NHANES
Your Health Information
NHANES Looks at Homocysteine On a National Level
Data Could Shed Light on Marker-or-Mediator Controversy
Many important scientific findings have come out of the U. S. Centers for Disease Control and Prevention's National Health and Nutrition Examination Survey, commonly referred to as NHANES. For example, it was NHANES data that demonstrated decreasing blood lead levels as lead was taken out of gasoline, and more recent data from NHANES III (1988- 1994) showed that folate fortification of U. S. grain products had significantly raised U. S. population serum and red cell folate levels, which has in turn has been shown to have a positive effect in reducing the incidence of neural tube defects caused by folate deficiency. Not only have NHANES data guided national health policy, but large-scale population studies such as NHANES also help researchers determine what exactly constitutes the "normal" range of laboratory test values for a given population, and which test methods are the best tools for measuring the various analytes. Two years ago, CDC kicked-off NHANES 1999+-the latest version of the NHANES assessment-with a new addition to the list of analytes that will be examined. The NHANES Central Laboratory at CDC will now measure plasma homocysteine, a somewhat controversial marker of coronary artery disease (CAD), in order to gather population-based data.
But when CDC decided to initiate homocysteine analyses in the NHANES laboratory, no acceptable commercial assays were available. Consequently, they were left with an interesting dilemma, namely how to gather national data on an analyte for which there was neither an adequate reference method nor reference material. The answer, pursued by scientists at the NHANES lab, was to customize an assay specifically suited to the unique requirements of the NHANES program.
A National Health Census
But just how good does an assay have to be in order to fulfill the stringent prerequisites of a national testing program? "Every assay that we use for NHANES has to be state of the art and validated to the max," explained Elaine Gunter, MT( ASCP), Deputy Director for Management and Operations of CDC's Division of Laboratory Services (Atlanta, GA.), and previously Chief of the NHANES Laboratory. "The assays we use in the NHANES program have to be the best that are available."
The reason for this level of excellence is simple; NHANES is a population-based survey designed to collect information on the health and nutrition of the U. S. population, and it is the means by which the nation's population reference values are established. "The data that is generated by this survey is used by many government agencies to set national health policy," explained Gunter, who oversees the NHANES program. For example, the FDA uses vitamin deficiency/overload data to determine whether more or less fortification should be included in food products, and whether levels of vitamins and minerals should be increased in supplements. The USDA uses other NHANES data to determine funding levels for programs such as food stamps and the Women, Infants, and Children clinics.
Not only that, NHANES is also an important tool for addressing issues of public health. Twenty years ago, data from NHANES II (1976- 1980) showed that as lead was taken out of gasoline, blood lead levels in the population dropped almost in parallel. This reduction was not predicted by statistical models created by the Environmental Protection Agency (EPA), and the data forced EPA to announce an end to leaded gasoline sales in the U. S. in 1981. More recently, NHANES III data on folate levels were used to show that the population response to folate fortification in U. S. cereal and grain products had resulted in a three-fold increase in serum folate levels and a doubling of RBC folate levels. "In essence, we practically wiped out folate deficiency, and NHANES was a big part of showing that fortification efforts were having the desired effect," said Gunter.
Why Add Homocysteine?
NHANES III also saw one of the first efforts at gathering population data for U. S. levels of homocysteine, although it didn't start out that way. "When NHANES III was in the planning stages, we didn't know the significance of homocysteine, so the administrators didn't initially include it in the survey," explained Gunter. However, by the time the survey reached its second phase (1991- 1994), it was starting to become obvious that homocysteine could be more than just a marker of vitamin deficiency.
One proponent of the value of homocysteine as a marker of CAD is Donald Jacobsen, PhD, director of the Laboratory for Homocysteine Research at the Cleveland Clinic's Lerner Research Institute in Cleveland, Ohio. "Homocysteine is certainly a marker of cardiovascular disease," noted Jacobsen," but does it directly mediate vascular cell damage? We're not sure yet, but data coming in from NHANES and prospective studies will eventually allow us to make that determination." Right now, said Jacobsen, the data is mixed: some studies point to a direct prothrombotic effect of homocysteine in vascular disease, while others studies show that vitamin B deficiency is the causative agent, with homocysteine merely a marker of low vitamin B levels.
Despite the fact that homocysteine analyses weren't initiated until nearly the end of the NHANES III survey, results from that survey have already shed some light into what constitutes "normal" homocysteine levels in the U. S. population. A 1999 study by NHANES investigators (Ann Intern Med 1999; 131: 331), showed that homocysteine levels continue to increase as a person ages, and that men have a higher mean concentration than women. Although this data from a nationally representative sample of Americans merely confirmed the age and sex differences observed in earlier, non-representative samples, it began the informal process of defining homocysteine reference ranges for the U. S. population.
NHANES Data Provides First Population-Based Reference Levels for Homocysteine |
|||
|---|---|---|---|
| NHANES 3 was the first national effort to accumulate information about homocysteine levels in a representative sample of the U. S. population. Their results, published in the Annals of Internal Medicine (1999; 131: 331), suggest the following reference ranges for serum total homocysteine in U. S. citizens: | |||
| 12- 19 years of age |
>=60 years of age |
cut- off for "high" levels |
|
| Male | 4.3- 9.9 µmol/ L |
5.9- 15.3 µmol/ L |
11.4 µmol/ L |
| Female | 3.3- 7.2 µmol/ L |
4.9- 11.6 µmol/ L |
10.4 µmol/ L |
"In 1990, people thought that a homocysteine level greater than 20 µmol/ L was the cutoff for developing disease," said Paul Jacques, a co-author of that article, and associate professor and chief of the Nutritional Epidemiology Program at the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University (Boston, Mass.) where the Phase II homocysteine analyses were performed. "That number has continued to go lower and lower, and today people think that a level somewhere around 12 µmol/ L should be the cutoff. Our data suggests that might be a little high, but we'll get a better picture of what's going on when some of the large-scale prospective studies are complete and the NHANES 99+ data are interpreted."
Building a Better Test
In fact, NHANES 1999+ may provide even more accurate information about homocysteine in the U. S. population than its predecessor, NHANES III. In NHANES III, researchers performed homocysteine determinations on available surplus sera- not the best matrix for determining homocysteine, because red blood cells continue to metabolize and excrete homocysteine as blood clots, which results in an artificially elevated homocysteine level. Most reference laboratories prefer to use plasma separated immediately from a whole-blood sample that was collected in an EDTA tube, and the NHANES Lab in Atlanta decided that it would go that route, too.
"It was a tough decision whether to switch assays between NHANES III and NHANES 1999+," recalled Christine Pfeiffer, PhD, Chief of the NHANES Lab in Atlanta. However, improved reagents and refined methods published within the past three years have made it possible to upgrade the homocysteine assay used in NHANES III. "The first thing we did was to make a very thorough survey of the literature to determine the advantages and disadvantages of the methods that were out there," said Pfeiffer.
After identifying potential candidate methods, Pfeiffer selected an assay that uses a stable, water-soluble phosphine derivative-tris (2-carboxethyl) phosphine-as a reducing agent (Clin Chem 1997; 43: 687), and a derivatization step in which ammonium-7- fluoro-2,1,3-benzoxadiazole-4-sulfonate (SBD-F) is used to derivatize thiols for fluorescent detection (Eur J Clin Chem Clin Biochem 1991; 29: 549). "The water soluble reducing agent really appealed to me, because you don't have to use the traditional tributyl phosphine [TBP], which requires special handling due to the fact that it is combustible, or an organic solvent to dissolve TBP," said Pfeiffer.
The homocysteine assay Dr. Pfeiffer selected was then subjected to rigorous optimization and validation procedures-including testing the accuracy and precision of the assay-before it was considered ready to join the NHANES test line-up. The validation process started with characterization of the imprecision of the assay, consisting of within-assay and between-assay determinations. "We usually do five replicates from the same sample- not just once, but over the course of 2-5 days so that we can see how the within-assay imprecision looks," explained Pfeiffer. "Then we do between-run imprecision and look at reproducibility for at least twenty days to see if there are problems in re-starting the assay after it hasn't been used for a few days or weeks." Few studies make an attempt to characterize long-term performance, said Pfeiffer, but because U. S. public health policy is in part underpinned by NHANES data, the NHANES program is obligated to make sure their assays produce the most accurate and precise data possible.
When the validation of the HPLC assay was completed, the NHANES Laboratory thoroughly evaluated the newly available and fully automated Abbott (Abbott Park, Ill.) IMx fluorescence polarization immunoassay. After finding very low imprecision and very good comparability to the newly devised in-house HPLC method, the CDC decided to use the Abbott assay as a routine, high-throughput method, and to perform the HPLC assay on a 10% subset of the NHANES population for reference purposes.
One remaining problem with homocysteine analyses is the lack of a certified reference material. "What you can't really do without a reference material is say for sure whether you are close to the 'true' value because you don't know what that is. So you just have an indirect measure of accuracy through spiking and recovery, which is the best thing you can do in the absence of a certified reference material," explained Pfeiffer.
A Fantastic Tool
NHANES is unique in that it offers insight into measurements that normally wouldn't be taken in clinical trials- or even in the course of normal health management- and accumulates this data into a huge national database. "The beauty of NHANES is that we have 5,000 variables that you can sort through- on a population basis-looking for disease associations," said CDC's Gunter.
The NHANES program interviews and tests 5,000 people across the U. S. each year, from one year to over a hundred-plus years of age. The program tracks the prevalence of cancers, diabetes, and coronary disease; monitors electrocardiography and blood pressure in the U. S population; allergy data, dietary intake, medication and supplement intake, and occupational exposures. The NHANES participants are tracked over time through annual mailing list contacts and through and association of the NHANES data with national cancer registries, and the national death index., "We gather a lot of demographic information about NHANES participants," said Gunter "to help interpret disease prevalence and health conditions. To help interpret homocysteine results and their relationship to CAD, we also have the whole battery of serum and RBC folate, vitamin B12, homocysteine, methylmalonic acid, lipid analyses, blood pressure, weight, and any pre-existing diseases that individual may have.
But where does all this information go, and what does it mean for national health policy? "We've added homocysteine to the NHANES test menu in the hope that NHLBI can sort all this data and determine whether homocysteine is a better predictor of development of CAD. NHANES is a fantastically useful tool in addition to the ongoing clinical trials that are looking at the relationship between homocysteine and CAD."
by David Sainato, Clinical Laboratory News, August 2001, volume 27, number 8