Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

Ability of Nonfasting and Fasting Triglycerides to Predict Coronary Artery Disease in Lupus Patients

Touma1,  Zahi, Urowitz2,  Murray, Ibanez1,  Dominique, Gladman1,  Dafna

University of Toronto, Toronto Western Hospital, Toronto, ON,
U of Toronto, Toronto Western Hospital, Toronto, ON


Hypertriglyceridemia is a metabolic disorder associated with atherosclerosis. We have previously shown that there is no clinically significant difference in individual lupus patients in the levels of fasting triglycerides (FTG) and nonfasting triglycerides (NTG). We aimed to determine whether nonfasting triglycerides predict coronary artery disease (CAD) in lupus patients.


Patients are followed at regular intervals (at 2–6 months) according to a standard protocol which includes: complete history and physical exam, SLE Activity Index (SLEDAI-2K), and Systemic Lupus International Collaborative Clinics Damage Index (SDI). Fasting lipid profile was measured once yearly and nonfasting was determined at all other visits. We included the entire patient cohort and all the values of TG available in our data. Time-dependent covariate survival analysis was conducted to determine the predictive ability of TG for CAD – whether fasting and nonfasting. Variables considered were: sex, age at diagnosis, age, SDI, SLEDAI-2K, smoker, glucocorticoid, antimalarial, immunosuppressive drugs, cholesterol. Variables retained for multivariate analysis were selected through the variable reduction strategy (Harell) – selecting variables which alter the parameter estimate of TG by ±10% when include in the model. Using this variable selection, age, SDI, and immunosuppressive drugs were selected to be included in the models.


We identified 1289 patients since the date of the first TG available. Eighty eight percent of the patients were female. Five hundred forty one patients had elevated cholesterol level and the length of follow-up from the 1st TG to CAD/last visit was 8.82±8.19 (table 1).

Table 1. Patients' demographics

Number of patients1289
Number of CAD events108 (8.1%)
Mean number of visits with TG available prior to CAD/last visit18.8 ± 19.8
Sex F1143 (88.7%)
Age at diagnosis (years)30.7 ± 13.7
Age * (years)34.9 ± 13.5
Disease Duration * (years)4.2 ± 5.8
SDI *0.33 ± 0.82
SLEDAI-2K *8.81 ± 7.21
Smoker *249 (19.7%)
Steroids *834 (64.9%)
Antimalarial *516 (40.2%)
Immunosuppressive *296 (23.0%)
Elevated Cholesterol *541 (42.1%)
Length of FU from 1st TG to CAD/last visit *years)8.82 ± 8.19
* as of date of 1st TG available

One hundred eight patients (8.1%) developed CAD. We identified 89 events of CAD in 1137 patients in the nonfasting model and 35 events of CAD in 707 patients in the fasting model (table 2). Both nonfasting and fasting model showed ability of the variables to predict CAD; triglycerides, age and SDI while immunosuppressive drug use predicted CAD only in the nonfasting model.

Table 2. Hazard ratio for CAD in both group of analysis

 Non-fasting ModelFasting Model
 HR95% CIpHR95% CIp
TG1.801.10, 2.930.022.761.07, 7.100.04
Age1.061.04, 1.08<0.00011.061.03, 1.09<0.0001
SDI1.111.00,, 1.370.04
Immunosuppressive2.201.41, 3.440.00051.270.61, 2.650.52


Both fasting and nonfasting TG predicted CAD in lupus patients. Nonfasting TG levels can be used in clinic to detect CAD event in lupus patients.

To cite this abstract, please use the following information:
Touma, Zahi, Urowitz, Murray, Ibanez, Dominique, Gladman, Dafna; Ability of Nonfasting and Fasting Triglycerides to Predict Coronary Artery Disease in Lupus Patients [abstract]. Arthritis Rheum 2009;60 Suppl 10 :1560
DOI: 10.1002/art.26634

Abstract Supplement

Meeting Menu