Change of dietary intake

Over the past several decades, Korean diet has undergone considerable changes. Several sources of data can be used to examine the change of diet. One such source is the food balance sheet. The amount of rice supplied per person was 213·0 g/d in 2013, 41 % decrease from that in 1980. Meanwhile, the amount of meat supplied per person was 134·8 g/d in 2013, 7 and 256 % increase from those in 2012 and 1980, respectively. The supply of milk and dairy products also increased from 29·5 g per person daily in 1980 to 168·2 g/d in 2013, an increment of 470 %^{( 17 , 18 )}.

Most of the information on food and nutrient intake was obtained from the National Nutrition Survey (1969–1995) and succeeding Korea National Health and Nutrition Examination Survey (1998–present). Based on these data, the proportion of animal product in Korean diet was under 10 % until the mid-1970s but increased to 21 % in 2013. Meat intake increased dramatically, 6·6 g in 1969, 67·8 g in 1995 and 105·0 g in 2013. Milk intake was under 10 g until 1981 and increased rapidly from the mid-1980s and is still gradually increasing, 79·7 g in 1998 and 111·4 g in 2013. However, rice intake has continuously decreased^{( 5 , 19 )} (Fig. 1).

Fig. 1. (Colour online) Secular trend of food intake among Koreans. Grains (), Meats (), Dairy products (), percentage of animal products (shaded). Data are from National Nutrition Survey (1969–1995), Korea National Health and Nutrition Examination Survey (1998–2013)^{( 5 )}.

The change in the composition of food consumption has resulted in significant change of nutrient intake. Fat intake increased from 16·9 g in 1969 to 47·7 g in 2013. With respect to calcium intake, it moderately increased from 444 mg in 1969 to 506·8 mg in 2013. However, carbohydrate intake gradually decreased from 423 g in 1969 to 313·2 g in 2013 (Fig. 2). In addition, in 2013, more than half the population took vitamin C and calcium less than the estimated average requirement in almost every age group, and a high percentage of the elderly had vitamin A and riboflavin less than the estimated average requirement^{( 5 )}.

Fig. 2. (Colour online) Secular trend of nutrient intake among Koreans. Energy (), Fat (), Carbohydrates (), Calcium (). Data are from National Nutrition Survey (1969–1995), Korea National Health and Nutrition Examination Survey (1998–2013)^{( 5 )}.

The aforementioned change was identified as diet transition from a traditional Korean diet that mainly consisted of rice and vegetables to a Westernised diet abundant in meat and milk in previous studies⁽ Reference Kim, Moon and Popkin ^{1 –} Reference Song, Joung and Engelhardt ^{3 )}. Kang et al. reported that the proportion of traditional dietary pattern was decreased over time and was less among the younger generation than in the older generation⁽ Reference Kang, Joung and Lim ^{4 )}.

Change of the status of metabolic syndrome and cancer

In addition to diet transition, a shift in the pattern of diseases towards non-communicable diseases, such as cancer and MetS has occurred in Korea. Based on the Korea National Cancer Incidence Database, overall cancer incidence rate per 100 000 was 445·3 in 2012 which increased by 3·3 % annually from 1999 to 2012 (Fig. 3). Data from Statistics Korea indicated that cancer became the leading cause of deaths, responsible for one in four deaths, in 2012. Lung, liver, stomach and colon/rectum cancers were the main causes of cancer death. Heart disease, diabetes mellitus and hypertensive diseases associated with MetS were also leading causes of death in Korea, accounting for approximately 16 % of all deaths⁽ Reference Jung, Won and Kong ^{20 )}. Prevalence of MetS in Korea was 24·9 % in 1998 and increased to 31·3 % in 2007 (Fig. 4)⁽ Reference Lim, Shin and Song ^{21 )}.

Fig. 3. (Colour online) Secular trend of cancer incidence among Koreans. Data are taken from Jung et al.⁽ Reference Jung, Won and Kong ^{20 )}

Fig. 4. (Colour online) Secular trend of metabolic syndrome prevalence among Koreans. Data are taken from Lim et al.⁽ Reference Lim, Shin and Song ^{21 )}

Meat and milk consumption and cancer and metabolic syndrome risk based on the results from meta-analyses

A joint report by the World Cancer Research Fund and the American Institute for Cancer Research recently concluded that there is convincing evidence on the association between red or processed meat and colorectal cancer, ‘limited to suggestive’ evidence on the association between red meat and cancer of oesophagus, lung, pancreas and endometrium; and between processed meat and cancer of oesophagus, lung, stomach and prostate^{( 22 )}. Meta-analyses of cohort and case–control studies for the association between red meat consumption and stomach cancer risk showed that increased consumption of red meat could be a risk factor for gastric cancer⁽ Reference Song, Lu and Yin ^{7 )} and breast cancer⁽ Reference Guo, Wei and Zhan ^{8 )}. In addition, another meta-analysis of prospective studies reported the significant association between beef consumption and increased risk of colorectal cancer and colon cancer, but no association between pork or poultry and the risk of colorectal adenomas or cancer⁽ Reference Carr, Walter and Brenner ^{9 )}. Poultry intake was rather associated with reduced risk of colorectal cancer⁽ Reference Shi, Yu and Zeng ^{23 )}.

Meat consumption has also been considered as a major risk factor for MetS. Recently, several meta-analyses showed significant associations between red meat or processed meat intake and MetS risk factors or related diseases. Meta-analysis of prospective cohorts⁽ Reference Micha, Michas and Mozaffarian ^{11 )} reported that consumption of processed meats was associated with 42 % higher risk of CHD and 51 % higher risk of diabetes. In addition, consumption of unprocessed red meats was associated with 19 % higher risk of diabetes, but no risk of CHD⁽ Reference Micha, Michas and Mozaffarian ^{11 )}. A meta-analysis of observational studies by Rouhani et al.⁽ Reference Rouhani, Salehi-Abargouei and Surkan ^{10 )} showed consumption of higher amount of red and processed meats increased the risk of obesity.

Dairy consumption has long been thought to play a role in cancer development. Previous meta-analyses noted the association between milk consumption and cancer at specific sites. A meta-analysis of prospective cohort studies on dairy consumption and the risk of breast cancer⁽ Reference Dong, Zhang and He ^{13 )} reported the decreased risk of breast cancer (relative risk (RR) 0·85, 95% CI 0·76, 0·95). Significant dose–response relationship of total dairy food, except milk, intake with breast cancer risk was also reported⁽ Reference Dong, Zhang and He ^{13 )}. For colorectal cancer RR was 0·91 (CI 0·85, 0·94) per 200 g/d of milk intake and 0·83 (CI 0·78, 0·88) per 400 g/d of total dairy products by a meta-analysis of cohort⁽ Reference Aune, Lau and Chan ^{12 )}. Meanwhile, a meta-analysis on dairy consumption and gastric cancer risk⁽ Reference Guo, Shan and Ren ^{24 )} including twenty-three epidemiologic studies suggested that total dairy food, except milk, might be associated with a reduced risk of gastric cancer.

A series of meta-analyses have consistently suggested that increased milk intakes are associated with reduced risk of MetS and MetS-related diseases. A meta-analysis of randomised controlled clinical trials⁽ Reference Abargouei, Janghorbani and Salehi-Marzijarani ^{14 )} examined the effect of dairy consumption on weight and body composition. Dairy consumption led to significantly greater reduction in fat mass and waist circumference and the effect was much greater with energy restriction. Another meta-analysis of randomised controlled trials⁽ Reference Chen, Pan and Malik ^{25 )} also reported beneficial effects of dairy consumption on body weight and fat loss under energy restriction, but the effects were not consistent in long-term or ad libitum studies. In addition, inverse association between low-fat and fluid dairy food consumption and elevated blood pressure was found in a meta-analysis of cohort studies⁽ Reference Ralston, Lee and Truby ^{15 )}. Aune et al.⁽ Reference Aune, Norat and Romundstad ^{16 )} focused on dairy products and type 2 diabetes and concluded that there was an inverse association between consumption of total dairy product and low-fat dairy product, and risk of type 2 diabetes.

Association of meat and milk consumption with cancer and metabolic syndrome in Korea and other Asian countries

From previous meta-analyses, we could conclude that excessive intake of red meat can result in elevated risk of cancer and MetS, whereas high intake of milk and its product is associated with the reduced risk of cancer and MetS. However, because most previous studies on the health effect of meat and milk were conducted in Western populations, there is little understanding on the association between meat and milk intake, and cancer and MetS in Asians. To elucidate the impact of meat and milk consumption on cancer and MetS, epidemiological studies conducted in Asia, including Korea were further reviewed.

The level of meat intake was very low in many Asian populations who had rice-based dietary pattern, but it began to rise rapidly since the late 20th century. Although meat consumption is still relatively low in Asians compared with Western populations, it requires more attention. Meat is a source of not only protein and various vitamins and minerals, but also SFA⁽ Reference Kim, Kim and Ahn ^{26 )} which are well known to be a major risk factor for various chronic diseases⁽ Reference Riccardi, Giacco and Rivellese ^{27 ,} Reference Sieri, Krogh and Ferrari ^{28 )}. Moreover, haem iron and non-haem iron, or some carcinogenic compounds such as heterocyclic amines and polycyclic aromatic hydrocarbons in meat are suspected to increase cancer risk⁽ Reference Fonseca-Nunes, Jakszyn and Agudo ^{29 ,} Reference Knize, Salmon and Pais ^{30 )}.

As in studies involving Western populations, consuming high amount of red meat was reported as a significant risk factor for cancer by a recent Korean cohort study⁽ Reference Wie, Cho and Kang ^{31 )}. Red meat, especially processed red meat, consumption was positively associated with overall cancer incidence in men. Korean males who ate red meat above the World Cancer Research Fund and the American Institute for Cancer Research recommended amount of 43 g/d had 41 % higher risk of overall cancer. Considering the recommended amount of 43 g/d is smaller than one serving size for a Korean, this implies that even small amounts of red meat intake may increase the risk of cancer in Koreans. The researchers explained the reason why the association was significant only in men as that the mean intake of red meat and the percentage of individuals who consumed more than the recommendation were higher in men. Later, a case–control study among Korean adults⁽ Reference Chun, Sohn and Song ^{32 )} added the evidence that red meat intake significantly increased risk for colorectal cancer incidence. However, there was a case–control study⁽ Reference Kim, Chang and Kim ^{33 )} which examined the association of beef or pork intake with gastric cancer and reported insignificant results. Similarly, findings from studies among Chinese population varied. Red meat intake and preserved meat intake were risk factors for bladder cancer⁽ Reference Isa, Xie and Hu ^{34 )}, and fresh meat intake was a risk factor for lung cancer⁽ Reference Shen, Chapman and He ^{35 )}. However, consumption of salt-preserved meats was associated with reduced risk of lung cancer in China⁽ Reference Shen, Chapman and He ^{35 )} (Table 1).

Table 1. Meat consumption and cancer in Korean and other Asian countries

I, incidence; OR, odds ratio; RR, relative risk; HR, hazard ratio; T, tertile.

*Values in parentheses are 95% CI.

Meat consumption seemed to increase the risk of MetS or its components in Asian countries. A series of observational studies among Chinese population demonstrated significant association between intake of red meat and excessive body fat, measured as abdominal obesity, high BMI, and overweight⁽ Reference Wang, Zhang and Zhai ^{36 –} Reference Xu, Yin and Tong ^{38 )}. A large cohort study of Japanese population⁽ Reference Kurotani, Nanri and Goto ^{39 )} reported significant effect of red meat consumption on the development of type 2 diabetes. The associations of processed red meat intake or poultry intake with type 2 diabetes risk were not found in the same study. The impact of meat on MetS among Koreans was not revealed due to limited information. However, considering Asian populations share common characteristics in their diet, it would be reasonable to expect similar results with other Asians (Table 2).

Table 2. Meat consumption and metabolic syndrome in Korean and other Asian countries

P, prevalence; I, incidence; OR, odds ratio; Q, quartile.

*Values in parentheses are 95% CI.

Milk, like meat, has not been an important component of traditional diet in Korea and most Asian countries until the late 20th century. However, milk and dairy foods have now become a crucial part of Korean diet. It is widely recommended to increase milk and dairy food consumption as a good source of calcium and other nutrients. A number of studies conducted in Western countries reported milk and dairy products have health benefits on chronic disease prevention⁽ Reference Aune, Lau and Chan ^{12 –} Reference Aune, Norat and Romundstad ^{16 )}. Milk consumption of Koreans has rapidly increased since the 1980s, but the level of consumption is still relatively lower than that of Western countries. The proportion of people who drink milk less than one serving size daily was 84 %⁽ Reference Lee and Joung ^{40 )} and of those who drink milk less than once per week was 34 %^{( 19 )} in Korea. The difference between frequent milk consumers and infrequent consumers seemed evident.

A substantial number of epidemiological studies aimed to verify the impact of milk on chronic disease risk in Korea and other Asian countries. In Korea, Kim et al.⁽ Reference Kim, Chang and Kim ^{33 )} reported significant inverse association between milk and milk products and gastric cancer risk with case–control study design, whereas Chun et al.⁽ Reference Chun, Sohn and Song ^{32 )} reported a positive association between milk intake and colorectal cancer risk with similar study design. In China, a case–control study on bladder cancer⁽ Reference Isa, Xie and Hu ^{34 )} reported strong health benefits of dairy foods, and another case–control study demonstrated potential protective effects of milk intake on breast cancer⁽ Reference Zhang, Ho and Fu ^{41 )}. In Japan, a case–control study⁽ Reference Mizoue, Kimura and Toyomura ^{42 )} found statistically significant inverse association between calcium and vitamin C intake and colorectal cancer, and potential inverse association between milk and other dairy food intake, and colorectal cancer risk. In a Japanese cohort study⁽ Reference Matsumoto, Ishikawa and Nakamura ^{43 )}, the associations with other cancers were unclear, except the association between increased butter and milk intake and deaths from hematopoietic neoplasm. Another cohort study focusing on prostate cancer⁽ Reference Kurahashi, Inoue and Iwasaki ^{44 )} showed that higher milk intake resulted in higher risk of prostate cancer and suspected calcium and fatty acids as the causative agents (Table 3). The findings of these studies were inconsistent. The results from a few studies on milk and MetS conducted in Asia were consistent. A study using the third Korea National Health and Nutrition Examination Survey data⁽ Reference Kwon, Lee and Park ^{45 )} showed that overweight adults who consume milk more frequently had significantly lower risk of MetS. The inverse relationship was supported by a study by Kim et al.⁽ Reference Kim ^{46 )} which reported significant association between higher consumption of milk or yoghurt and lower risk of MetS based on the fifth Korea National Health and Nutrition Examination Survey data. The study also found the association between higher dairy food intake and lower risk of MetS components, namely abdominal obesity, high fasting glucose and low-HDL cholesterol level. Additionally, the prevalence of MetS was significantly lower in subjects with higher dairy foods. Furthermore, Shin et al.⁽ Reference Shin, Yoon and Lee ^{47 )} analysed the data of middle-aged Koreans from Anseong and Ansan cohort and suggested that dairy product consumption might reduce the risk of MetS, especially abdominal obesity. In subjects who consumed milk and dairy products once daily, the risk of MetS reduced by about 21 and 25 %, respectively. Another study in Japan⁽ Reference Akter, Kurotani and Nanri ^{48 )} revealed that among three types of dairy products classified by fat content, only full-fat dairy products had significant inverse relationship with fasting insulin and homeostasis model of insulin resistance index (Table 4).

Table 3. Milk consumption and cancer in Korean and other Asian countries

I, incidence; M, mortality; OR, odds ratio; RR, relative risk; HR, hazard ratio; T, tertile; Q, quartile.

*Values in parentheses are 95% CI.

Table 4. Milk consumption with metabolic syndrome in Korean and other Asian countries

I, incidence; P, prevalence; HR, hazard ratio; OR, odds ratio; Q, quartile; HOMA-IR, homeostasis model assessment of insulin resistance.

*Values in parentheses are 95% CI.