The 1999–2000 and 2001–2002 NHANES conducted by the Centers for Disease Control and Prevention (CDC) were designed to be nationally representative of the noninstitutionalized, U.S. civilian population on the basis of a complex, multistage probability sample. Approximately 9,965 persons, 2 months to 85 years of age, were studied in NHANES 1999–2000, and 11,039 persons were included in NHANES 2001–2002. Details of the NHANES protocol and all testing procedures are available elsewhere (NCHS 2006a
). The study protocol was reviewed and approved by the CDC institutional review board; additionally, informed written consent was obtained from all subjects before they took part in the study.
PCDDs, PCDFs, PCBs, and OC pesticides were measured in serum from a random one-third subsample of people ≥ 12 years of age in 1999 and 2000. In 2001 and 2002, PCDDs, PCDFs, and coplanar PCBs were measured in a random one-third subsample of people ≥ 20 years of age, and OC pesticides and other PCBs were measured in these people and in a random one-third subsample of people ≥ 12–19 years of age.
The NHANES data collection included a standardized home interview followed by a detailed physical examination in a mobile evaluation clinic or the participant’s home. Information on demographic characteristics, ethnicity, and medical history of diabetes was obtained in a household interview. Venous blood and urine samples were collected and shipped weekly at –20°C. PCDDs, PCDFs, PCBs, and OC pesticides were all measured as individual chemicals by high-resolution gas chromatography/high-resolution mass spectrometry using isotope dilution for quantification. All of these analytes were measured in approximately 5 mL serum using a modification of the method of Turner et al. (1997)
. The POPs were reported on a lipid-adjusted basis using concentrations of serum total cholesterol and triglycerides.
Although 49 POPs were measured in both NHANES 1999–2000 and 2001–2002, to avoid bias in estimation among those below the limit of detection (LOD), we selected the 21 POPs for which at least 60% of study subjects had concentrations > LOD: 3 PCDDs, 3 PCDFs, 5 dioxin-like PCBs, 6 nondioxin-like PCBs, and 4 OC pesticides. There were a total of 1,054 study participants ≥ 40 years of age with information on serum concentrations of the 21 selected POPs. After excluding 165 diabetic participants, including newly diagnosed cases, the final sample size was 889. We excluded those with diabetes from the present study because simple adjustment for diabetic status may not be enough to exclude its effect due to strong associations between serum concentrations of POPs and diabetes (Lee et al. 2006b
). However, inclusion of diabetic patients did not materially change results.
For each POP, subjects with serum concentrations < LOD were regarded as the reference group, and subjects with detectable values were categorized by cutoff points of 25th, 50th, and 75th percentile values. To yield a cumulative measure of three PCDDs, we summed the ranks according to magnitude of detectable levels of the three POPs that belong to the PCDDs, using the rank 0 for any nondetectable value. The summary values were categorized by cutoff points of 25th, 50th, and 75th percentile values. We assigned and cumulated POP subclasses similarly for the three PCDFs, the five dioxin-like PCBs, the six nondioxin-like PCBs, and the four OC pesticides. Thus, depending on the sum of ranks of the several POPs belonging to the specific POP subclass under consideration, the subject could be in the lowest quartile or in a higher quartile; however, if all POPs in the subclass were nondetectable, the subject would be placed in the lowest quartile.
Participants were considered to have prevalent CVD if they answered “yes” to any of the following questions:
- “Has a doctor or other health professional ever told you that you had CHD?”
- “Has a doctor or other health professional ever told you that you had angina/angina pectoris?”
- “Has a doctor or other health professional ever told you that you had heart attack/myocardial infarction?”
- “Has a doctor or other health professional ever told you that you had a stroke?”
We used logistic regression models to calculate multivariate-adjusted odds ratios (ORs). Adjusting CVD risk factors were age (years), race/ethnicity, poverty income ratio (continuous), body mass index (BMI; continuous), cigarette smoking (never, former, or current), cotinine levels (nanograms per milligram), alcohol consumption (grams per day), leisure time physical activity (vigorous, moderate, or none), status of hypertension (yes/no), total cholesterol (continuous), HDL (high-density lipoprotein)-cholesterol (continuous), triglyceride (continuous), and C-reactive protein (continuous). We substituted median values of noncases for missing BMI, poverty income ratio, cotinine levels, or alcohol consumption in 152 subjects. Exclusion of these individuals did not change any conclusions, but it did limit power in some analyses; therefore, they were retained.
All statistical analyses were performed using SAS 9.1 (SAS Institute Inc., Cary, NC, USA) and SUDAAN 9.0 (Research Triangle Institute, Research Triangle Park, NC, USA). Estimates of main results were calculated accounting for stratification and clustering (Korn and Graubard 1991
), and adjusting for age, race, ethnicity, and poverty income ratio instead of using sample weights; this adjustment is regarded as a good compromise between efficiency and bias (Graubard and Korn 1999
; Korn and Graubard 1991
). Because results were very similar with both SAS 9.1 and SUDAAN 9.0, we present the results based on SAS 9.1.