A new 129-question food survey designed for Korean adults with obesity shows modest accuracy for ranking people’s eating habits and good consistency when used twice, according to Gram Research analysis. The survey correctly placed 68-82% of people in the same food intake category as detailed food records, with better results when averaged from two administrations. However, it shouldn’t be used to measure exact individual food amounts.
Researchers in Korea created a new 129-question food survey to help doctors and scientists understand what people with obesity are eating. They tested it with 97 Korean adults to see if it accurately measured food intake. The survey worked pretty well at ranking people from eating less to more food, and when they used it twice, they got similar results both times. However, the survey isn’t perfect at measuring exact amounts of what each person eats. Scientists say it’s useful for studying large groups of people and how diet connects to weight and health problems.
Key Statistics
A 2026 study of 97 Korean adults with obesity found that a new 129-item food survey achieved 68-82.5% agreement with detailed food records for ranking energy and nutrient intake across categories.
According to research reviewed by Gram, the food survey showed good reproducibility with 81.4-90.7% of participants falling into the same or adjacent intake categories when the survey was administered twice.
A 2026 validation study found that averaging results from two administrations of the food survey improved accuracy, with energy-adjusted correlation for sugar intake reaching 0.54 compared to food records.
The 2026 research demonstrated correlation coefficients between 0.50-0.69 for nutrient intake when the food survey was given twice, indicating good consistency for tracking dietary patterns in Korean adults with obesity.
The Quick Take
- What they studied: Does a new 129-question food survey accurately measure what Korean adults with obesity eat, and does it give consistent answers when used twice?
- Who participated: 97 Korean adults between ages 19 and 64 who have obesity participated in the study.
- Key finding: The survey was good at ranking people from low to high food intake (68-82% agreement with detailed food records), and gave consistent results when used twice (81-90% agreement). However, it wasn’t perfect at measuring exact amounts each person ate.
- What it means for you: If you’re in a research study about diet and weight, this survey can help researchers understand your eating patterns compared to others. It works better for group studies than for tracking your personal exact intake. Results may vary based on individual memory and reporting accuracy.
The Research Details
Researchers created a new food survey with 129 questions designed specifically for Korean adults with obesity. They tested it with 97 people aged 19-64. To check if the survey was accurate, they compared answers from the survey to detailed food diaries where people wrote down everything they ate for 3 days, four times throughout the year (covering all seasons). They also gave the same survey twice to the same people to see if they got similar answers both times.
The researchers used several methods to measure how well the survey worked. They looked at whether people fell into the same category (like low, medium, or high eaters) when using the survey versus the food diaries. They also calculated correlation numbers, which show how closely two measurements match. They even made special charts called Bland-Altman plots to see if there were big differences between the two methods.
Food surveys are important tools for scientists studying how diet affects obesity and diseases like diabetes and heart disease. However, asking people to remember everything they eat is hard. This study shows whether a carefully designed survey can reliably measure what Korean adults with obesity eat. Understanding this helps researchers decide if they can trust this survey for future studies about diet and health in this population.
The study used a reasonable sample size (97 people) and compared the survey to detailed food records, which is a solid approach. The researchers tested the survey twice to check consistency. However, the study only included Korean adults with obesity, so results may not apply to other groups. The study didn’t report some details like exact age ranges or how diverse the participants were. The wide limits of agreement shown in the charts suggest individual measurements can vary quite a bit.
What the Results Show
When researchers compared the food survey to detailed 3-day food records, they found moderate agreement. For energy (calories) and 10 nutrients measured, between 68% and 82.5% of people fell into the same or adjacent ranking categories. Correlation numbers ranged from 0.30 to 0.56, which indicates modest but not perfect agreement.
When the researchers averaged the answers from both times the survey was given, the results improved. For example, the correlation for sugar intake improved to 0.54 when adjusted for total calories. This suggests that using the survey twice and averaging the results gives more reliable information.
The survey was very consistent when given twice. Between 81.4% and 90.7% of people fell into the same or adjacent categories when comparing the first and second survey. Nutrient correlation coefficients between the two surveys ranged from 0.50 to 0.69, showing good reproducibility.
However, the Bland-Altman plots revealed that while average differences were small, individual measurements could vary quite a bit. The survey tended to overestimate flavonoid intake (plant compounds found in foods), suggesting some foods may have been reported inaccurately.
The survey showed good consistency for specific food and dish categories when given twice. This means people gave similar answers about how often they ate certain Korean foods both times. The consistency was similar to what researchers found for individual nutrients, suggesting the survey reliably captures eating patterns for common foods in the Korean diet.
This study adds to existing research on food surveys for people with obesity. Previous surveys have shown similar levels of agreement (moderate validity with correlation coefficients in the 0.30-0.60 range). The finding that averaging two surveys improves accuracy aligns with other research suggesting that repeated measurements reduce error. The overestimation of certain nutrients like flavonoids has been noted in other food surveys as well.
The survey works better for comparing groups of people than for measuring exact intake in individuals. The wide limits of agreement mean some people’s measurements could be quite different from their actual intake. The study only included Korean adults with obesity aged 19-64, so results may not apply to younger people, older people, or people without obesity. The survey may not capture all foods people eat or may have errors based on memory. Some nutrients, like flavonoids, were overestimated, suggesting certain foods may be reported inaccurately.
The Bottom Line
This survey is useful for researchers conducting large studies about diet and obesity in Korean populations (moderate-to-high confidence). It should not be used to measure exact food intake for individual patients (low confidence for this purpose). For research purposes, using the survey twice and averaging results improves accuracy (moderate-to-high confidence). Healthcare providers should be cautious about using this survey alone to assess individual dietary intake without additional methods.
Researchers studying diet and obesity in Korean populations should find this survey useful. Public health officials planning nutrition studies in Korea may use this tool. People with obesity participating in diet research may encounter this survey. Healthcare providers in Korea may use it as part of larger research efforts. This survey is less useful for individual dietary counseling or precise personal intake tracking.
If you’re in a research study using this survey, expect results to be meaningful when comparing your eating patterns to others in the study. Changes in diet tracked over weeks to months should be detectable. However, precise measurements of your individual intake may take longer to stabilize and should be interpreted cautiously.
Frequently Asked Questions
How accurate is this Korean food survey for measuring what I eat?
The survey ranks your eating patterns fairly well (68-82% accuracy) compared to detailed food records, but it’s not precise for exact amounts. It works better for comparing groups than individual measurements. Using it twice and averaging improves accuracy.
Can I use this food questionnaire to track my personal diet for weight loss?
This survey is designed for research studies, not personal tracking. While it reliably shows eating patterns compared to others, it may not accurately measure your exact intake. Talk to a dietitian about tools better suited for personal dietary tracking.
Does this food survey work for all Korean people or just those with obesity?
This survey was specifically tested on Korean adults aged 19-64 with obesity. Results may not apply to younger people, older people, or those without obesity. Different surveys may be needed for other populations.
What foods does this 129-item survey measure?
The survey includes 129 questions about Korean foods and dishes commonly eaten by Korean adults. It covers major food categories and traditional Korean meals, though the specific foods aren’t detailed in this research summary.
Why does the survey overestimate some nutrients like flavonoids?
People may misremember portion sizes or frequency of certain foods, or some foods may be reported inaccurately. This is common in food surveys. The researchers recommend caution when interpreting individual nutrient estimates, especially for flavonoids.
Want to Apply This Research?
- Track food intake using the 129-item categories from this survey weekly, comparing your rankings (low, medium, high intake) across food groups rather than focusing on exact portion sizes.
- Use the survey’s food categories to identify which Korean foods and dishes you eat most frequently, then set goals to increase or decrease specific categories based on health objectives.
- Complete the survey every 4-6 weeks and compare your food group rankings over time. Average results from two survey completions taken 1-2 weeks apart for more reliable tracking of dietary pattern changes.
This research describes a tool for measuring food intake in research studies, not a medical treatment or diagnostic tool. This survey should only be used as part of formal research studies or under professional guidance. Individual dietary assessment and weight management should be discussed with a qualified healthcare provider, registered dietitian, or physician. The findings apply specifically to Korean adults with obesity and may not generalize to other populations. This information is for educational purposes and should not replace professional medical or nutritional advice.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.
