News · Nutrition & Diet
The traditional Chinese diet tracked with fewer deaths, more obesity
A European Journal of Nutrition study of 11,158 Chinese adults found high adherence to the traditional diet linked to lower heart disease and death rates, and higher obesity rates.
Based on a peer-reviewed cohort study in European Journal of Nutrition
- Researchers used the China Health and Nutrition Survey from 1997 to 2011, publishing in European Journal of Nutrition.
- The sample was 11,158 adults aged 18 to 74, free of the relevant conditions at baseline and present in at least two survey waves.
- Adherence was scored with a purpose-built traditional Chinese diet index and split into low, medium and high bands.
- High adherence carried lower cardiovascular disease risk (OR, 0.73; 95% CI, 0.61-0.87) and lower all-cause mortality (HR, 0.73; 95% CI, 0.63-0.87).
- The same high adherence carried HIGHER obesity risk (OR, 1.19; 95% CI, 1.08-1.32) and central obesity risk (OR, 1.19; 95% CI, 1.12-1.26).
- When total energy intake was standardized, the obesity association reversed direction.
- That reversal is the finding. It suggests the obesity signal reflected how much people ate, not what.
- It is observational, and diet was self-reported over a period when China's food supply changed enormously.
Traditional diets get studied when they are Mediterranean and mostly ignored when they are not. The one feeding the largest population on earth had never been tested this way.
A study in European Journal of Nutrition closes that gap, and returns a split answer that is more interesting than a clean one would have been. The authors note that no research had examined associations between the traditional Chinese diet, defined using an a priori method, and disease outcomes or all-cause mortality.
Why defining the diet in advance matters
There are two ways to study a dietary pattern. You can let statistics find whatever combination of foods clusters together in your data, or you can define the pattern first and then test it.
The second is harder and more honest, because a pattern derived from the same data you are testing can be shaped, consciously or not, to fit the outcome. Adherence here was assessed using a recently developed index, specified before the analysis.
Who was followed, and for how long
The data source is a long-running national survey. The longitudinal study drew on data from the China Health and Nutrition Survey between 1997 and 2011.
Entry rules were sensible. Eligible participants were aged 18-74, had no relevant conditions at baseline, and participated in at least two survey waves. Requiring two waves means nobody’s diet is a single guess.
The sample was 11,158 participants with a mean age of 42.2 years, 51.3% women. Outcomes covered cardiovascular disease, diabetes, cancer, obesity, hypertension and death.
The good half
On the outcomes people fear most, high adherence looked protective.
High adherence was associated with lower risk of cardiovascular disease and all-cause mortality - both estimates around 0.73, and both with intervals comfortably clear of 1.
A mortality signal of that size from a dietary pattern is substantial, and death is the one outcome nobody misclassifies. It also lines up with what has been found for other traditional patterns built on grains, vegetables and legumes with little processed food.
The awkward half
Then the same diet went the wrong way on body weight.
High adherence was also associated with higher risk of obesity and central obesity, both around 1.19, both statistically clear. A pattern linked to fewer deaths and more obesity is not an obvious result, and a less careful paper might have buried it.
The sentence that resolves it
The explanation is in the same paragraph, and it changes the reading entirely.
When standardizing the energy intake on a common basis, adherence was inversely associated with obesity and central obesity.
Put plainly: once you account for how much people ate, the traditional pattern went from looking fattening to looking protective. The obesity signal was about quantity, not composition. A diet of rice and vegetables can be nutritionally sound and still deliver a great deal of energy if the bowl is large enough, and white rice in volume is a serious carbohydrate load.
That distinction is the practical content of the paper. It is also the reason “eat traditionally” is incomplete advice without “and watch the portion”.
What it cannot settle
It is observational, and diet was self-reported.
The larger problem is the calendar. Between 1997 and 2011 China’s food supply transformed faster than almost any in history, with processed food, cooking oil and meat all rising sharply. “Traditional adherence” in 1997 and in 2011 are not quite the same exposure, and people who held to the older pattern were more likely rural, poorer and more physically active in ways adjustment cannot fully absorb.
The authors’ conclusion stays measured: higher adherence was associated with reduced risk of cardiovascular disease and all-cause mortality in this sample of Chinese adults, while the positive association with obesity highlights the need to consider total energy intake in future dietary recommendations.
Which is a careful way of saying the food was fine and there was too much of it.
People also ask
What counts as the traditional Chinese diet here?
The researchers used a purpose-built index defined in advance rather than derived from the data, which matters because data-derived patterns can be shaped to fit the outcome. Broadly the traditional pattern is built on rice and other grains, vegetables, legumes and soy foods, modest amounts of fish and pork, and very little dairy, processed food or sugary drink. Higher index scores mean closer adherence, and participants were grouped into low, medium and high bands.
How can a diet be linked to less death and more obesity at once?
Because quantity and composition are different things. A rice-and-vegetables pattern can be nutritionally sound and still deliver a lot of energy if portions are large, and refined white rice in quantity carries a substantial carbohydrate load. The study's own analysis supports this reading: when energy intake was put on a common basis, the obesity association flipped to inverse. In other words the obesity signal appears to have been about how much, not what.
Why is the energy-standardized result the important one?
Because it separates two explanations that the raw result confuses. Without standardizing, you cannot tell whether people following the traditional pattern were heavier because of the pattern or because they simply ate more. Standardizing shows the pattern itself pointed the other way. The authors highlight the need to consider total energy intake in future dietary recommendations, which is the practical lesson.
Does this apply outside China?
The clinical direction may travel; the specifics will not. This is a Chinese cohort eating Chinese food, and the traditional pattern is defined against that food supply. What generalizes is the broader point that a grain-and-vegetable-centered pattern low in processed food tracked with less cardiovascular disease and lower mortality, which is consistent with findings for Mediterranean and Nordic patterns elsewhere.
What are the main weaknesses?
Two stand out. Diet was self-reported, and the follow-up spanned 1997 to 2011, a period in which China's food environment changed faster than almost anywhere on earth, so 'traditional' meant something different at the start than at the end. The design also cannot rule out that people who ate traditionally differed in rural residence, income, activity and smoking in ways adjustment does not fully capture.
References
- Niu J, Li B, Shi Z, Papadaki A. Association between the traditional Chinese diet, non-communicable diseases and all-cause mortality: a longitudinal study based on China Health and Nutrition Survey (CHNS). European Journal of Nutrition (2026).
- World Health Organization. Noncommunicable diseases.
- Harvard T.H. Chan School of Public Health. Healthy Eating Plate.