2023 Volume 30 Issue 8 Pages 861-862
See article vol. 30: 919-933
Yokoyama conducted an interesting ecological study using National Health and Nutrition Examination Survey of Japan (NHNES), showing continuous increase in high density lipoprotein (HDL) cholesterol in Japanese men and women over 30 years and mainly addressing the relationship between HDL cholesterol increase and its related factors, including meat and fish intake1). The author concluded that this increasing trend is real for the measurement of HDL cholesterol, and it is also not explained by the trend of meat and fish intake, lipid-related medication, and other lifestyle factors.
Yokoyama et al. previously pointed out this phenomenon2) and presently reconfirmed it. The author denied the possibility of artificial measurement bias by two lines of evidence; seamless increase in HDL cholesterol not related to timing of the introduction of new technology and increase in ApoA-I, another independent HDL parameter1). The average HDL cholesterol increased from 51.5 to 57.5 mg/dl and from 58.2 to 68.2mg/dl by 6 and 10 mg/dl or 12% and 17% in men and women, respectively, for the period of analysis of 30 years. In contrast, non-HDL cholesterol and triglyceride showed no chronological change during the same study period. These findings did not cause significant difference in analysis for excluding the data with lipid-related medication.
It is known that low HDL cholesterol is a risk factor for coronary heart diseases, major cardiovascular diseases3, 4), and total mortality5). Yokoyama tried to find associated factors for the increasing trend of HDL cholesterol, including dietary intake, especially fish and meat consumption, and other lifestyle behaviors, such as, smoking, alcohol consumption, body mass index, and exercise. It was observed that omega-3 polyunsaturated fatty acids, which are mainly extracted from fish, are related to HDL cholesterol in men as reported by the International Study of Macronutrients and Blood Pressure (INTERMAP) study, in Japan and Hawaii6). However, Yokoyama did not find any association of fish intake or fish/meat ratio with HDL cholesterol trend based on NHNES data. Smoking, alcohol consumption, and body mass index did not fully support the increasing trend in HDL cholesterol but they were only partially related to the increasing trend of HDL cholesterol. Then, the author concluded that none of them seemed to largely contribute to the steady increase in HDL cholesterol for Japanese men and women.
Yokoyama’s study (the so called ecological study) is fully based on group mean data, and not on individual data. In some cases, ecological study observed the association between chronological stroke mortality and population blood pressure trend in Japan7). The limitation of ecological study is not fully control possible and well-known confounding factors due to the reason not using individual data. If we use the individual data, we can control confounding factors by multiple regression analysis. Therefore, it is not easy to find possible reasons contributing steady increase in HDL cholesterol in Japan by ecological study.
One of the possible approaches may be to use multiple regression method for time series data8); setting the dependent variable for HDL-cholesterol and independent variables for well-known factors such as smoking rate or amount, alcohol consumption, body mass index, dietary intake, etc.
As the increasing trend of HDL cholesterol seems to be real, it is also interesting how this HDL cholesterol increase causes influence on cardiovascular diseases.
The author has no conflict of interest.