Obesity and type 2 diabetes are closely linked to diet. The increasing global prevalence of obesity and type 2 diabetes implies that there might have been some common, worldwide changes happened in diet. Indeed, a significant change, happened since around the mid-20th century, is food fortification with B-vitamins. The present ecological study found that the nationwide prevalence of obesity and diabetes in the US in the past 50 years increased in close parallel with the per capita consumption of niacin, thiamin or riboflavin, with a 10-and 26-year lag, respectively. It is obvious that the B-vitamins fortification has been followed first by an increase in the prevalence of obesity, and then by an increase in the prevalence of diabetes. Thus, it seems that the high level consumption of the B vitamins, primarily due to the mandatory grain fortification with the vitamins, may be an attractive candidate for the dietary changes responsible for the increased prevalence of obesity and diabetes.
B-vitamins fortification of grains and the change in high-carbohydrate diet effect
The prevalence of diabetes began to rapidly increase in the US in the 1960s, which was thought to be possibly due to a shift towards a dietary pattern characterized by low fiber and high saturated fats and sugar. Since around the early 1970s, a series of preventive measures have been taken, including reducing consumption of saturated fats and sugar and increasing intake of grains and dietary fiber. Moreover, the standards of B-vitamins fortification were updated in 1974, which has led to a further significant increase in the B-vitamins contents in grain products [14]. Unexpectedly, all of these preventive measures have been followed by a sharp nationwide increase in the prevalence of obesity started from mid-1980s and a second rapid increase in the prevalence of diabetes in the late 1990s. Because one of the most significant changes during this period was the significant increase in the per capita grain consumption, it is suspected that there must be something happened in dietary carbohydrates [11].
Carbohydrates are the main energy source for the body. The role of carbohydrate in the diabetes meal plan remains controversial. Traditionally, high carbohydrate, low fat diets were associated with lower prevalence of obesity and type 2 diabetes in the US in the early 20th century (i.e., before the mandatory B-vitamins fortification), and high-carbohydrate and low-fat diets were used for treating type 2 diabetes [16, 17]. High carbohydrate diets, although inducing hyperlipidemia [33, 34], were still found to improve glucose tolerance in the late 1960s and the early 1970s [35] (i.e., more than 20 years after the implementation of mandatory fortification). However, recently, increasing studies have shown that high carbohydrate diets increase the risk for obesity and type 2 diabetes, and that low carbohydrate diets may be beneficial for preventing obesity and type 2 diabetes in the past decade [8–10, 36] (i.e., about 20 years after update of the fortification standards). Low-carbohydrate diets became a major weight loss and health maintenance trend in the US during the late 1990s and early 2000s [9, 37]. The present study also revealed that the abrupt increase in prevalence of adult obesity was in parallel with the re-increase in the per capita carbohydrate consumption started from the early 1970s with a 10-year lag. It is assumed that the change in carbohydrate-diet effect may involve a change in consuming different type of carbohydrate [11]. Gross et al. suggested that the increased prevalence of diabetes in the US may be due to a high consumption of refined carbohydrates and a lack of fiber [8]. However, the fact is that there was a significant increasing trend in the per capita grain and fiber consumption (Figure 4E and Figure 7E, open cycles) and a decreasing trend in sugar contribution to per capita carbohydrate consumption (Figure 6D, open cycles) since about the early 1970s, which has been followed by a sharp increase, rather than by a decrease, in the prevalence of obesity and diabetes in the following three decades. Moreover, although it is suspected that increasing the consumption of fructose (mainly from beet or cane, high fructose corn syrup, fruits, and honey) may play a role in the prevalence of obesity and type 2 diabetes, there is, however, no unequivocal evidence that fructose intake at moderate doses is directly related with adverse metabolic effects [12]. Thus, it seems unlikely that the change in sugar consumption is responsible for the sharply increasing nationwide prevalence of obesity in the US started since the early 1980s.
Grain products, a major source of carbohydrate, are used as vehicles for the mandatory B-vitamins fortification. The mandatory grain fortification has led to a nationwide increase in B-vitamins intake [14]. Therefore, the adverse effects of B-vitamins fortification, if there are any, should be nationwide. Indeed, the present study found that the increased prevalence of adult obesity and diabetes in the US is highly correlated with the consumption of B-vitamin-fortified grains. Each of the two sharp increases in the vitamin contents, induced respectively by the initiation of the fortification and the update of the fortification standards, was followed by a nationwide increase in the prevalence of diabetes with a 26-year lag. More significantly, the update of grain fortification in 1974 and the subsequent increase use of fortified-grain products was followed by an abrupt increase in the prevalence of obesity among both the adults (Figure 3 and 5) and the children in the US with a 10-year lag [22].
Obesity is known to be associated with excessive energy intake. Indeed, the present population-based study also revealed a high correlation between the obesity prevalence and the per capita energy consumption (Figure 6F). Most of energy consumed by the US population is derived from grains, sugars, meat and fats/oils. The present data clearly showed that the contributions to per capita energy consumption from the known dietary risk factors for obesity and type 2 diabetes, such as meat and animal fats, are not increased or even decreased since the early 1970s. Therefore, it seems unlikely that these known dietary risk factors alone are responsible for the nationwide sharp increase in the prevalence of obesity since the late 1970s. An interesting finding from this analysis was the strong lag-correlation between high obesity prevalence and high fortified-grain contribution to the per capita energy consumption since the early 1970s, which is totally different from the association pattern of high unfortified-grain contribution to energy consumption with very low obesity prevalence in the early 20 century. Increase in fortified-grain contribution to the total energy consumption means an increase in the intake of fortified-grain and B-vitamins, which may lead to an excessive B-vitamin intake. Because B-vitamins can stimulate appetite [15], chronic excess B-vitamins may trigger excessive energy intake, which may contribute to the different outcomes of unfortified-grains and fortified-grains. This interpretation was further supported by the finding that the per capita B-vitamin consumption was lag-correlated not only with the per capita energy consumption but also with the prevalence of obesity and diabetes. Taken together, it seems quite possible that the nationwide increased prevalence of obesity and type 2 diabetes in the US in the late half of 20th century may involve an increase in B-vitamin consumption primarily due to the implementation of mandatory grain fortification with B-vitamins.
It should be noted that the standards of B-vitamins fortification vary from country to country in the world. For example, the level of wheat flour fortification with niacin in the US and the UK is 52.9 mg/kg and 16 mg/kg [13], respectively. Moreover, unlike in the US, the fortification in the UK is voluntary [13]. These differences may underlie regional differences in the study of carbohydrate effect. For example, even in the early 2000, a study from the UK still found that a low-fat, high-carbohydrate diet in overweight individuals with abnormal intermediary metabolism led to moderate weight loss and some improvement in serum cholesterol [38]. Thus, it seems necessary that the content of vitamins should be taken into consideration in the study of the relationship between carbohydrates and the development of obesity and diabetes.
Excess niacin consumption and the obesity and diabetes prevalence
Niacin, one of the most stable of B vitamins, is resistant to heat, light, air, acid, and alkali [39], which means that, once added to grains, little is lost during food processing and cooking [40]. The well-known common adverse effects of niacin are metabolic disturbances, such as insulin resistance and glucose intolerance, and liver injury [15, 19–22], all of which are the hallmarks of obesity and type 2 diabetes [2, 4]. Our previous studies suggested that type 2 diabetes and obesity may involve excess niacin intake [21, 22]. Although the prevalence of obesity and diabetes is also highly correlated with thiamin and riboflavin, however, so far as we know, there is no evidence yet indicating that either thiamin or riboflavin may induce glucose intolerance or insulin resistance [15]. Thus, it seems that the high prevalence of obesity and diabetes may involve niacin consumption.
Human dietary niacin comes mainly from two major sources: animal flesh (meat, poultry and fish) and grains, which accounts for about 70% of dietary niacin consumption in the US in the early 20th century [14]. The amount of the daily per capita niacin consumption from grains and animal flesh in the US was estimated to be 3.7 and 6.8 mg, respectively, in 1930s (just before the introduction of mandatory niacin-fortification), and has increased to 14.8 and 11.8 mg, respectively, in 2000, according the contribution of meat and grain to total niacin given in the literature [14, 22]. The per capita niacin consumption from grains has increased four-fold since the implementation of niacin fortification. By the early 2000s, the US per capita daily niacin consumption has reached 33 mg [23], which is much higher than the RDA (see Introduction) [15]. Thus, long-term excess niacin intake may be very common in the US population after the implementation of mandatory niacin fortification, which may be mainly responsible for rapid increase in the prevalence of obesity and diabetes. According to the regression equations given in Figure 1B and Figure 3B, if the per capita niacin consumption is remained at the current levels (33 mg/d per capita), the prevalence of diabetes in the US would increase from the current about 6% to 7.6% by 2025, whereas the prevalence of obesity in the adults has reached its peak level. In agreement with this prediction, the recent NHANES data have shown that there was no significant change in the prevalence of obesity between 2003-2004 and 2005-2006 for either men or women [41].
B-vitamins fortification and the global increasing obesity and diabetes prevalence
Grain fortification with B vitamins, a strategy for preventing B-vitamins deficiency, was first mandatorily implemented in the US in the early 1940s [42]. Soon after, many other industrialized countries, following the US model, have set up their own B-vitamins fortification programs [13, 39]. During the last few decades, B-vitamins fortification of grains has also been introduced to developing countries [13]. Nowadays, vitamin fortification has become so popular in the world that, besides of grain fortification, other foods, such as most powdered milk and infant milk [43], have already been fortified with niacin and other B vitamins. Moreover, niacin has also been widely used in meat processing to maintain the bright red color of meat [44]. Although there are no data available concerning the relationships between B-vitamins fortification of a variety of foods and the global increasing prevalence of obesity and type 2 diabetes, the notable facts are that: (1) compared with formula-feeding, breastfeeding is associated with a reduction in risk of later overweight and obesity [45–47]; (2) high consumption of processed meat is a strong risk factor for type 2 diabetes [48]; and (3) the global prevalence of obesity and type 2 diabetes showed a trend of spreading from the earliest fortified-countries to the latter fortified-countries, whereas the non-vitamin-fortified countries including western developed countries, such as Norway, have a low prevalence of obesity and diabetes, compared with the earliest fortified-countries, such as the US and Canada [49, 50]. The relevant evidence shows that, when the US was experiencing a rapid increase in the prevalence of diabetes in 1960s (i.e. more than 20 years after the introduction of B-vitamins fortification), the trend in the incidence of diabetes in Norwegian adults was fairly constant, or even significantly decreased in Norwegian women aged 40-59 years [51].
It has been well-recognized that physical inactivity also contributes substantially to the global epidemic of obesity and diabetes. A large scale investigation among the US physicians found that sweat-inducing exercise once a week may effectively reduce the risk of diabetes [52]. Although the exact mechanism of sweat-inducing physical activity is unclear, a well-known fact is that water-soluble B vitamins can be eliminated through sweat [53, 54]. Our recent study also demonstrated that sauna-induced sweating may also effectively eliminate excess nicotinamide from the body and thus reduce the generation of toxic metabolites of nicotinamide [21]. In contrast, excess nicotinamide cannot be effectively eliminated through urine because of its reabsorption by the renal tubules [55]. Therefore, sweating is expected to play an important role in eliminating excess B-vitamins from the body. Unfortunately, modern lifestyle makes the sweat gland less active due to low physical activity and an increase in time spent in air-conditioned environments. The combination of high B-vitamins intake and low sweat elimination of excess B-vitamins may lead to chronic B-vitamins overload. Thus, there is a strong possibility that the worldwide spread of B-vitamins fortification of foods may play a role in the global prevalence of obesity and type 2 diabetes.
The limitations of the current study
Diagnostic criteria are important for estimating the prevalence of diabetes. Because the NHIS survey was redesigned, two changes may have affected trends. First, the diabetes question was changed. Second, proxy respondents (i.e., household members responding for absent adult members) who tend to under report disease were no longer used in the survey [26, 27]. The NHIS survey was redesigned in 1997, and since then gestational diabetes has been excluded. All diagnosed cases of diabetes in the US consist of type 2 diabetes (90% to 95%), type 1 diabetes (5% to 10%), gestational diabetes (2% to 5%) and other specific types of diabetes (1% to 2%) [56]. In this case, the prevalence of diabetes estimated should be lower after 1997 than before 1997 because of the exclusion of gestational diabetes. However, the fact is that the prevalence of diabetes has shown a steadily increasing trend since the late 1990s, and is significantly correlated with the per capita niacin consumption either before or after 1997 (Figure 1A). Moreover, before the significant increase in diabetes prevalence started from the mid-1990s, there had been a sudden increase in the prevalence of obesity in mid-1980s (Figure 3A). The diabetes prevalence is strongly correlated with the obesity prevalence with a time lag of 16 years. Considering that obesity is a major risk factor for type 2 diabetes, it is unlikely that the correlation between the B-vitamins consumption and the prevalence of diabetes is due to a sampling bias.
Ecological studies investigate relationships at the level of the group, rather than at the level of the individual. The inherent limitation of ecological studies is the ecological fallacy, i.e., the data of exposure and disease obtained from populations cannot be linked to individuals. To overcome this shortcoming, we used country-level food and nutrient disappearance data only from within the US, thus, the bias, if any, would at least be uniform for the same population. Also, we addressed the issue from different angles, by which the potential biases due to age or sex had been excluded. More distinctly, although this study is an ecological one, the exposure factor (the B-vitamins consumption) analysed in this study involves every adult inhabitant of the US due to the mandatory grain fortification. Moreover, the US standard of flour fortification with niacin is 5.29 mg/100 g, similar to the niacin content in meat, one of the richest sources of niacin (around 3.6 to 8.2 mg/100 g, depending on the particular type of meat product) [57]. In this case, no matter what dietary pattern is chosen, either high grain diet or high meat diet, the niacin exposure is essentially similar after the implementation of mandatory fortification.
It is important to examine evidence from a variety of sources and to look for congruence between epidemiologic, clinical and laboratory research findings before establishing causality between a diet factor and a human disease [58]. Although the present study provided only correlative evidence linked the B-vitamins consumption to the prevalence of obesity and type 2 diabetes, the following clinical and laboratory research findings provide support for a likely causal relationship between the high level of consumption of the B vitamins and the development of obesity and diabetes: (1) High grain intake-induced increase in the risk of obesity and type 2 diabetes in the US has occurred only after the implementation of B-vitamins fortification, especially after the update of the fortification standards in 1974. However, traditionally, low-fat, high-carbohydrate diets were beneficial for treatment of diabetes. (2) High intake of meat, a niacin-rich food, increases the risk for diabetes [7, 48]. (3) Obesity is a well-known risk factor of type 2 diabetes, and the prevalence of obesity precedes the prevalence of diabetes. The present results showed that the time lag for the prevalence of obesity was much shorter than that for the prevalence of diabetes under the same exposure to the B-vitamins. (4) Niacin is well known to induce glucose intolerance and insulin resistance which are the key features of obesity and type 2 diabetes. (5) The prevalence of obesity and type 2 diabetes has spread in a way similar to that of B-vitamins fortification spread in the world, i.e., from developed countries to developing countries, but the non-vitamin-fortified countries including the developed countries have been less affected. (6) High-niacin feeding has been demonstrated to induce fatty liver in rats [59], and excess niacin intake may be involved in the development of nonalcoholic fatty liver disease, a disease closely associated obesity and type 2 diabetes [60]. (7) Niacin is a potent stimulator of appetite and may play a role in the development of obesity [22]. From the findings of this study, it seems that prospective studies are needed to evaluate the possible role of the high-level consumption of niacin, thiamin and riboflavin in the prevalence of obesity and type 2 diabetes.