adjusted for LDL-C, HDL-C, and triglyceride.
Model four: not adjusted for other variables.
Model five: adjusted for age, sex, BMI, systolic blood pressure, LDL-C, HDL-C, triglyceride and log-transformed urinary albumin to creatinine ratio, and drug use (statins, other lipid lowering drugs, renin-angiotensin system inhibitors, oral antidiabetes drugs [OADs] only, glucagon-like peptide-1 based treatment, basal insulin based treatment, and premixed insulin based treatment).
Model six: adjusted for the variables listed in model five and hypoglycaemia.