The National Health and Nutrition Examination Survey (NHANES) is a continuous, nationally representative survey that measures the health and nutritional status of people living in the United States. It is run by the CDC's National Center for Health Statistics (NCHS), and it is the only U.S. national health survey that combines in-person interviews, physical examinations, and laboratory tests for participants of all ages.

For medical trainees and early-career researchers, NHANES is one of the most valuable publicly available datasets in existence. It's free, well-documented, and has supported tens of thousands of peer-reviewed publications — including papers in the New England Journal of Medicine and JAMA. If you're trying to publish original research without spending months on chart reviews or patient recruitment, NHANES is often the fastest credible path there.

This guide walks through what NHANES is, what kinds of questions it can answer, what its limitations are, and how it compares to the other public databases trainees commonly use.

NHANES at a Glance

Here are the core facts worth knowing before you go any further:

• Run by: CDC's National Center for Health Statistics (NCHS)
• Sample size: Approximately 10,000 participants per year, drawn from 15 counties across the U.S. annually
• Population covered: The civilian, non-institutionalized U.S. population — all ages, all 50 states and Washington, D.C.
• Data collection method: In-home interviews followed by physical exams and lab tests in mobile examination centers (MECs)
• Cycle: Continuous since 1999, with data publicly released in 2-year cycles
• Cost to access: Free, no application or approval required for public-use data
• Oversampled groups: Adults 60+, African Americans, Asian Americans, and Hispanic Americans (to allow more precise subgroup estimates)

A Brief History (And Why It Matters)

NHANES didn't start out as one continuous survey. The first National Health Examination Survey (NHES) was conducted from 1960 to 1962 and focused on adults aged 18–79. Two follow-up surveys covered children and adolescents through the 1960s. In the early 1970s, the survey was expanded to include nutritional measurements and renamed the National Health and Nutrition Examination Survey.

From 1971 through 1994, NHANES was conducted as a series of discrete surveys (NHANES I, II, and III). Beginning in 1999, NHANES became a continuous survey, with data released in two-year cycles.

Why does this history matter for your research? Two reasons. First, when you're searching for variables, you'll often see references to "NHANES III" or "Continuous NHANES" — these refer to different eras of the survey with different file structures. Second, the shift to continuous collection in 1999 is what makes NHANES so powerful for modern research: you can combine multiple cycles to increase sample size, study trends over time, and ask questions that require comparing pre- and post-event data (for example, before and after the COVID-19 pandemic).

What NHANES Actually Measures

This is where NHANES becomes unusual. Most national surveys rely entirely on what participants tell interviewers. NHANES combines what people report about themselves with what's directly measured during a physical exam. That combination is what makes it irreplaceable for certain kinds of research.

The data are organized into five major components, each released as a separate set of files for every two-year cycle:

Demographics. Age, sex, race and ethnicity, household income, education, marital status, country of birth, and military service history.

Dietary. Two 24-hour dietary recalls per participant, along with dietary supplement use, food security status, and dietary behaviors. NHANES is one of the most detailed sources of population-level nutritional data in the world.

Examination. Body measurements (weight, height, BMI, body composition), blood pressure, dental exams, and — depending on the cycle — vision, hearing, dermatology, balance, grip strength, and respiratory testing.

Laboratory. Blood, urine, and other biospecimen results, including markers for chronic disease (cholesterol, A1c, kidney function), infectious disease, environmental exposures (lead, BPA, PFAS), and nutritional biomarkers (vitamin D, ferritin).

Questionnaire. Self-reported information on chronic conditions, prescription medications (verified by examining bottle labels during the home interview), preventive care, mental health, sexual and reproductive health, occupation, and dozens of other topics.

The combination of measured data (from the exam and lab) with self-reported data (from the interviews) is what makes NHANES so unusual. You can study not just what people say about their health, but what their bodies actually show.

What Kinds of Research Questions NHANES Can Answer

Because of its scope, NHANES can support a wide range of research questions. A few examples of the types of questions trainees commonly investigate:

• Prevalence and trends. What percentage of U.S. adults have undiagnosed diabetes? How has the prevalence of obesity in children changed over the past 20 years?
• Risk factor associations. Is there an association between sleep duration and hypertension in middle-aged adults? Does dietary fiber intake correlate with markers of inflammation?
• Disparities research. Are there racial or socioeconomic disparities in cardiovascular risk factor control? How does insurance status affect access to preventive care?
• Environmental exposures. What population-level biomarkers exist for exposure to specific chemicals or pollutants? How do exposure levels vary by demographic group?
• Nutritional epidemiology. How do dietary patterns relate to specific health outcomes? Are Americans meeting federal nutrition guidelines?

NHANES has been used to generate national estimates for obesity, diabetes, and blood pressure, and to establish national reference standards for measurements like height, weight, and blood pressure. The same dataset that supports those landmark public health figures is available to you as a trainee.

Limitations to Know Before You Start

NHANES is powerful, but it's not the right tool for every question. The most important limitations to understand:

Cross-sectional design. Each cycle is a snapshot in time. NHANES generally cannot prove causation — only association. If you need to follow the same individuals over years, NHANES is not your dataset.

Complex survey design. NHANES uses oversampling and weighted sampling to produce nationally representative estimates. This means standard statistical methods don't work — you must use survey weights, strata, and primary sampling units in your analysis, or your results will be biased. This is the most common reason trainee projects using NHANES go wrong.

Variables change between cycles. Some measurements are collected in every cycle; others appear only in specific years. Before you commit to a question, you need to confirm your variables of interest exist in the cycles you want to study.

Limited geographic resolution. Because of how the sample is drawn, NHANES cannot produce state-level or city-level estimates. For state-level research, datasets like BRFSS are better suited.

Self-reported components. Even though NHANES includes measured data, large parts of the survey still rely on self-report (diet recalls, behavior questionnaires), which introduces the usual recall and social desirability biases.

These limitations aren't dealbreakers — they're the boundaries that define what NHANES is good for. Knowing them upfront protects you from designing a study NHANES can't actually answer.

How NHANES Compares to MEPS, NHIS, and HINTS

NHANES is one of four major publicly available U.S. health databases that Lumono supports. Each serves different research purposes:

• NHANES is the only one of the four that includes physical exams and lab tests. Best for questions involving measured biomarkers, nutritional status, or clinical measurements.
• MEPS (Medical Expenditure Panel Survey) focuses on healthcare costs, utilization, and insurance coverage. Best for health economics and policy questions.
• NHIS (National Health Interview Survey) is the longest-running U.S. health survey and the largest by sample size. Best for tracking long-term trends in health status, healthcare access, and behaviors.
• HINTS (Health Information National Trends Survey) focuses on how Americans access and use health information. Best for questions about health communication, digital health, and patient-provider information exchange.

If your question involves a lab value, a body measurement, or a nutritional biomarker, NHANES is almost certainly the right starting point. If it involves cost, coverage, or communication, one of the others will likely serve you better.

A Note on Statistical Analysis

This is worth saying directly: NHANES is not a dataset you can drop into a basic statistics workflow and expect to get correct answers. The complex survey design means that running an ordinary t-test or logistic regression on NHANES data will give you the wrong standard errors and potentially the wrong point estimates.

To analyze NHANES correctly, you need to:

• Use survey-aware procedures (svy commands in Stata, the survey package in R, or PROC SURVEY* in SAS)
• Apply the correct sample weights for the cycles and subpopulation you're studying
• Account for the survey's stratification and clustering
• Handle subpopulation analyses carefully — never exclude observations from a weighted dataset; you must use subpopulation commands instead

This is the step where most trainee projects stall. It's also the step where Lumono is built to help.

How Lumono Helps You Use NHANES

Lumono is built specifically to make publicly available health databases — including NHANES — usable for medical trainees who don't have years of biostatistics training. We handle the data preparation, walk you through formulating a feasible research question, run the statistically appropriate analyses, and help you interpret the results in the context of existing literature.

If you're considering an NHANES project, the database study timeline post is a good next read — it walks through what the project will actually look like from idea to submission. And if you want a broader view of the public databases trainees should know about, our overview of publicly available health datasets covers the full landscape.

NHANES has supported decades of high-impact research, and it's free for anyone willing to learn how to use it. With the right framing and the right support, it can be the dataset behind your first — or next — publication.