点击显示 收起
1 From the Nutritional Epidemiology Program, National Institute of Health and Nutrition, Tokyo, Japan (KMurakami, SS, and YT); the Laboratory of Physiological Nutrition, Kagawa Nutrition University, Saitama, Japan (KU); the Department of Health and Nutrition Science, Faculty of Health and Social Welfare Science, Nishikyushu University, Saga, Japan (MY); Department of Human Environmental Science, Fukuoka Women's University, Fukuoka, Japan (HH); Department of Nutrition, School of Food and Nutritional Sciences and COE21, University of Shizuoka, Shizuoka, Japan (TG); Department of Food and Nutrition, Faculty of Home Economics, Tokyo Kasei University, Tokyo, Japan (JO); Department of Nutrition, Mie Chukyo University Junior College, Matsusaka, Japan (KB); Graduate School of Science for Living System, Showa Women's University, Tokyo, Japan (KO); the Department of Food Science and Nutrition, School of Agriculture, Kinki University, Nara, Japan (TK); Course of Food and Nutrition, Department of Human Life Environments, Niigata Women's College, Niigata, Japan (KMuramatsu); and Department of Nutrition Management, Faculty of Health and Nutrition, Minamikyushu University, Miyazaki, Japan (MF)
2 Supported by grants from the Ministry of Health, Labour, and Welfare of Japan. 3 Address reprint requests to S Sasaki, Nutritional Epidemiology Program, National Institute of Health and Nutrition, Toyama 1-23-1, Shinjuku-ku, Tokyo 162-8636, Japan. E-mail: stssasak{at}nih.go.jp.
ABSTRACT
Background: Animal studies suggest the beneficial effect of hardness of diet on body weight and adiposity. No human studies have examined hardness of diet in relation to obesity.
Objective: We examined cross-sectional associations of hardness of the habitual diet with body mass index (BMI; in kg/m2) and waist circumference in free-living humans.
Design: Subjects were 454 female Japanese dietetic students aged 18–22 y. Dietary hardness was assessed as an estimate of masticatory muscle activity for the habitual diet (ie, the difficulty of chewing the food). The consumption of a total of 107 foods was estimated by means of a self-administered, comprehensive diet history questionnaire, and masticatory muscle activity during the ingestion of these foods was estimated according to published equations. Waist circumference was measured at the level of the umbilicus.
Results: Mean BMI was 21.4 (95% CI: 21.1, 21.6), and mean waist circumference was 73.6 (72.9, 74.3) cm. Mean dietary hardness was 178 (175, 181) mVs/1000 kcal. Dietary hardness was not significantly associated with BMI. However, it was negatively associated with waist circumference (P for trend = 0.005). This association remained after adjustment not only for potential confounding factors (P for trend = 0.028) but also for BMI (P for trend = 0.002).
Conclusions: Whereas no association between dietary hardness and BMI was seen, increasing dietary hardness was associated with lower waist circumference even after adjustment for BMI in free-living young Japanese women. This finding could make innovative contributions to the literature and raise issues for future studies regarding diet and obesity.
Key Words: Hardness of diet • body mass index • waist circumference • Japanese • women • diet history questionnaire • epidemiology
INTRODUCTION
Because the human genome has hardly changed since the emergence of behaviorally modern humans 10 000 y ago, contemporary humans are still genetically adapted for the foods consumed by our remote ancestors (1–5). The dietary choices of that time would necessarily have been limited to minimally processed or unprocessed—and, often, uncooked—wild plant and animal foods (2). In contrast, the contemporary diet in affluent societies mainly consists of foods that could not have been regularly consumed before the development of agriculture, industrialization, and advanced technology such as food-processing procedures; these foods include dairy products, cereals, refined cereals, refined sugars, refined vegetable oils, fatty meats, salt, and combinations of these foods (3). The collision of our ancient genome with the new conditions of life in affluent nations, including the dietary qualities of recently introduced foods is considered to be the ultimate factor underlying diseases of civilization, including obesity (3–5). Given that probability, the differences between the ancient dietary patterns and those currently prevalent in industrialized countries appear to have important implications for the prevention and treatment of contemporary chronic diseases, including obesity. A dietary characteristic that would differ greatly between the ancient dietary patterns and the contemporary dietary patterns in developed societies is hardness of the diet, referred to hereafter as dietary hardness.
However, no human studies have examined with diligence the possible association between dietary hardness and diseases of civilization, such as obesity. In contrast, several studies in mice (6, 7) and rats (8, 9) suggested the beneficial effect of dietary hardness on obesity. In this preliminary study, we tried to examine hardness of the habitual diet in relation to body mass index (BMI; in kg/m2) and waist circumference (WC; in cm) among young free-living Japanese women. For this examination, we assessed dietary hardness by using an estimate of masticatory muscle activity for the habitual diet, obtained with data on the consumption of a total of 107 foods estimated by a self-administered comprehensive diet history questionnaire (DHQ) (10–12) and data on masticatory muscle activities during the ingestion of these foods estimated according to published equations (13).
SUBJECTS AND METHODS
Subjects
The present study was based on a multicenter nutritional survey conducted in February and March 2006 among female dietetics students from 10 institutions in Japan. All measurements at each institution were conducted according to the survey protocol. Briefly, staff at each institution explained an outline of the survey to potential subjects. Subjects who responded positively were then provided detailed written and oral explanations of the general purpose and procedure of the survey. A total of 474 women took part. For the present analysis, we selected 454 women who met the following 3 inclusion criteria: they were 18–22 y old (n = 467); were not currently receiving dietary counseling from a doctor or dietitian (n = 468); and had a reported energy intake in the range of 1000–3500 kcal/d (n = 467).
Written informed consent was obtained from each subject, and also from a parent for subjects aged <20 y. The study protocol was approved by the Ethics Committee of the Japanese National Institute of Health and Nutrition (of Japan).
Dietary assessment
Dietary habits during the preceding month were assessed by using a self-administered comprehensive DHQ (10–12). Responses to the DHQ, as well as those to a lifestyle questionnaire, were checked at least twice for completeness. When necessary, forms were reviewed with the subject to ensure the clarity of answers. The DHQ is a 16-page structured questionnaire that consists of 7 sections: general dietary behavior; major cooking methods; consumption frequency and amount of 6 alcoholic beverages; consumption frequency and semiquantitative portion size of 118 selected food and nonalcoholic beverage items; dietary supplements; consumption frequency and semiquantitative portion size of 19 cereals (rice, bread, and noodles), soup consumed with noodles, and miso (fermented soybean paste) soup; and open-ended items for foods consumed regularly (1 time/wk) but not appearing in the DHQ (10). The food and beverage items were selected as foods commonly consumed in Japan, mainly from a food list used in the National Nutrition Survey of Japan, and standard portion sizes were derived mainly from several books of recipes for Japanese dishes (10). Estimates of dietary intake for a total of 150 food and beverage items (including 5 seasonings), energy, and nutrients were calculated by using an ad hoc computer algorithm for the DHQ based on the Standard Tables of Food Composition in Japan (14). Information on dietary supplements and data from the open-ended questionnaire items were not used in the calculation of dietary intake (10). Detailed descriptions of the methods used to calculate dietary intake and the validity of the DHQ with respect to nutrients have been published elsewhere (10–12). Pearson's correlation coefficients between the DHQ and 3-d estimated dietary records were 0.48 for energy and 0.48–0.55 for macronutrients in 47 women (10). In addition, Pearson's correlation coefficient between the DHQ and 16-d weighed dietary records was 0.71 for dietary fiber in 92 women, and the mean value of Spearman's correlation coefficients of food groups was 0.44 (range: 0.13–0.77; S Sasaki, unpublished observations, 2006).
Estimation of dietary hardness
In the present study, dietary hardness was assessed as estimated masticatory muscle activity needed for the habitual diet (ie, the difficulty of chewing the food in the diet). Whereas the habitual diet was assessed by DHQ (10–12) as described above, estimates of masticatory muscle activity for each food in the DHQ were obtained from equations published by Yanagisawa et al (13). Those authors measured the activities of 6 muscle regions (mVs) involved in mastication (right and left masseters and anterior and posterior temporalis) by using electromyography during the ingestion of the same volume (1.3 x 1.3 x 1.3 cm) of 16 selected foods with various physical properties by 20 healthy Japanese adults (10 men and 10 women) with a mean age of 21 y. They found that masticatory muscle activities (mVs/2.197 cm3) were highly correlated with the physical properties of foods (ie, firmness, cohesiveness, and strain) as measured with a texturometer (GTX-2; Zenken KK Inc, Chiba, Japan) and developed the following equations (13):
RESULTS
Basic characteristics of the subjects are shown in Table 2. Mean BMI was 21.4 (95% CI: 21.1, 21.6), and mean WC was 73.6 (72.9, 74.3) cm. Mean dietary hardness was 178 (175, 181) mVs/1000 kcal (range: 101–289 mVs/1000 kcal). The top contributor to dietary hardness was well-milled rice (27.0%), and next were spaghetti (4.1%), pork (3.9%), green leafy vegetables (3.7%), and cabbage (3.4%), as shown in Table 1. Potential confounding factors are shown by quintile of dietary hardness in Table 3. There was a negative association between dietary hardness and rate of eating. Dietary hardness was negatively associated with energy and fat intakes and positively associated with protein and dietary fiber intakes.
View this table:
TABLE 2. Basic characteristics of 454 Japanese women aged 18–22 y
View this table:
TABLE 3. Selected characteristics of 454 Japanese women aged 18–22 y according to quintile (Q) of dietary hardness
Mean values of BMI and WC across quintiles of dietary hardness are shown in Table 4. Dietary hardness was not significantly associated with BMI, regardless of adjustment for potential confounding factors. Conversely, dietary hardness was significantly and negatively associated with WC (in model 1, the mean difference in WC between the lowest and highest quintiles of dietary hardness was –2.9 cm; P for trend = 0.005). The significant negative association between dietary hardness and WC remained after adjustment for potential confounding factors (in model 2, mean difference: –2.7 cm; P for trend = 0.028) and also BMI (in model 4, mean difference: –2.4 cm; P for trend = 0.002). This inverse association seemed mainly due to the composition of the diet, because it disappeared after further adjustment for dietary intake (models 3 and 5).
View this table:
TABLE 4. BMI and waist circumference according to quintile (Q) of dietary hardness in 454 Japanese women aged 18–22 y
DISCUSSION
To our knowledge, this is the first study to examine dietary hardness in relation to BMI and WC in humans. We found that, whereas there was no association with BMI, dietary hardness was negatively associated with WC even after adjustment for BMI in free-living young Japanese women. No human studies have examined the association between dietary hardness and obesity, but several animal studies have suggested the beneficial effect of a hard diet on obesity. Mice fed a hard diet from age 4 wk had a significantly lower body weight at age 36 wk than did mice fed a normal diet (6). In addition, body-weight gain from 4 to 9 wk of age was significantly smaller in male (but not female) mice fed a hard diet than in those fed a soft diet (7), and body-weight gain at age 6 wk was significantly smaller in rats fed a hard diet from age 1 wk than in those fed a soft diet (8). Furthermore, rats fed a hard diet from age 4–26 wk had significantly lower body weight and abdominal white adipose tissue than did those fed a soft diet (9).
We do not know why we found unexpected null association with BMI (but found the expected inverse association with WC). Several limitations of the present study, such as the narrow range of BMIs in the subjects, the study's cross-sectional design, and the use of a new and as yet unestablished method for assessing dietary hardness, may at least partly explain the null finding on BMI. Alternatively, the difference in abdominal white adipose tissue was much larger (22%) than that in body weight (6%) between rats fed a soft and those fed a hard diet (9), which may suggest that dietary hardness affects abdominal obesity (eg, WC) more strongly than it affects overall obesity (eg, BMI).
The negative association between dietary hardness and WC was independent of energy intake. Less body weight gain with a hard diet was related to decreased food intake in a study of rats (8). Conversely, the effect of dietary hardness on obesity was independent of the amount of foods consumed in other studies of mice (6) and rats (9), which may be due to increased thermogenesis (9) or unknown mechanisms. However, the negative association between dietary hardness and WC was not independent of diet composition, because that association disappeared after control for dietary composition. This finding is not consistent with findings from animal studies, because dietary hardness had a beneficial effect on obesity independent of diet composition (6–9). However, the question of whether the association of dietary hardness with obesity is independent of dietary composition should be examined and interpreted with caution, because, whereas dietary hardness can freely be changed in animal models while dietary composition remains constant, dietary hardness is associated with dietary composition in the diet of free-living humans. In the present study, greater dietary hardness was associated with healthier dietary patterns, including lower energy and fat and higher protein and dietary fiber. Several human studies have supported the favorable effects of healthy dietary patterns, including a high intake of dietary fiber (18–21) and a low intake of dietary fat (18, 19), on WC, which does not conflict with our finding.
Several limitations of the present study should be acknowledged. First, our subjects were selected female dietetics students, not a random sample of Japanese women, and the exact response rate was unknown because of our recruitment procedure; these elements of the design may produce recruitment bias. Thus, it may be that our results cannot be extrapolated to the general Japanese population.
Second, because this was a cross-sectional study, reverse causation may have occurred. However, it is unlikely that subjects with a large WC would intentionally change the hardness of their diet as a result of an increase in WC, because the notion that dietary hardness is associated with a measure of obesity is not well known. Furthermore, adjustment for intentional dietary change within the preceding year (yes or no), assessed as part of the DHQ, did not materially change the present results (data not shown). It is therefore reasonable to consider that our findings are not due to reverse causation.
Third, our DHQ was not designed specifically to measure dietary hardness, and the validity of the DHQ with respect to dietary hardness was unknown. The satisfactory validity of the DHQ for a wide range of nutrients and foods (10–12; S Sasaki, unpublished observations, 2006), however, may provide some reassurance. In addition, the DHQ may not adjust sufficiently for cooking methods in the calculation of dietary hardness. Our mean estimate of dietary hardness [crude
Finally, although we attempted to adjust for a wide range of potential confounding variables, we could not rule out residual confounding. Physical activity in particular was assessed relatively roughly from only 5 different activities, a number that may not have been sufficient. In addition, whereas dental status has an influence on food and nutrient intakes and on obesity (28–30), particularly in older persons, we unfortunately had no information on the subjects' dental status, which could confound the present results for young women. Although impaired dental status may be less pervasive in young than in elderly populations, and although the percentage of subjects in a similar population (3828 Japanese female dietetics students aged 18–20 y) who had been diagnosed by a dentist as having a dental disease was relatively small (8%) (S Sasaki, unpublished observations, 2007), further research on dietary hardness and health should take the subjects' dental status into account.
In conclusion, the results of the present study showed that, whereas there was no association between dietary hardness and BMI, dietary hardness was a significant independent determinant of WC in a group of free-living young Japanese women. Because these observations are generally consistent with the results of several animal studies (6–9), the present findings could make innovative contributions to the literature and raise issues for future studies on diet and obesity. However, because this is a preliminary study with a novel, as yet unestablished method of assessing dietary hardness, the results should be interpreted with great caution; nevertheless, applications of the method of assessing dietary hardness to other similar datasets would be of some interest. To better understand the influence of dietary hardness on obesity, further observational and intervention studies are clearly needed. To conduct such investigations, it is urgent to develop a database of values for a variable indicating hardness (eg, masticatory muscle activity) of various food items.
ACKNOWLEDGMENTS
We thank Yukie Yanagisawa (Wayo Women's University) for technical advice regarding the estimation of dietary hardness.
The author's responsibilities were as follows—KMurakami: contributed to the concept and design of the study, the study protocol, data management, and coordinated the field work, calculated the dietary hardness, analyzed and interpreted the data, and wrote the manuscript; SS: the concept and design of the study, the study protocol, and data management, and contributed to the writing and editing of the manuscript; YT: the writing and editing of the manuscript; KU: the concept and design of the study, the study protocol, and data collection; MY, HH, TG, JO, KB, KO, TK, KMuramatsu, and MF: data collection. All authors contributed to the preparation of the manuscript and approved the final version submitted for publication. None of the authors had any personal or financial conflict of interest.
REFERENCES