AI-Powered Analytics Platform

Explore Health Data with AI Intelligence

StatwiseAI is the AI-powered platform for researchers, students, and public health professionals. Explore complex health data through natural language and interactive visualizations.No code required.

1000+
Health Variables
20+
Years of Data
3
Major Datasets
StatwiseAI - AI-Powered Health Data Analytics Platform
Interactive AI Assistant

Ask Questions, Get Answers

Interact with health data like never before. Our AI understands natural language and provides instant insights.

StatwiseAI Chatbot
You

What is the correct weight variable for LBXTC?

For LBXTC (total cholesterol), the correct weight is usually the MEC exam weight: WTMEC2YR (because cholesterol is a lab measure taken in the Mobile Examination Center, not just in the interview).

Which weight to use (rule of thumb)

  • If LBXTC was measured on (basically) all examined participants in that cycle:
    • • Use WTMEC2YR for a single 2-year cycle
    • • If combining cycles, you create a combined MEC weight from WTMEC2YR per NHANES guidelines (e.g., divide by number of 2-year cycles you combine, assuming you stack cycles correctly).
  • If LBXTC was measured only in a fasting/subsample group in that cycle:
    • • Use the fasting/subsample weight, commonly WTSAF2YR (and an analogous multi-year version if you build it).

What not to use

• Don't use WTINT2YR for LBXTC (interview weights are not appropriate for lab analytes).

Quick Python check (to confirm if it's a subsample)

If you're unsure whether LBXTC is full-sample or fasting/subsample in your specific cycle, a practical check is whether many people have LBXTC missing while still having MEC weights:

# df is your merged NHANES dataset for a given cycle
# Columns expected: LBXTC, WTMEC2YR, (maybe WTSAF2YR)

df[["LBXTC", "WTMEC2YR", "WTSAF2YR"]].isna().mean()

If WTSAF2YR exists and LBXTC is only present among those with non-missing/nonzero WTSAF2YR, that's a clue you need the fasting/subsample weight.

Next steps

  • • Which NHANES cycle(s) are you using (e.g., 2017–2018, 2015–2016)?
  • • Are you analyzing LBXTC alone, or together with other fasting lipids (like triglycerides), which would force you onto fasting weights?
  • • If you want, paste your column names and I'll tell you exactly which weight to use for your merged file.
Core Capabilities

Powerful Features

Powerful features designed to make health data analysis accessible to everyone, from beginners to experts

AI-Powered Analysis

Ask questions in natural language and get instant insights from complex health datasets

Comprehensive Data

Access NHANES 1999-2020, HRS Longitudinal Waves, and MIDUS datasets with full metadata

Designed for You

Built for researchers, students, data scientists, public health professionals, and policy analysts

Powerful Performance

Get instant responses and real-time data analysis with optimized performance

Rich Metadata

Comprehensive variable documentation, survey cycles, and data quality information

Secure & Reliable

Enterprise-grade security with reliable data access and privacy protection

How It Works

Simple Three-Step Process

From exploring variables to sharing insights, we've made health data analysis effortless

1

Explore NHANES Variables

Access 1000+ health variables including demographics, health outcomes, biomarkers, and behavioral data from 1999-2020.

2

Get Instant Insights

Ask natural language questions to uncover trends, correlations, and patterns in health data without writing code.

3

Export & Share Results

Download visualizations as charts, get Python/R/Stata code for reproducibility, and share findings with your team.