Obesity Research & Clinical Practice (2007) 1, 79—89
ORIGINAL ARTICLE
Correlation between food intake change patterns and body weight loss in middle-aged women in Japan Makiko Nakade a,∗, Jung Su Lee a, Kiyoshi Kawakubo b, Yuki Amano a, Katsumi Mori a, Akira Akabayashi a a
Department of Health Promotion Sciences, Division of Health Sciences and Nursing, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku 113-0033, Japan b Department of Food Sciences and Nutrition, Kyoritsu Women’s University, 2-2-1 Hitotsubashi, Chiyoda-ku 101-0033, Japan Received 10 May 2006 ; received in revised form 11 January 2007; accepted 30 January 2007
KEYWORDS Overweight; Obesity; Body weight loss; Food intake change patterns; Energy intake
∗
Abstract The patterns of food intake change which are effective for weight loss have not been clearly researched yet. The objective of this study was to examine the correlation between food intake change patterns and body weight loss. One hundred and two overweight women completed a 2-day dietary record before and after a 12week weight-reduction program, and 28 food groups were classified. Patterns of food intake change were derived by cluster analysis, and compared with the changes of physical measurements and nutrition intake. As a result, decreasing Japanese foods pattern (DJP), increasing healthy foods pattern (IHP) and changing staple foods pattern (CSP) were classified. DJP and CSP mainly changed staple food intake. IHP decreased sugars, oils, beans except soybeans and meat intake, and increased fruits and seafood intake. DJP decreased fat and carbohydrate intake most but CSP showed least change. IHP also decreased fat and carbohydrate intake but maintained protein intake. Although no significant differences were seen in the change of energy intake between IHP and DJP or CSP, subjects of IHP showed the largest reduction in mean body weight, BMI, %body fat, waist circumference and serum triacylglycerol after adjusting for age and baseline values. Body weight, BMI and %body fat maintained their significance further adjusting for changes in energy intake and the number of walking steps. Food intake change patterns affected the magnitude of body weight loss independent of energy intake. In addition to energy intake, assessment of food intake change patterns could be useful for effective weight loss. © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +81 3 5841 3618; fax: +81 3 5841 3319.
1871-403X/$ — see front matter © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2007.01.001
80
M. Nakade et al.
Introduction
Subjects and methods
Obesity is currently the most common metabolic disease in the world. Almost all countries are experiencing an obesity epidemic. WHO estimates that more than one billion people are overweight (BMI 25 kg/m2 ), and of these, 300 million are considered obese (BMI 30 kg/m2 ) [1]. The risk of cardiovascular disease, diabetes and hypertension rises continuously with increasing body weight [2]. And, body weight reduction is essential in the prevention and treatment of these diseases [3]. Although the majority of the Japanese have relatively smaller BMI than people in western countries, more than 30% of men 30—69 years old are overweight and overweight men of all ages have been increasing since 20 years ago [4]. Over 30% of women 40 years old and above are overweight, and especially overweight women 60 years old and above have been increasing [4]. On the other hand, energy intake has been decreasing since 20 years ago according to the National Health and Nutrition Survey [4]. Conventional nutrition education for overweight people has been based on the principle of energy intake reduction [5—8]. However, significant relationships have not always been seen between the magnitude of weight loss and the change in energy intake [8]. Recently, dietary pattern analyses which focused on the combination of food intake were conducted in cross-sectional studies [9—15] or longitudinal studies [16—19], and dietary patterns correlated with obesity, coronary heart disease [20,21], colorectal cancer [22,23], and type 2 diabetes [24] have been reported. Furthermore, intervention trials for blood pressure or cardiovascular complications using dietary pattern approaches have been carried out and their effectiveness has been reported [25,26]. There are several studies which examined the correlation between the changes of food intake and weight loss in weight-reduction programs [27,28]. However, these studies focused on the change of a single food group. Change of diet is followed by changes of many kinds of food intake. Therefore, evaluation of effectiveness of patterns of food intake change for body weight loss is needed. The primary aim of this study was to classify the food intake change patterns and examine which food intake change patterns are effective for weight loss in overweight women who participated in a short-term 12-week weight reduction program. The changes in BMI, %body fat, waist and hip circumference, serum lipid, plasma glucose and blood pressure were also examined.
Study subjects Subjects were a total of 102 middle-aged women who completed a weight-reduction program at a municipal health center in Tokyo between 2001 and 2004. The subjects were recruited through municipal advertisement and an eligibility criterion was BMI 24.0 kg/m2 .
Weight-reduction program In the program, professional health educators lectured about weight loss (30 min) once a week for 12 weeks. The lecture topics included the mechanism of weight gain, calorie check of prepared foods, desirable life style, effective exercise for weight loss, and how to prevent a rebound. Subjects were instructed to increase their daily walking steps to ten thousand per day. In addition to the lectures, the subjects received individual dietary advice for reducing energy intake to 1600 kcal/day by dietitians at the beginning of the program, and each subject kept a diary which included a simplified daily dietary record during the program. The dietitians confirmed adherence to the dietary advice by checking the diary every week and encouraged subjects to maintain a reduced energy intake. Aerobic exercises (60 min) were also performed once a week. One goal of the program was more than 5% body weight loss from initial weight.
Physical measurements The following study variables were measured 1 week before and at the 12th week of the program. Height was measured with shoes off and body weight was measured with light clothes to the nearest 0.1 cm and 0.1 kg, respectively, in the fasting state in the morning. BMI was calculated from weight (kg) divided by height (m2 ) of each subject. Waist and hip circumference (cm) were measured using an inelastic tape with the subjects in a standing position at the horizontal level of the navel and the greater trochanter, respectively. Triceps and subscapula subcutaneous fat skinfold thickness (mm) were measured on the right side of the body in a natural standing position with an Eiken Caliper Skinfold Thickness Meter. Percentage of body fat was calculated from the estimated body density for the Japanese equation [29] and the Brozek equation [30]. Blood samples were collected from the anterior cubital vein in the overnight fasting state, and serum total cholesterol (TC), HDL cholesterol (HDL-C), triacylglycerol (TG), and fasting plasma
Correlation between food intake change patterns and body weight loss glucose (FPG) were immediately analyzed in the Foundation of Yobo Igaku Kyoukai (Tokyo, Japan). TC, HDL-C, TG and FPG levels were determined by enzymatic methods using an automatic analyzer (Nippon Denshi BM-12, Japan). Quality control for blood testing was performed every day using pooled standard blood samples. The coefficients of variation of TC, HDL-C, TG and FPG levels obtained using this measurement system were 1.81, 0.64, 1.47, and 1.36, respectively. LDL cholesterol (LDL-C) was calculated using the Friedward equation [31]. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured using a standard mercury sphygmomanometer after the subject had sat at rest more than 10 min. If the blood pressure was high, it was measured again a few minutes later and the lower value was used. The frequency and duration of exercise and smoking habits were obtained by a questionnaire 1 week before the program. Those classified as habitual exercisers were defined as subjects who exercised two times or more per week, 30 min or more at one time and exercise duration was 1 year or more. To determine smoking habits, subjects were asked about smoking states (current smokers/past smokers/non smokers) and those classified as smokers were defined as subjects who were current smokers. The number of walking steps was assessed with a pedometer (EC-200, Yamasa Company, Japan) during the entire period of participation in the program. The subjects were instructed to wear a pedometer all day even during exercise except when sleeping. For analysis, we calculated the mean number of walking steps for 1 week before and at the 12th week of the program. Written information including the purpose of study, use and application of study, assurance of refusal, the benefits and risks, and security of personal information was handled to each participant. All participants signed the consent form to participate in this study.
Dietary assessment A 2-day dietary record was used 1 week before the program and again at the end of the program. Dietitians instructed each subject how to record detailed descriptions of all foods and beverages consumed (ingredients, cooking methods, and information on whether foods were eaten at home or in a restaurant). The dietitians checked the dietary records of each subject and personally clarified any ambiguous information to ensure completeness.
81
The mean intake of food weight per day was considered as the amount of food intake, and energy and nutrition intake of 87 food groups were calculated using nutrition calculation software, Excel eiyou-kun, Version 3.0, for Windows (Kenpakusha, Japan). Twenty-two food groups were formed based on the classification of food groups by the National Nutrition Survey in Japan, then, they were reclassified into 28 food groups (Appendix A). In reclassifying the food groups, the wheat, oils, beverages and seasonings groups were separated into more detailed food groups. The bean group was separated into a soybean group and beans except soybean group. The dairy products group was separated into a regular-fat dairy group and a low-fat dairy group.
Food intake change patterns The mean intake of food for two days for each food group was calculated, and then changes of food intake before and after the program were computed. The change of each food intake was standardized by the standardized score (z), because the portion size was quite different among food groups. To derive food intake change patterns, a cluster analysis was performed using the z-score. Cluster analysis is one of the methods of dietary pattern analysis [32], and the K-means method was used. To decide how many clusters would be used, 2 to 10-cluster solutions were run and a 3cluster solution was selected based on the pseudo F statistic. The food groups which became the characteristic of each cluster were selected based on the criterion that the z-score was 0.5 or more.
Statistical analysis The results were expressed as means ± S.D. To compare the mean of anthropometric data, serum lipid, plasma glucose, blood pressure and nutrition intake before and after the program, the paired t test was used. To compare age, the number of walking steps, body weight and food intake before the program and changes in the number of walking steps, anthropometric data, serum lipid, plasma glucose and nutrition intake among three food intake change patterns, analysis of variance (ANOVA) was used. Moreover, for adjusting for age, baseline values, changes in energy intake and the number of walking steps, analysis of covariance (ANCOVA) was applied. When the P-value was 0.05 or less, Tukey’s multiple comparison test was performed. To compare habitual exercisers and smokers among food intake change patterns, the chi-square test and Fisher’s exact test were used. Differences were considered
82
M. Nakade et al.
significant at P < 0.05. All statistical analyses were carried out using SAS (version 8.2; SAS Institute, Cary, NC) software.
Results Characteristics of the subjects and changes after the program Characteristics of the subjects before the program, and anthropometric data, serum lipid, plasma glucose, blood pressure, the number of walking steps and nutrition intake before and after the program are shown in Table 1. The mean age of the subjects was 51.0 ± 7.6 years, and 41% of the subjects were habitual exercisers, while 12% were smokers. Before the program, the mean body weight of the subjects was 63.6 ± 6.8 kg and BMI was 26.5 ± 2.2 kg/m2 . After the program, the mean body weight and BMI decreased 3.4 ± 2.1 kg and 1.4 ± 0.8 kg/m2 , respectively. Body weight, BMI, %body fat, waist and hip circumference decreased significantly after the Table 1
program. And also TC, LDL-C, TG, FPG, SBP and DBP decreased significantly but HDL-C did not. The number of walking steps increased significantly after the program. Intake of energy, fat, carbohydrate and protein decreased significantly. Energy% from fat and carbohydrate also decreased, but was not significant. Energy% from protein increased significantly. Dietary fiber intake (g) did not change, but dietary fiber intake (g/1000 kcal) increased significantly.
Characteristics of food intake change patterns Three food intake change patterns were derived by cluster analysis. The mean change in food intake of 28 food groups is shown in Table 2. Subjects of the first pattern decreased intake of rice and miso, and increased intake of noodles and beans except soybeans. Because rice and miso are main foods in Japanese diet, this pattern was named ‘‘Decreasing Japanese foods Pattern (DJP)’’.
Characteristics of study subjects before and after the weight-reduction program Before
Age (years) Exercise habit [n (%)] Smoking habit [n (%)]
After
51.0 ± 7.6 42 (41) 12 (12)
— — — — 60.1 ± 25.1 ± 33.7 ± 87.2 ± 93.7 ±
Height (cm) Body weight (kg) BMI (kg/m2 ) Body fat (%) Waist circumference (cm) Hip circumference (cm)
154.7 63.6 26.5 37.6 91.7 96.3
± ± ± ± ± ±
5.6 6.8 2.2 6.9 7.7 4.4
TC (mg/dl) HDL-C (mg/dl) LDL-C (mg/dl)a TG (mg/dl) FPG (mg/dl) SBP (mmHg) DBP (mmHg)
231.3 62.7 145.8 113.1 95.0 123.9 74.5
± ± ± ± ± ± ±
35.1 14.3 31.5 70.1 14.1 15.6 10.6
Number of walking steps per day Energy (kcal) Fat (g) Carbohydrate (g) Protein (g) Dietary fiber (g) Fat (%) Carbohydrate (%) Protein (%) Dietary fiber (g/1000 kcal) Values are means ± S.D. Paired t-test; * P < 0.01, ** P < 0.001. a LDL-C was calculated when subject’s TG < 400 mg/dl (n = 101).
7341 ± 3188 1853 57.1 248.9 71.6 16.2 27.5 57.0 15.5 8.8
± ± ± ± ± ± ± ± ±
371 19.3 55.7 17.1 5.0 6.7 7.6 2.0 2.3
208.0 63.8 127.9 78.2 91.8 116.7 71.5
± ± ± ± ± ± ±
6.4* 2.2* 6.8* 7.5* 4.4* 32.3* 14.0 27.5* 49.5* 13.8* 13.8* 10.1*
9929 ± 3418* 1541 47.0 204.5 66.1 16.0 27.2 55.6 17.2 10.7
± ± ± ± ± ± ± ± ±
293* 15.2* 42.1* 14.2** 4.2 5.6 5.9 2.4* 3.4*
Correlation between food intake change patterns and body weight loss Table 2
83
Changes in food intake among food intake change patterns by cluster analysis
Changes in food intake (g) Rice Bread, wheat flour Noodles Other grains Nuts and seeds Potatoes Sugars and sweeteners Sweets Butter Vegetable oils Miso Soybeans Beans except soybeans Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms Seaweed Seafood Meats Regular-fat dairy Low-fat dairy Eggs Alcohol Other beverages Seasonings Mayonnaise
Food intake change patterns
Means of changes in food intake (g) (n = 102)
DJP (n = 36)
IHP (n = 27)
CSP (n = 39)
−92.8 −3.4 16.1 0.1 −8.3 −0.8 0.0 −18.7 −0.9 0.0 −5.3 −3.8 9.6 −86.8 10.9
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
113.6 (−0.6) 35.7 (0.0) 82.2 (0.6) 2.2 (0.0) 18.1 (−0.4) 56.8 (−0.1) 11.0 (0.1) 44.8 (0.0) 5.4 (0.0) 4.6 (0.0) 10.8 (−0.6) 82.7 (0.0) 19.4 (0.6) 148.9 (−0.4) 58.3 (−0.4)
−18.3 −17.1 −64.3 −2.2 −2.3 11.9 −7.2 −30.2 −3.9 −6.1 3.0 −19.3 −9.1 44.1 76.2
26.5 3.4 −74.1 1.7 −0.4 10.4 2.5 −27.8 0.8 3.4 3.3 6.1 −3.7 −23.0 53.1
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
111.6 (0.5) 28.3 (0.2) 73.0 (−0.5) 6.5 (0.3) 4.7 (0.2) 68.5 (0.1) 10.3 (0.3) 28.4 (−0.2) 3.3 (0.4) 7.3 (0.5) 6.2 (0.3) 84.6 (0.1) 12.5 (−0.2) 92.5 (0.0) 80.3 (0.1)
−8.7 −1.8 1.6 −2.4 −12.0 −28.2 8.1 36.5 −3.1 −45.5 −85.3 −5.1 −3.6
± ± ± ± ± ± ± ± ± ± ± ± ±
106.4 (−0.2) −39.3 ± 95.3 (−0.4) 18.6 (0.2) −4.0 ± 19.5 (0.2) 33.2 (0.0) 6.1 ± 23.8 (0.1) 13.2 (−0.4) 14.6 ± 22.3 (0.4) 49.9 (−0.2) 52.5 ± 55.9 (0.8) 40.2 (−0.2) −50.9 ± 48.7 (−0.7) 142.7 (0.1) −49.5 ± 169.5 (−0.4) 72.8 (0.1) 55.6 ± 112.3 (0.3) 35.5 (−0.2) −4.3 ± 33.0 (−0.2) 151.1 (0.0) −81.5 ± 280.2 (−0.2) 139.0 (0.2) −212.1 ± 315.6 (−0.3) 20.9 (−0.4) 6.6 ± 23.1 (0.1) 8.0 (−0.2) −1.5 ± 8.3 (0.0)
58.0 −14.1 0.9 9.5 −10.0 3.9 20.3 16.7 14.1 −23.7 −98.3 9.8 −0.9
± ± ± ± ± ± ± ± ± ± ± ± ±
120.3 (0.4) 8.7 ± 115.6 36.2 (−0.2) −7.1 ± 27.3 19.6 (−0.1) 2.5 ± 26.0 25.0 (0.2) 6.7 ± 21.8 61.4 (−0.1) 5.9 ± 62.3 42.9 (0.4) −22.0 ± 48.6 109.9 (0.2) −2.5 ± 140.8 54.5 (−0.2) 34.0 ± 80.3 28.6 (0.3) 3.2 ± 33.2 95.8 (0.1) −46.7 ± 179.2 254.9 (0.1) −123.8 ± 244.2 22.8 (0.3) 3.7 ± 23.0 6.8 (0.1) −2.0 ± 7.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
74.5 (0.1) 43.0 (−0.4) 64.2 (−0.4) 8.7 (−0.4) 14.6 (0.1) 51.3 (0.1) 12.2 (−0.5) 35.7 (−0.3) 4.5 (−0.6) 7.3 (−0.8) 8.4 (0.3) 80.4 (−0.2) 13.1 (−0.6) 128.6 (0.5) 81.2 (0.4)
−27.4 −4.5 −39.6 0.1 −3.7 6.8 −1.0 −25.2 −1.0 −0.3 0.2 −4.1 −0.5 −27.7 44.3
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
115.1 35.8 84.5 6.3 13.7 59.9 11.7 36.8 4.8 7.4 9.5 82.6 17.1 133.4 77.5
Values are means ± S.D. (z-score). Underlined values were the absolute z-scores were 0.5 or more. DJP, decreasing Japanese foods pattern; IHP, increasing healthy foods pattern; CSP, changing staple foods pattern.
Subjects of the second pattern decreased intake of sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats, and increased intake of fruits and seafood. This pattern decreased intake of energy-dense foods such as sugars and oils, and increased fruits. The pattern also increased seafood intake and decreased meat intake. Therefore, it was named ‘‘Increasing Healthy foods Pattern (IHP)’’. Subjects of the third pattern decreased intake of noodles, and increased intake of rice and vegetable oils, and showed relatively smaller changes in other food groups. This pattern was named ‘‘Changing Staple foods Pattern (CSP)’’.
Comparing baseline characteristics among three food intake change patterns Age, exercise and smoking habits, the number of walking steps, body weight and food intake among the three patterns were compared (Table 3).
Subjects with IHP were significantly younger than DJP. However, no significant differences were seen in exercise habit or smoking habit among the three patterns. There was no significant difference in the number of walking steps among the three patterns, although the number of walking steps in IHP tended to be less than the other patterns. Body weight was not significantly different among the patterns. These results did not change after adjusting for age. The other anthropometrical and clinical data were also not significantly different (data not shown). Intake of sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats was greater in IHP than the other patterns and significant differences were seen in the intake of sugars and sweeteners and vegetable oils between IHP and the other two patterns, while butter and meats were significantly different between IHP and CSP. Bean except soybean intake in IHP was significantly greater than in DJP. However, fruit intake in IHP was
84 Table 3
M. Nakade et al. Comparison of baseline characteristics with food intake change patterns Food intake change patterns DJPb (n = 36)
Age (years) Exercise habit [n (%)] Smoking habit [n (%)] Number of walking steps per day Body weight (kg) Rice Bread, wheat flour Noodles Other grains Nuts and seeds Potatoes Sugars and sweeteners Sweets Butter Vegetable oils Miso Soybeans Beans except soybeans Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms Seaweed Seafood Meats Regular-fat dairy Low-fat dairy Eggs Alcohol Other beverages Seasonings Mayonnaise
53.6 ± 7.8 15 (42) 4 (11) 7452 ± 3681
IHPc (n = 27)
Tukey’s test
Tukey’s testa
b—c NSe NSf NS
NS
CSPd (n = 39)
47.8 ± 9.0 14 (52) 1 (4) 7025 ± 3100
50.9 ± 5.5 13 (50) 7 (18) 7457 ± 2806
61.9 280.4 41.0 47.9 0.3 9.7 43.0 8.9 33.3 4.0 8.0 12.9 66.3 0.6 187.6 97.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
5.8 137.0 34.5 57.2 1.2 17.7 42.7 6.3 41.7 4.2 4.8 10.5 67.3 2.6 152.2 48.4
64.4 226.8 58.8 91.7 2.7 5.4 41.2 16.8 39.3 5.6 12.3 8.1 79.4 9.5 122.7 93.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
7.3 95.4 31.3 63.9 8.7 13.0 36.6 11.3 36.5 4.0 6.9 6.4 72.2 12.7 87.5 56.7
64.5 252.1 39.0 97.6 0.0 2.5 35.5 9.5 38.3 2.5 6.7 7.5 64.7 5.8 104.4 76.5
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
7.3 104.3 30.1 89.3 0.0 4.3 44.3 8.4 26.4 2.9 4.6 5.2 66.0 11.3 94.3 62.4
NS NS NS b—d c—d b—d NS b—c,c—d NS c—d b—c,c—d b—c,b—d NS b—c c—d NS
NS NS NS b—d c—d NS NS b—c,c—d NS c—d b—c,c—d b—c,b—d NS b—c c—d NS
143.7 16.5 15.8 12.2 89.4 75.3 110.2 0.0 27.6 70.3 238.8 39.1 5.5
± ± ± ± ± ± ± ± ± ± ± ± ±
92.7 15.8 17.9 15.2 52.4 42.6 95.2 0.0 26.9 154.8 243.3 19.4 7.9
171.4 19.0 12.9 6.8 77.8 93.7 159.2 0.0 37.5 145.6 327.9 37.9 5.2
± ± ± ± ± ± ± ± ± ± ± ± ±
75.5 18.5 16.5 6.9 44.5 45.7 139.6 0.0 23.9 385.0 302.7 17.7 6.6
105.1 27.7 12.1 10.5 75.5 53.4 119.4 0.0 27.4 43.8 221.7 34.3 4.2
± ± ± ± ± ± ± ± ± ± ± ± ±
56.7 35.7 12.7 13.5 50.3 34.1 100.9 0.0 22.4 99.4 256.1 14.7 5.4
c—d NS NS NS NS c—d NS NS NS NS NS NS NS
c—d NS NS NS NS c—d NS NS NS NS NS NS NS
Values are means ± S.D. a Adjusted for age. Tukey’s test; p < 0.05. NS: not significant. b DJP: decreasing Japanese foods pattern. c IHP: increasing healthy foods pattern. d CSP: changing staple foods pattern. e Chi-square test. f Fisher’s exact test.
not significantly different from those of the other patterns and there was no difference in seafood intake among the three patterns. Rice intake was not significantly different among the three patterns and noodle intake in DJP was significantly less than in CSP. Miso intake in DJP was significantly greater than in IHP and CSP. Other grains and other vegetables in IHP were significantly greater than in CSP, and nut and seed intake in DJP was significantly greater than in CSP, although these foods were not characteristic of the food intake change patterns.
After adjusting for age, only nut and seed intake was not significant.
Comparing the change in physical activity, nutrition intake and clinical characteristics among three food intake change patterns The changes in the number of walking steps, nutrition intake, anthropometric data, serum lipid, plasma glucose and blood pressure among the three patterns were compared (Table 4).
Correlation between food intake change patterns and body weight loss
85
Table 4 Comparison of changes in physical activity, nutrition intake and clinical characteristics with food intake change patterns Tukey’s testa
Food intake change patterns DJPc (n = 36) Number of walking steps per dayf Energy (kcal)f Fat (g)f Carbohydrate (g)f Protein (g)f Dietary fiber (g)f Fat (%)f Carbohydrate (%)f Protein (%)f Dietary fiber (g/1000 kcal)f Body weight (kg)f BMI (kg/m2 )f Body fat (%)f Waist circumference (cm)f Hip circumference (cm)f TC (mg/dl)f HDL-C (mg/dl)f LDL-C (mg/dl)f,g TG (mg/dl)f FPG (mg/dl)f SBP (mmHg)f DBP (mmHg)f
2506 ± 3958
IHPd (n = 27) 3551 ± 2558
Tukey’s testb
CSPe (n = 39) 2603 ± 2871
NS
−430 −14.5 −60.8 −11.8 −1.4 −1.1 −0.6 1.5 1.6
± ± ± ± ± ± ± ± ±
50 2.9 7.1 2.2 0.7 5.3 5.0 0.5 0.6
−366 −13.2 −48.1 −0.4 0.8 −3.7 −0.4 3.5 2.6
± ± ± ± ± ± ± ± ±
61 3.7 8.3 2.7 0.8 6.2 5.9 0.5 0.7
−199 −1.4 −32.5 −4.4 −0.1 −8.3 7.6 1.2 1.6
± ± ± ± ± ± ± ± ±
48 2.9 6.7 2.1 0.6 5.1 4.8 0.4 0.5
c—e c—e, d—e c—e c—d, c—e NS NS NS c—d, d—e NS
−3.0 −1.2 −3.8 −4.4
± ± ± ±
0.3 0.1 0.5 0.6
−4.5 −1.9 −5.5 −6.0
± ± ± ±
0.4 0.1 0.6 0.7
−3.1 −1.3 −2.7 −3.7
± ± ± ±
0.3 0.1 0.5 0.6
c—d, d—e c—d, d—e d—e d—e
c—d c—d d—e NS
NS
NS
NS NS NS d—e NS NS NS
NS NS NS NS NS NS NS
−1.9 ± 0.4 −22.8 2.9 −20.7 −33.2 −3.1 −7.0 −2.6
± ± ± ± ± ± ±
4.1 1.4 3.2 5.6 1.5 1.5 1.2
−3.3 ± 0.4 −31.0 0.2 −21.5 −50.5 −4.3 −8.8 −4.2
± ± ± ± ± ± ±
4.8 1.6 3.7 6.5 1.8 1.8 1.4
−2.6 ± 0.3 −18.5 0.1 −13.1 −25.7 −2.6 −6.4 −2.4
± ± ± ± ± ± ±
3.9 1.3 3.0 5.3 1.4 1.4 1.1
Values are means ± S.D. NS: not significant. a Adjusted for age and baseline values. Tukey’s test; p < 0.05. NS: not significant. b Adjusted for age, baseline values, change in energy intake, change in the number of walking steps. c DJP: decreasing Japanese foods pattern. d IHP: increasing healthy foods pattern. e CSP: changing staple foods pattern. f Changes before and after the program. g LDL-C was calculated when subject’s TG < 400 mg/dl (n = 101).
The change in the number of walking steps was not significantly different among the three patterns. Adjustment for age and baseline the number of walking steps did not change this finding. In DJP, intake of energy, fat, carbohydrate, protein and dietary fiber (g) decreased most. On the other hand, CSP showed less change in energy, fat and carbohydrate intake than DJP. Although no significant differences were seen in changes of energy, fat, and carbohydrate intake between DJP and IHP, IHP showed less change in protein intake than DJP. Furthermore, IHP increased the most energy% from protein than the other two patterns. There were no significant differences in change of energy% from fat, carbohydrate, dietary fiber intake (g) and dietary fiber intake (g/1000 kcal) among the three patterns.
In IHP, body weight, BMI, %body fat, and waist circumference decreased most, and significant differences were seen in body weight and BMI decreases compared with the other patterns even after adjusting for age and baseline values. Furthermore, %body fat and waist circumference decrease of subjects with IHP were significantly greater than CSP. After adjusting for changes in energy intake and the number of walking steps in addition to age and baseline values, IHP had greater body weight and BMI decrease than DJP, and %body fat decrease was significantly greater than CSP. There was no significant difference in hip circumference change among the three patterns. Decreases in TC, LDL-C, TG, FPG, SBP and DBP were greatest in IHP, however, significant difference was seen only in the TG decrease between IHP and CSP.
86
Discussion The primary objective of this study was to examine which food intake change patterns are effective for body weight loss. The values of the changes of food intake before and after the program for each food group were standardized, and a cluster analysis was performed. As a result, three food intake change patterns were derived. The mean body weight decreased in all of the food intake change patterns. However, IHP, which decreased intake of energy-dense foods such as sugars, oils and meats, and increased intake of fruits and seafood, reduced anthropometric data the most, and significant differences were seen in body weight, BMI, and %body fat decrease after adjusting for age, baseline values, the changes in energy intake and the number of walking steps compared with DJP or CSP. DJP and CSP mostly showed that change in staple foods like rice or noodles and it was different from IHP. In the program, participants received individual dietary advice to reduce energy intake, and the changes of food groups were left to each participant. IHP showed more change in energy intake than CSP, and anthropometric data, serum lipid, plasma glucose and blood pressure improved more. On the other hand, there was no significant difference in the change of energy intake between DJP and IHP, but IHP saw more decreases in body weight, BMI, other anthropometric data and serum lipids than DJP. This result that decreasing energy intake and weight loss are not always correlated with each other, is an interesting observation, and is also consistent with a previous study [8]. Therefore, the results of this study support the idea that it would be important to evaluate not only energy intake but also food intake change patterns for weight reduction of overweight people. A decreasing of protein intake in DJP was due to a decrease in seafood and meat intake. On the other hand, IHP decreased meat intake but increased seafood intake, thus protein intake was relatively maintained. Okuda et al. [27] compared changes of a single food intake between a successful weight loss group (weight loss was 2.0 kg and more) and an unsuccessful group (weight loss was less than 2.0 kg) in a weight-reduction program in Japan. The successful group increased fish intake and a significant difference was seen only in the change of fish intake. In this study, an increase of seafood intake was seen in the greatest weight loss group (IHP) and the result was consistent with the previous study. As for protein intake and weight loss, WesterterpPlantenga et al. [33] suggested that high protein intake sustained body weight maintenance after
M. Nakade et al. weight loss in humans. Furthermore, diet-induced thermogenesis of protein is highest [34]. High dietinduced thermogenesis may lead to more energy expenditure. Therefore, IHP might be more effective for weight loss and the maintenance of weight loss than the other patterns during the program. In this study, IHP which had the greatest weight loss also decreased meat and oil intake, and increased fruit and seafood intake. Ledikwe et al. [14] and Quatromoni et al. [17] reported that dietary patterns which included meats and oils were correlated with high BMI, and patterns which included fruits and fish were correlated with low BMI. These results were adjusted for energy intake. Hence, these combinations of foods might be independent of energy intake and associated with BMI. Thus, our results are similar to these previous studies and these combinations of food would be important for effective weight loss. However, a previous study of dietary patterns reported poultry intake was correlated with low BMI, and processed meat was correlated with high BMI [14]. Another previous study reported that lean poultry intake was correlated with a lower risk for becoming overweight and meat except for lean poultry was correlated with a higher risk for becoming overweight [17]. Therefore, different kinds of meats may be associated with obesity. To investigate these questions, it may be necessary to examine food intake change patterns using more specific meat groups. IHP decreased intake of beans except soybeans. Many previous studies of dietary patterns reported that bean intake was associated with low BMI [9,11,14,17]. This was not consistent with the results of the current study. It may be due to a difference in the portion size of bean intake among ethnic groups. Maskarinec et al. [9] compared dietary patterns across ethnic groups; a high bean intake pattern was very common among Japanese women. In the current study, the total amount of beans except soybeans and soybean intake for each pattern before the program was 66.9 g, 88.9 g, and 70.5 g per day, respectively. On the other hand, in a previous study in America [14] in which a high-nutrient-dense pattern was associated with low BMI, bean intake was very low compared with the results of this study. Hence, in many previous studies, increasing bean intake might be associated with low BMI, while in the current study, decreasing bean intake was associated with weight loss. The result of this study also suggested that different kinds of beans (soybeans or beans except soybeans) had different effects on weight loss, because the difference was seen only in the change of beans except soybean intake. In Japan, beans except
Correlation between food intake change patterns and body weight loss soybeans are usually used for sweet desserts with sugar. Many studies reported that in addition to these foods, sweets were associated with high BMI [14,17] and low-fat dairy was associated with low BMI [16,19]. But in this study, these food changes were not different among the three patterns, because the mean intake of sweets decreased and low-fat dairy intake increased in all of the patterns. The weight-reduction program in this study was only for 12 weeks, but the improvements of serum lipid, plasma glucose and blood pressure were large in all of the patterns, therefore we could not detect significant differences except TG. Significant difference was seen in TG reduction between IHP and CSP. There were no significant differences in baseline exercise habit, smoking habit, the number of walking steps, body weight, the other anthropometrical and clinical data among the three patterns. However, baseline food intake was different among the patterns. Food groups whose baseline intake was greater or less compared with the other patterns became characteristic of each pattern (e.g., noodles, miso and beans except soybeans in DJP or sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats in IHP or noodles and vegetable oils in CSP). On the other hand, baseline rice and seafood intake were not significantly different among the three patterns and fruit intake in IHP was not significantly different from those of the other patterns. Therefore, most of these food groups changed depend on the amount of baseline intake while rice intake in DJP and fruit and seafood intake in IHP changed after the program, these results might be associated with the degree of body weight loss in each food intake change patterns. The authors were not aware of any studies which derived patterns using change of food intake and examined the correlation between the patterns and body weight loss. The results of this study suggested that there were significant differences in the magnitude of weight loss among food intake change patterns even if change in energy intake was not significantly different. Although cluster analysis is a method used to classify subjects based on their similarity in food intake change, some subjects had the food groups whose food intake change was in the opposite direction of the pattern characteristic. We defined these food groups as unfit food groups, and classified subjects who had at least two unfit food groups into ‘‘less fit’’ and the other subjects into ‘‘more fit’’. There were no significant differences in body weight change or change in metabolic parameters between less fit and more fit in DJP and IHP;
87
and although significant difference was seen in the change of HDL-C in CSP, it improved more in more fit subjects than less fit subjects (data not shown). The results showed that food intake change patterns derived by cluster analysis were useful, although there were some unfit food groups. In this study, there were four subjects with an increase in body weight after the program. The mean weight gain of these four subjects (weight gain group) was 0.9 ± 0.8 kg (range 0.3—2.0 kg). The weight gain group was older, had a lower body weight and a greater number of walking steps than the weight loss group (n = 98) at baseline. There was no difference in energy intake. After the program, the change in energy intake was not different between the groups. However, the weight gain group decreased the number of walking steps while the weight loss group increased it. Therefore, this might be associated with the increasing of body weight (data not shown). This study has some limitations. This study used a dietary record because it is a gold standard. There are several studies which used a 7-day dietary record [16,19], but in this study, only a 2-day dietary record was used, so habitual food intake is not reflected. However, many studies of dietary patterns use food frequency questionnaires [9—12,15,17,18] and the kinds of foods they can investigate are limited. On the other hand, the current study investigated all of the foods which subjects ate. Physical activity was assessed only by the number of walking steps, thus the effect of exercise may not be well evaluated. However, the study subjects performed the same exercise during the program. Therefore, we evaluated the change in physical activity by step count data. The subjects in this study were also limited. The subjects were only women and most of them were middle aged, so other food intake change patterns may be derived in case of men or other age brackets. Thus, more studies are needed using a various subjects. In summary, this study examined the correlation between food intake change patterns and body weight loss, and three food intake patterns were derived. The mean body weight decreased in all of the food intake change patterns. Among the three patterns, body weight loss was significantly greater in IHP, which decreased intake of sugars, oils, beans except soybeans and meats, and increased intake of fruits and seafood, than DJP or CSP, although the change in energy intake was not significantly different. In conclusion, nutrition education and assessment of food intake change patterns in addition to energy intake could be useful for effective weight loss.
88
M. Nakade et al. Appendix A (Continued)
Acknowledgement Contibutions. A part of this study was supported by a grant from the Foundation for Total Health Promotion (Title: Daily Dietary Patterns and Obesity; Principal investigator: Jung Su LEE). We would like to express our thanks to the study participants and to the Koto Health Promotion Center for their cooperation.
22 Food groups
28 Food groups
Examples of foods in the groups
Meats
Meats
Beef, pork, poultry, processed meat
Dairy products
Regular-fat dairy
Regular-fat milk and dairy products Low-fat milk and dairy products
Low-fat dairy
Appendix A. Food groups used in the food intake change pattern analysis in the study 22 Food groups
28 Food groups
Examples of foods in the groups
Rice
Rice
Rice
Wheat
Bread, wheat flour Noodles
Other grains Nuts and seeds
Other grains Nuts and seeds
Potatoes
Potatoes
Sugars and sweeteners Sweets
Sugars and sweeteners Sweets
Bread, wheat flour Soba, udon noodles, pasta, Chinese noodles Cereals Chestnut, sesame, other nuts and seeds Sweet potato, white potato Sugar, honey
Oils
Butter Vegetable oils
Butter Vegetable oils, margarine
Miso
Miso
Miso (soybean paste)
Beans
Soybeans
Soybean and soybean products except miso Adzuki beans, kidney beans
Beans except soybeans Fruits Brightly colored vegetables
Pickles Mushrooms
Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms
Seaweed
Seaweed
Seafood
Seafood
Other vegetables
Cake, cookies
Fruits, fruit juice Carrots, spinach
Lettuce, onions Pickles Shiitake mushroom Laver, wakame seaweed Fish, shellfish, processed fish
Eggs
Eggs
Eggs, processed eggs
Beverages
Alcohol Other beverages
Sake, beer, wine Coffee, tea, green tea
Seasonings
Seasonings
Soy sauce, dressing without oil Mayonnaise, dressing with oil
Mayonnaise
References [1] Formiguera X, Canton A. Obesity: epidemiology and clinical aspects. Best Pract Res Clin Gastroenterol 2004;18: 1125—46. [2] World Health Organization. Diet, nutrition and the prevention of chronic diseases. Report of a joint WHO/FAO expert consultation, WHO Technical Report Series No. 916; 2003. Geneva. [3] Sowers JR. Obesity as a cardiovascular risk factor. Am J Med 2003;115(Suppl.):37S—41S. [4] Ministry of Health, Labour and Welfare. The results of the national nutrition survey in Japan, 2002. Japan: Daiichi Publishers; 2004 [in Japanese]. [5] Japan Society for the Study of Obesity. Manual for the treatment of obesity. 2nd ed. Tokyo: Ishiyaku Publishers; 2003 [in Japanese]. [6] Tetsuguchi M, Abe K, Mimura K. Effects of the exercise and nutritional prescription on body composition in middle age obese woman. J Educ Health Sci 2000;46:1012—9 [in Japanese]. [7] Imaizumi T, Asami T. Effect of weight reduction program on visceral and subcutaneous fat tissues in white collar male workers. Bull Phys Fitness Res Inst 1995;89:24—31 [in Japanese]. [8] Andersson I, Lennernas M, Rossner S. Meal pattern and risk factor evaluation in one-year completers of a weight reduction program for obese men—–the ‘Gustaf’ study. J Intern Med 2000;247:30—8. [9] Maskarinec G, Novotny R, Tasaki K. Dietary patterns are associated with body mass index in multiethnic women. J Nutr 2000;130:3068—72. [10] Gittelsohn J, Wolever TM, Harris SB, Harris-Giraldo R, Hanley AJ, Zinman B. Specific patterns of food consumption and preparation are associated with diabetes and obesity in a Native Canadian community. J Nutr 1998;128: 541—7.
Correlation between food intake change patterns and body weight loss [11] Sichieri R. Dietary patterns and their associations with obesity in the Brazilian city of Rio de Janeiro. Obes Res 2002;10:42—8. [12] Wirfalt AK, Jeffery RW. Using cluster analysis to examine dietary patterns: nutrient intakes, gender, and weight status differ across food pattern clusters. J Am Diet Assoc 1997;97:272—9. [13] Wirfalt E, Hedblad B, Gullberg B, Mattisson I, Andren C, Rosander U, et al. Food patterns and components of the metabolic syndrome in men and women: a cross-sectional study within the Malmo Diet and Cancer cohort. Am J Epidemiol 2001;154:1150—9. [14] Ledikwe JH, Smiciklas-Wright H, Mitchell DC, Miller CK, Jensen GL. Dietary patterns of rural older adults are associated with weight and nutritional status. J Am Geriatr Soc 2004;52:589—95. [15] Lin H, Bermudez OI, Tucker KL. Dietary patterns of Hispanic elders are associated with acculturation and obesity. J Nutr 2003;133:3651—7. [16] Newby PK, Muller D, Hallfrisch J, Qiao N, Andres R, Tucker KL. Dietary patterns and changes in body mass index and waist circumference in adults. Am J Clin Nutr 2003;77:1417—25. [17] Quatromoni PA, Copenhafer DL, D’Agostino RB, Millen BE. Dietary patterns predict the development of overweight in women: The Framingham Nutrition Studies. J Am Diet Assoc 2002;102:1239—46. [18] Togo P, Osler M, Sorensen TI, Heitmann BL. A longitudinal study of food intake patterns and obesity in adult Danish men and women. Int J Obes Relat Metab Disord 2004;28:583—93. [19] Newby PK, Muller D, Hallfrisch J, Andres R, Tucker KL. Food patterns measured by factor analysis and anthropometric changes in adults. Am J Clin Nutr 2004;80:504—13. [20] Hu FB, Rimm EB, Stampfer MJ, Ascherio A, Spiegelman D, Willett WC. Prospective study of major dietary patterns and risk of coronary heart disease in men. Am J Clin Nutr 2000;72:912—21. [21] Fung TT, Willett WC, Stampfer MJ, Manson JE, Hu FB. Dietary patterns and the risk of coronary heart disease in women. Arch Intern Med 2001;161:1857—62. [22] Fung T, Hu FB, Fuchs C, Giovannucci E, Hunter DJ, Stampfer MJ, et al. Major dietary patterns and the risk of colorectal cancer in women. Arch Intern Med 2003;163:309—14.
89
[23] Slattery ML, Boucher KM, Caan BJ, Potter JD, Ma KN. Eating patterns and risk of colon cancer. Am J Epidemiol 1998;148:4—16. [24] van Dam RM, Rimm EB, Willett WC, Stampfer MJ, Hu FB. Dietary patterns and risk for type 2 diabetes mellitus in U.S. men. Ann Intern Med 2002;136:201—9. [25] Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, et al. A clinical trial of the effects of dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med 1997;336:1117—24. [26] de Lorgeril M, Salen P, Martin JL, Monjaud I, Delaye J, Mamelle N. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation 1999;99:779—85. [27] Okuda N, Okamura T, Kadowaki T, Tanaka T, Uesima H. Weight-control intervention in overweight subjects at high risk of cardiovascular disease: A trial of a public health practical training program in a medical school. Jpn J Public Health 2004;51:552—9 [in Japanese]. [28] Nisigami T, Imamura H, Moriwaki T, Yoshimura Y, Shimada R, Nishimura T, et al. Effects of dietary education and physical exercise on mildly obese and hypertensive women, vol. 24. Japan Society of Health Evaluation and Promotion; 1997. p. 375—81 [in Japanese]. [29] Nagamine S, Suzuki S. Anthropometry and body composition of Japanese young men and women. Hum Biol 1964;36:8—15. [30] Brozek J, Grande F, Anderson JT, Keys A. Densitometric analysis of body composition: revision of some quantitative assumption. Ann NY Acad Sci 1963;110:113—40. [31] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499—502. [32] Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 2002;13: 3—9. [33] Westerterp-Plantenga MS, Lejeune MP, Nijs I, van Ooijen M, Kovacs EM. High protein intake sustains weight maintenance after body weight loss in humans. Int J Obes Relat Metab Disord 2004;28:57—64. [34] de Jonge L, Bray GA. The thermic effect of food and obesity: a critical review. Obes Res 1997;5:622—3.
ORIGINAL ARTICLE
Correlation between food intake change patterns and body weight loss in middle-aged women in Japan Makiko Nakade a,∗, Jung Su Lee a, Kiyoshi Kawakubo b, Yuki Amano a, Katsumi Mori a, Akira Akabayashi a a
Department of Health Promotion Sciences, Division of Health Sciences and Nursing, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku 113-0033, Japan b Department of Food Sciences and Nutrition, Kyoritsu Women’s University, 2-2-1 Hitotsubashi, Chiyoda-ku 101-0033, Japan Received 10 May 2006 ; received in revised form 11 January 2007; accepted 30 January 2007
KEYWORDS Overweight; Obesity; Body weight loss; Food intake change patterns; Energy intake
∗
Abstract The patterns of food intake change which are effective for weight loss have not been clearly researched yet. The objective of this study was to examine the correlation between food intake change patterns and body weight loss. One hundred and two overweight women completed a 2-day dietary record before and after a 12week weight-reduction program, and 28 food groups were classified. Patterns of food intake change were derived by cluster analysis, and compared with the changes of physical measurements and nutrition intake. As a result, decreasing Japanese foods pattern (DJP), increasing healthy foods pattern (IHP) and changing staple foods pattern (CSP) were classified. DJP and CSP mainly changed staple food intake. IHP decreased sugars, oils, beans except soybeans and meat intake, and increased fruits and seafood intake. DJP decreased fat and carbohydrate intake most but CSP showed least change. IHP also decreased fat and carbohydrate intake but maintained protein intake. Although no significant differences were seen in the change of energy intake between IHP and DJP or CSP, subjects of IHP showed the largest reduction in mean body weight, BMI, %body fat, waist circumference and serum triacylglycerol after adjusting for age and baseline values. Body weight, BMI and %body fat maintained their significance further adjusting for changes in energy intake and the number of walking steps. Food intake change patterns affected the magnitude of body weight loss independent of energy intake. In addition to energy intake, assessment of food intake change patterns could be useful for effective weight loss. © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +81 3 5841 3618; fax: +81 3 5841 3319.
1871-403X/$ — see front matter © 2007 Asian Oceanian Association for the Study of Obesity. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.orcp.2007.01.001
80
M. Nakade et al.
Introduction
Subjects and methods
Obesity is currently the most common metabolic disease in the world. Almost all countries are experiencing an obesity epidemic. WHO estimates that more than one billion people are overweight (BMI 25 kg/m2 ), and of these, 300 million are considered obese (BMI 30 kg/m2 ) [1]. The risk of cardiovascular disease, diabetes and hypertension rises continuously with increasing body weight [2]. And, body weight reduction is essential in the prevention and treatment of these diseases [3]. Although the majority of the Japanese have relatively smaller BMI than people in western countries, more than 30% of men 30—69 years old are overweight and overweight men of all ages have been increasing since 20 years ago [4]. Over 30% of women 40 years old and above are overweight, and especially overweight women 60 years old and above have been increasing [4]. On the other hand, energy intake has been decreasing since 20 years ago according to the National Health and Nutrition Survey [4]. Conventional nutrition education for overweight people has been based on the principle of energy intake reduction [5—8]. However, significant relationships have not always been seen between the magnitude of weight loss and the change in energy intake [8]. Recently, dietary pattern analyses which focused on the combination of food intake were conducted in cross-sectional studies [9—15] or longitudinal studies [16—19], and dietary patterns correlated with obesity, coronary heart disease [20,21], colorectal cancer [22,23], and type 2 diabetes [24] have been reported. Furthermore, intervention trials for blood pressure or cardiovascular complications using dietary pattern approaches have been carried out and their effectiveness has been reported [25,26]. There are several studies which examined the correlation between the changes of food intake and weight loss in weight-reduction programs [27,28]. However, these studies focused on the change of a single food group. Change of diet is followed by changes of many kinds of food intake. Therefore, evaluation of effectiveness of patterns of food intake change for body weight loss is needed. The primary aim of this study was to classify the food intake change patterns and examine which food intake change patterns are effective for weight loss in overweight women who participated in a short-term 12-week weight reduction program. The changes in BMI, %body fat, waist and hip circumference, serum lipid, plasma glucose and blood pressure were also examined.
Study subjects Subjects were a total of 102 middle-aged women who completed a weight-reduction program at a municipal health center in Tokyo between 2001 and 2004. The subjects were recruited through municipal advertisement and an eligibility criterion was BMI 24.0 kg/m2 .
Weight-reduction program In the program, professional health educators lectured about weight loss (30 min) once a week for 12 weeks. The lecture topics included the mechanism of weight gain, calorie check of prepared foods, desirable life style, effective exercise for weight loss, and how to prevent a rebound. Subjects were instructed to increase their daily walking steps to ten thousand per day. In addition to the lectures, the subjects received individual dietary advice for reducing energy intake to 1600 kcal/day by dietitians at the beginning of the program, and each subject kept a diary which included a simplified daily dietary record during the program. The dietitians confirmed adherence to the dietary advice by checking the diary every week and encouraged subjects to maintain a reduced energy intake. Aerobic exercises (60 min) were also performed once a week. One goal of the program was more than 5% body weight loss from initial weight.
Physical measurements The following study variables were measured 1 week before and at the 12th week of the program. Height was measured with shoes off and body weight was measured with light clothes to the nearest 0.1 cm and 0.1 kg, respectively, in the fasting state in the morning. BMI was calculated from weight (kg) divided by height (m2 ) of each subject. Waist and hip circumference (cm) were measured using an inelastic tape with the subjects in a standing position at the horizontal level of the navel and the greater trochanter, respectively. Triceps and subscapula subcutaneous fat skinfold thickness (mm) were measured on the right side of the body in a natural standing position with an Eiken Caliper Skinfold Thickness Meter. Percentage of body fat was calculated from the estimated body density for the Japanese equation [29] and the Brozek equation [30]. Blood samples were collected from the anterior cubital vein in the overnight fasting state, and serum total cholesterol (TC), HDL cholesterol (HDL-C), triacylglycerol (TG), and fasting plasma
Correlation between food intake change patterns and body weight loss glucose (FPG) were immediately analyzed in the Foundation of Yobo Igaku Kyoukai (Tokyo, Japan). TC, HDL-C, TG and FPG levels were determined by enzymatic methods using an automatic analyzer (Nippon Denshi BM-12, Japan). Quality control for blood testing was performed every day using pooled standard blood samples. The coefficients of variation of TC, HDL-C, TG and FPG levels obtained using this measurement system were 1.81, 0.64, 1.47, and 1.36, respectively. LDL cholesterol (LDL-C) was calculated using the Friedward equation [31]. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured using a standard mercury sphygmomanometer after the subject had sat at rest more than 10 min. If the blood pressure was high, it was measured again a few minutes later and the lower value was used. The frequency and duration of exercise and smoking habits were obtained by a questionnaire 1 week before the program. Those classified as habitual exercisers were defined as subjects who exercised two times or more per week, 30 min or more at one time and exercise duration was 1 year or more. To determine smoking habits, subjects were asked about smoking states (current smokers/past smokers/non smokers) and those classified as smokers were defined as subjects who were current smokers. The number of walking steps was assessed with a pedometer (EC-200, Yamasa Company, Japan) during the entire period of participation in the program. The subjects were instructed to wear a pedometer all day even during exercise except when sleeping. For analysis, we calculated the mean number of walking steps for 1 week before and at the 12th week of the program. Written information including the purpose of study, use and application of study, assurance of refusal, the benefits and risks, and security of personal information was handled to each participant. All participants signed the consent form to participate in this study.
Dietary assessment A 2-day dietary record was used 1 week before the program and again at the end of the program. Dietitians instructed each subject how to record detailed descriptions of all foods and beverages consumed (ingredients, cooking methods, and information on whether foods were eaten at home or in a restaurant). The dietitians checked the dietary records of each subject and personally clarified any ambiguous information to ensure completeness.
81
The mean intake of food weight per day was considered as the amount of food intake, and energy and nutrition intake of 87 food groups were calculated using nutrition calculation software, Excel eiyou-kun, Version 3.0, for Windows (Kenpakusha, Japan). Twenty-two food groups were formed based on the classification of food groups by the National Nutrition Survey in Japan, then, they were reclassified into 28 food groups (Appendix A). In reclassifying the food groups, the wheat, oils, beverages and seasonings groups were separated into more detailed food groups. The bean group was separated into a soybean group and beans except soybean group. The dairy products group was separated into a regular-fat dairy group and a low-fat dairy group.
Food intake change patterns The mean intake of food for two days for each food group was calculated, and then changes of food intake before and after the program were computed. The change of each food intake was standardized by the standardized score (z), because the portion size was quite different among food groups. To derive food intake change patterns, a cluster analysis was performed using the z-score. Cluster analysis is one of the methods of dietary pattern analysis [32], and the K-means method was used. To decide how many clusters would be used, 2 to 10-cluster solutions were run and a 3cluster solution was selected based on the pseudo F statistic. The food groups which became the characteristic of each cluster were selected based on the criterion that the z-score was 0.5 or more.
Statistical analysis The results were expressed as means ± S.D. To compare the mean of anthropometric data, serum lipid, plasma glucose, blood pressure and nutrition intake before and after the program, the paired t test was used. To compare age, the number of walking steps, body weight and food intake before the program and changes in the number of walking steps, anthropometric data, serum lipid, plasma glucose and nutrition intake among three food intake change patterns, analysis of variance (ANOVA) was used. Moreover, for adjusting for age, baseline values, changes in energy intake and the number of walking steps, analysis of covariance (ANCOVA) was applied. When the P-value was 0.05 or less, Tukey’s multiple comparison test was performed. To compare habitual exercisers and smokers among food intake change patterns, the chi-square test and Fisher’s exact test were used. Differences were considered
82
M. Nakade et al.
significant at P < 0.05. All statistical analyses were carried out using SAS (version 8.2; SAS Institute, Cary, NC) software.
Results Characteristics of the subjects and changes after the program Characteristics of the subjects before the program, and anthropometric data, serum lipid, plasma glucose, blood pressure, the number of walking steps and nutrition intake before and after the program are shown in Table 1. The mean age of the subjects was 51.0 ± 7.6 years, and 41% of the subjects were habitual exercisers, while 12% were smokers. Before the program, the mean body weight of the subjects was 63.6 ± 6.8 kg and BMI was 26.5 ± 2.2 kg/m2 . After the program, the mean body weight and BMI decreased 3.4 ± 2.1 kg and 1.4 ± 0.8 kg/m2 , respectively. Body weight, BMI, %body fat, waist and hip circumference decreased significantly after the Table 1
program. And also TC, LDL-C, TG, FPG, SBP and DBP decreased significantly but HDL-C did not. The number of walking steps increased significantly after the program. Intake of energy, fat, carbohydrate and protein decreased significantly. Energy% from fat and carbohydrate also decreased, but was not significant. Energy% from protein increased significantly. Dietary fiber intake (g) did not change, but dietary fiber intake (g/1000 kcal) increased significantly.
Characteristics of food intake change patterns Three food intake change patterns were derived by cluster analysis. The mean change in food intake of 28 food groups is shown in Table 2. Subjects of the first pattern decreased intake of rice and miso, and increased intake of noodles and beans except soybeans. Because rice and miso are main foods in Japanese diet, this pattern was named ‘‘Decreasing Japanese foods Pattern (DJP)’’.
Characteristics of study subjects before and after the weight-reduction program Before
Age (years) Exercise habit [n (%)] Smoking habit [n (%)]
After
51.0 ± 7.6 42 (41) 12 (12)
— — — — 60.1 ± 25.1 ± 33.7 ± 87.2 ± 93.7 ±
Height (cm) Body weight (kg) BMI (kg/m2 ) Body fat (%) Waist circumference (cm) Hip circumference (cm)
154.7 63.6 26.5 37.6 91.7 96.3
± ± ± ± ± ±
5.6 6.8 2.2 6.9 7.7 4.4
TC (mg/dl) HDL-C (mg/dl) LDL-C (mg/dl)a TG (mg/dl) FPG (mg/dl) SBP (mmHg) DBP (mmHg)
231.3 62.7 145.8 113.1 95.0 123.9 74.5
± ± ± ± ± ± ±
35.1 14.3 31.5 70.1 14.1 15.6 10.6
Number of walking steps per day Energy (kcal) Fat (g) Carbohydrate (g) Protein (g) Dietary fiber (g) Fat (%) Carbohydrate (%) Protein (%) Dietary fiber (g/1000 kcal) Values are means ± S.D. Paired t-test; * P < 0.01, ** P < 0.001. a LDL-C was calculated when subject’s TG < 400 mg/dl (n = 101).
7341 ± 3188 1853 57.1 248.9 71.6 16.2 27.5 57.0 15.5 8.8
± ± ± ± ± ± ± ± ±
371 19.3 55.7 17.1 5.0 6.7 7.6 2.0 2.3
208.0 63.8 127.9 78.2 91.8 116.7 71.5
± ± ± ± ± ± ±
6.4* 2.2* 6.8* 7.5* 4.4* 32.3* 14.0 27.5* 49.5* 13.8* 13.8* 10.1*
9929 ± 3418* 1541 47.0 204.5 66.1 16.0 27.2 55.6 17.2 10.7
± ± ± ± ± ± ± ± ±
293* 15.2* 42.1* 14.2** 4.2 5.6 5.9 2.4* 3.4*
Correlation between food intake change patterns and body weight loss Table 2
83
Changes in food intake among food intake change patterns by cluster analysis
Changes in food intake (g) Rice Bread, wheat flour Noodles Other grains Nuts and seeds Potatoes Sugars and sweeteners Sweets Butter Vegetable oils Miso Soybeans Beans except soybeans Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms Seaweed Seafood Meats Regular-fat dairy Low-fat dairy Eggs Alcohol Other beverages Seasonings Mayonnaise
Food intake change patterns
Means of changes in food intake (g) (n = 102)
DJP (n = 36)
IHP (n = 27)
CSP (n = 39)
−92.8 −3.4 16.1 0.1 −8.3 −0.8 0.0 −18.7 −0.9 0.0 −5.3 −3.8 9.6 −86.8 10.9
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
113.6 (−0.6) 35.7 (0.0) 82.2 (0.6) 2.2 (0.0) 18.1 (−0.4) 56.8 (−0.1) 11.0 (0.1) 44.8 (0.0) 5.4 (0.0) 4.6 (0.0) 10.8 (−0.6) 82.7 (0.0) 19.4 (0.6) 148.9 (−0.4) 58.3 (−0.4)
−18.3 −17.1 −64.3 −2.2 −2.3 11.9 −7.2 −30.2 −3.9 −6.1 3.0 −19.3 −9.1 44.1 76.2
26.5 3.4 −74.1 1.7 −0.4 10.4 2.5 −27.8 0.8 3.4 3.3 6.1 −3.7 −23.0 53.1
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
111.6 (0.5) 28.3 (0.2) 73.0 (−0.5) 6.5 (0.3) 4.7 (0.2) 68.5 (0.1) 10.3 (0.3) 28.4 (−0.2) 3.3 (0.4) 7.3 (0.5) 6.2 (0.3) 84.6 (0.1) 12.5 (−0.2) 92.5 (0.0) 80.3 (0.1)
−8.7 −1.8 1.6 −2.4 −12.0 −28.2 8.1 36.5 −3.1 −45.5 −85.3 −5.1 −3.6
± ± ± ± ± ± ± ± ± ± ± ± ±
106.4 (−0.2) −39.3 ± 95.3 (−0.4) 18.6 (0.2) −4.0 ± 19.5 (0.2) 33.2 (0.0) 6.1 ± 23.8 (0.1) 13.2 (−0.4) 14.6 ± 22.3 (0.4) 49.9 (−0.2) 52.5 ± 55.9 (0.8) 40.2 (−0.2) −50.9 ± 48.7 (−0.7) 142.7 (0.1) −49.5 ± 169.5 (−0.4) 72.8 (0.1) 55.6 ± 112.3 (0.3) 35.5 (−0.2) −4.3 ± 33.0 (−0.2) 151.1 (0.0) −81.5 ± 280.2 (−0.2) 139.0 (0.2) −212.1 ± 315.6 (−0.3) 20.9 (−0.4) 6.6 ± 23.1 (0.1) 8.0 (−0.2) −1.5 ± 8.3 (0.0)
58.0 −14.1 0.9 9.5 −10.0 3.9 20.3 16.7 14.1 −23.7 −98.3 9.8 −0.9
± ± ± ± ± ± ± ± ± ± ± ± ±
120.3 (0.4) 8.7 ± 115.6 36.2 (−0.2) −7.1 ± 27.3 19.6 (−0.1) 2.5 ± 26.0 25.0 (0.2) 6.7 ± 21.8 61.4 (−0.1) 5.9 ± 62.3 42.9 (0.4) −22.0 ± 48.6 109.9 (0.2) −2.5 ± 140.8 54.5 (−0.2) 34.0 ± 80.3 28.6 (0.3) 3.2 ± 33.2 95.8 (0.1) −46.7 ± 179.2 254.9 (0.1) −123.8 ± 244.2 22.8 (0.3) 3.7 ± 23.0 6.8 (0.1) −2.0 ± 7.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
74.5 (0.1) 43.0 (−0.4) 64.2 (−0.4) 8.7 (−0.4) 14.6 (0.1) 51.3 (0.1) 12.2 (−0.5) 35.7 (−0.3) 4.5 (−0.6) 7.3 (−0.8) 8.4 (0.3) 80.4 (−0.2) 13.1 (−0.6) 128.6 (0.5) 81.2 (0.4)
−27.4 −4.5 −39.6 0.1 −3.7 6.8 −1.0 −25.2 −1.0 −0.3 0.2 −4.1 −0.5 −27.7 44.3
± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
115.1 35.8 84.5 6.3 13.7 59.9 11.7 36.8 4.8 7.4 9.5 82.6 17.1 133.4 77.5
Values are means ± S.D. (z-score). Underlined values were the absolute z-scores were 0.5 or more. DJP, decreasing Japanese foods pattern; IHP, increasing healthy foods pattern; CSP, changing staple foods pattern.
Subjects of the second pattern decreased intake of sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats, and increased intake of fruits and seafood. This pattern decreased intake of energy-dense foods such as sugars and oils, and increased fruits. The pattern also increased seafood intake and decreased meat intake. Therefore, it was named ‘‘Increasing Healthy foods Pattern (IHP)’’. Subjects of the third pattern decreased intake of noodles, and increased intake of rice and vegetable oils, and showed relatively smaller changes in other food groups. This pattern was named ‘‘Changing Staple foods Pattern (CSP)’’.
Comparing baseline characteristics among three food intake change patterns Age, exercise and smoking habits, the number of walking steps, body weight and food intake among the three patterns were compared (Table 3).
Subjects with IHP were significantly younger than DJP. However, no significant differences were seen in exercise habit or smoking habit among the three patterns. There was no significant difference in the number of walking steps among the three patterns, although the number of walking steps in IHP tended to be less than the other patterns. Body weight was not significantly different among the patterns. These results did not change after adjusting for age. The other anthropometrical and clinical data were also not significantly different (data not shown). Intake of sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats was greater in IHP than the other patterns and significant differences were seen in the intake of sugars and sweeteners and vegetable oils between IHP and the other two patterns, while butter and meats were significantly different between IHP and CSP. Bean except soybean intake in IHP was significantly greater than in DJP. However, fruit intake in IHP was
84 Table 3
M. Nakade et al. Comparison of baseline characteristics with food intake change patterns Food intake change patterns DJPb (n = 36)
Age (years) Exercise habit [n (%)] Smoking habit [n (%)] Number of walking steps per day Body weight (kg) Rice Bread, wheat flour Noodles Other grains Nuts and seeds Potatoes Sugars and sweeteners Sweets Butter Vegetable oils Miso Soybeans Beans except soybeans Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms Seaweed Seafood Meats Regular-fat dairy Low-fat dairy Eggs Alcohol Other beverages Seasonings Mayonnaise
53.6 ± 7.8 15 (42) 4 (11) 7452 ± 3681
IHPc (n = 27)
Tukey’s test
Tukey’s testa
b—c NSe NSf NS
NS
CSPd (n = 39)
47.8 ± 9.0 14 (52) 1 (4) 7025 ± 3100
50.9 ± 5.5 13 (50) 7 (18) 7457 ± 2806
61.9 280.4 41.0 47.9 0.3 9.7 43.0 8.9 33.3 4.0 8.0 12.9 66.3 0.6 187.6 97.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
5.8 137.0 34.5 57.2 1.2 17.7 42.7 6.3 41.7 4.2 4.8 10.5 67.3 2.6 152.2 48.4
64.4 226.8 58.8 91.7 2.7 5.4 41.2 16.8 39.3 5.6 12.3 8.1 79.4 9.5 122.7 93.7
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
7.3 95.4 31.3 63.9 8.7 13.0 36.6 11.3 36.5 4.0 6.9 6.4 72.2 12.7 87.5 56.7
64.5 252.1 39.0 97.6 0.0 2.5 35.5 9.5 38.3 2.5 6.7 7.5 64.7 5.8 104.4 76.5
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
7.3 104.3 30.1 89.3 0.0 4.3 44.3 8.4 26.4 2.9 4.6 5.2 66.0 11.3 94.3 62.4
NS NS NS b—d c—d b—d NS b—c,c—d NS c—d b—c,c—d b—c,b—d NS b—c c—d NS
NS NS NS b—d c—d NS NS b—c,c—d NS c—d b—c,c—d b—c,b—d NS b—c c—d NS
143.7 16.5 15.8 12.2 89.4 75.3 110.2 0.0 27.6 70.3 238.8 39.1 5.5
± ± ± ± ± ± ± ± ± ± ± ± ±
92.7 15.8 17.9 15.2 52.4 42.6 95.2 0.0 26.9 154.8 243.3 19.4 7.9
171.4 19.0 12.9 6.8 77.8 93.7 159.2 0.0 37.5 145.6 327.9 37.9 5.2
± ± ± ± ± ± ± ± ± ± ± ± ±
75.5 18.5 16.5 6.9 44.5 45.7 139.6 0.0 23.9 385.0 302.7 17.7 6.6
105.1 27.7 12.1 10.5 75.5 53.4 119.4 0.0 27.4 43.8 221.7 34.3 4.2
± ± ± ± ± ± ± ± ± ± ± ± ±
56.7 35.7 12.7 13.5 50.3 34.1 100.9 0.0 22.4 99.4 256.1 14.7 5.4
c—d NS NS NS NS c—d NS NS NS NS NS NS NS
c—d NS NS NS NS c—d NS NS NS NS NS NS NS
Values are means ± S.D. a Adjusted for age. Tukey’s test; p < 0.05. NS: not significant. b DJP: decreasing Japanese foods pattern. c IHP: increasing healthy foods pattern. d CSP: changing staple foods pattern. e Chi-square test. f Fisher’s exact test.
not significantly different from those of the other patterns and there was no difference in seafood intake among the three patterns. Rice intake was not significantly different among the three patterns and noodle intake in DJP was significantly less than in CSP. Miso intake in DJP was significantly greater than in IHP and CSP. Other grains and other vegetables in IHP were significantly greater than in CSP, and nut and seed intake in DJP was significantly greater than in CSP, although these foods were not characteristic of the food intake change patterns.
After adjusting for age, only nut and seed intake was not significant.
Comparing the change in physical activity, nutrition intake and clinical characteristics among three food intake change patterns The changes in the number of walking steps, nutrition intake, anthropometric data, serum lipid, plasma glucose and blood pressure among the three patterns were compared (Table 4).
Correlation between food intake change patterns and body weight loss
85
Table 4 Comparison of changes in physical activity, nutrition intake and clinical characteristics with food intake change patterns Tukey’s testa
Food intake change patterns DJPc (n = 36) Number of walking steps per dayf Energy (kcal)f Fat (g)f Carbohydrate (g)f Protein (g)f Dietary fiber (g)f Fat (%)f Carbohydrate (%)f Protein (%)f Dietary fiber (g/1000 kcal)f Body weight (kg)f BMI (kg/m2 )f Body fat (%)f Waist circumference (cm)f Hip circumference (cm)f TC (mg/dl)f HDL-C (mg/dl)f LDL-C (mg/dl)f,g TG (mg/dl)f FPG (mg/dl)f SBP (mmHg)f DBP (mmHg)f
2506 ± 3958
IHPd (n = 27) 3551 ± 2558
Tukey’s testb
CSPe (n = 39) 2603 ± 2871
NS
−430 −14.5 −60.8 −11.8 −1.4 −1.1 −0.6 1.5 1.6
± ± ± ± ± ± ± ± ±
50 2.9 7.1 2.2 0.7 5.3 5.0 0.5 0.6
−366 −13.2 −48.1 −0.4 0.8 −3.7 −0.4 3.5 2.6
± ± ± ± ± ± ± ± ±
61 3.7 8.3 2.7 0.8 6.2 5.9 0.5 0.7
−199 −1.4 −32.5 −4.4 −0.1 −8.3 7.6 1.2 1.6
± ± ± ± ± ± ± ± ±
48 2.9 6.7 2.1 0.6 5.1 4.8 0.4 0.5
c—e c—e, d—e c—e c—d, c—e NS NS NS c—d, d—e NS
−3.0 −1.2 −3.8 −4.4
± ± ± ±
0.3 0.1 0.5 0.6
−4.5 −1.9 −5.5 −6.0
± ± ± ±
0.4 0.1 0.6 0.7
−3.1 −1.3 −2.7 −3.7
± ± ± ±
0.3 0.1 0.5 0.6
c—d, d—e c—d, d—e d—e d—e
c—d c—d d—e NS
NS
NS
NS NS NS d—e NS NS NS
NS NS NS NS NS NS NS
−1.9 ± 0.4 −22.8 2.9 −20.7 −33.2 −3.1 −7.0 −2.6
± ± ± ± ± ± ±
4.1 1.4 3.2 5.6 1.5 1.5 1.2
−3.3 ± 0.4 −31.0 0.2 −21.5 −50.5 −4.3 −8.8 −4.2
± ± ± ± ± ± ±
4.8 1.6 3.7 6.5 1.8 1.8 1.4
−2.6 ± 0.3 −18.5 0.1 −13.1 −25.7 −2.6 −6.4 −2.4
± ± ± ± ± ± ±
3.9 1.3 3.0 5.3 1.4 1.4 1.1
Values are means ± S.D. NS: not significant. a Adjusted for age and baseline values. Tukey’s test; p < 0.05. NS: not significant. b Adjusted for age, baseline values, change in energy intake, change in the number of walking steps. c DJP: decreasing Japanese foods pattern. d IHP: increasing healthy foods pattern. e CSP: changing staple foods pattern. f Changes before and after the program. g LDL-C was calculated when subject’s TG < 400 mg/dl (n = 101).
The change in the number of walking steps was not significantly different among the three patterns. Adjustment for age and baseline the number of walking steps did not change this finding. In DJP, intake of energy, fat, carbohydrate, protein and dietary fiber (g) decreased most. On the other hand, CSP showed less change in energy, fat and carbohydrate intake than DJP. Although no significant differences were seen in changes of energy, fat, and carbohydrate intake between DJP and IHP, IHP showed less change in protein intake than DJP. Furthermore, IHP increased the most energy% from protein than the other two patterns. There were no significant differences in change of energy% from fat, carbohydrate, dietary fiber intake (g) and dietary fiber intake (g/1000 kcal) among the three patterns.
In IHP, body weight, BMI, %body fat, and waist circumference decreased most, and significant differences were seen in body weight and BMI decreases compared with the other patterns even after adjusting for age and baseline values. Furthermore, %body fat and waist circumference decrease of subjects with IHP were significantly greater than CSP. After adjusting for changes in energy intake and the number of walking steps in addition to age and baseline values, IHP had greater body weight and BMI decrease than DJP, and %body fat decrease was significantly greater than CSP. There was no significant difference in hip circumference change among the three patterns. Decreases in TC, LDL-C, TG, FPG, SBP and DBP were greatest in IHP, however, significant difference was seen only in the TG decrease between IHP and CSP.
86
Discussion The primary objective of this study was to examine which food intake change patterns are effective for body weight loss. The values of the changes of food intake before and after the program for each food group were standardized, and a cluster analysis was performed. As a result, three food intake change patterns were derived. The mean body weight decreased in all of the food intake change patterns. However, IHP, which decreased intake of energy-dense foods such as sugars, oils and meats, and increased intake of fruits and seafood, reduced anthropometric data the most, and significant differences were seen in body weight, BMI, and %body fat decrease after adjusting for age, baseline values, the changes in energy intake and the number of walking steps compared with DJP or CSP. DJP and CSP mostly showed that change in staple foods like rice or noodles and it was different from IHP. In the program, participants received individual dietary advice to reduce energy intake, and the changes of food groups were left to each participant. IHP showed more change in energy intake than CSP, and anthropometric data, serum lipid, plasma glucose and blood pressure improved more. On the other hand, there was no significant difference in the change of energy intake between DJP and IHP, but IHP saw more decreases in body weight, BMI, other anthropometric data and serum lipids than DJP. This result that decreasing energy intake and weight loss are not always correlated with each other, is an interesting observation, and is also consistent with a previous study [8]. Therefore, the results of this study support the idea that it would be important to evaluate not only energy intake but also food intake change patterns for weight reduction of overweight people. A decreasing of protein intake in DJP was due to a decrease in seafood and meat intake. On the other hand, IHP decreased meat intake but increased seafood intake, thus protein intake was relatively maintained. Okuda et al. [27] compared changes of a single food intake between a successful weight loss group (weight loss was 2.0 kg and more) and an unsuccessful group (weight loss was less than 2.0 kg) in a weight-reduction program in Japan. The successful group increased fish intake and a significant difference was seen only in the change of fish intake. In this study, an increase of seafood intake was seen in the greatest weight loss group (IHP) and the result was consistent with the previous study. As for protein intake and weight loss, WesterterpPlantenga et al. [33] suggested that high protein intake sustained body weight maintenance after
M. Nakade et al. weight loss in humans. Furthermore, diet-induced thermogenesis of protein is highest [34]. High dietinduced thermogenesis may lead to more energy expenditure. Therefore, IHP might be more effective for weight loss and the maintenance of weight loss than the other patterns during the program. In this study, IHP which had the greatest weight loss also decreased meat and oil intake, and increased fruit and seafood intake. Ledikwe et al. [14] and Quatromoni et al. [17] reported that dietary patterns which included meats and oils were correlated with high BMI, and patterns which included fruits and fish were correlated with low BMI. These results were adjusted for energy intake. Hence, these combinations of foods might be independent of energy intake and associated with BMI. Thus, our results are similar to these previous studies and these combinations of food would be important for effective weight loss. However, a previous study of dietary patterns reported poultry intake was correlated with low BMI, and processed meat was correlated with high BMI [14]. Another previous study reported that lean poultry intake was correlated with a lower risk for becoming overweight and meat except for lean poultry was correlated with a higher risk for becoming overweight [17]. Therefore, different kinds of meats may be associated with obesity. To investigate these questions, it may be necessary to examine food intake change patterns using more specific meat groups. IHP decreased intake of beans except soybeans. Many previous studies of dietary patterns reported that bean intake was associated with low BMI [9,11,14,17]. This was not consistent with the results of the current study. It may be due to a difference in the portion size of bean intake among ethnic groups. Maskarinec et al. [9] compared dietary patterns across ethnic groups; a high bean intake pattern was very common among Japanese women. In the current study, the total amount of beans except soybeans and soybean intake for each pattern before the program was 66.9 g, 88.9 g, and 70.5 g per day, respectively. On the other hand, in a previous study in America [14] in which a high-nutrient-dense pattern was associated with low BMI, bean intake was very low compared with the results of this study. Hence, in many previous studies, increasing bean intake might be associated with low BMI, while in the current study, decreasing bean intake was associated with weight loss. The result of this study also suggested that different kinds of beans (soybeans or beans except soybeans) had different effects on weight loss, because the difference was seen only in the change of beans except soybean intake. In Japan, beans except
Correlation between food intake change patterns and body weight loss soybeans are usually used for sweet desserts with sugar. Many studies reported that in addition to these foods, sweets were associated with high BMI [14,17] and low-fat dairy was associated with low BMI [16,19]. But in this study, these food changes were not different among the three patterns, because the mean intake of sweets decreased and low-fat dairy intake increased in all of the patterns. The weight-reduction program in this study was only for 12 weeks, but the improvements of serum lipid, plasma glucose and blood pressure were large in all of the patterns, therefore we could not detect significant differences except TG. Significant difference was seen in TG reduction between IHP and CSP. There were no significant differences in baseline exercise habit, smoking habit, the number of walking steps, body weight, the other anthropometrical and clinical data among the three patterns. However, baseline food intake was different among the patterns. Food groups whose baseline intake was greater or less compared with the other patterns became characteristic of each pattern (e.g., noodles, miso and beans except soybeans in DJP or sugars and sweeteners, butter, vegetable oils, beans except soybeans and meats in IHP or noodles and vegetable oils in CSP). On the other hand, baseline rice and seafood intake were not significantly different among the three patterns and fruit intake in IHP was not significantly different from those of the other patterns. Therefore, most of these food groups changed depend on the amount of baseline intake while rice intake in DJP and fruit and seafood intake in IHP changed after the program, these results might be associated with the degree of body weight loss in each food intake change patterns. The authors were not aware of any studies which derived patterns using change of food intake and examined the correlation between the patterns and body weight loss. The results of this study suggested that there were significant differences in the magnitude of weight loss among food intake change patterns even if change in energy intake was not significantly different. Although cluster analysis is a method used to classify subjects based on their similarity in food intake change, some subjects had the food groups whose food intake change was in the opposite direction of the pattern characteristic. We defined these food groups as unfit food groups, and classified subjects who had at least two unfit food groups into ‘‘less fit’’ and the other subjects into ‘‘more fit’’. There were no significant differences in body weight change or change in metabolic parameters between less fit and more fit in DJP and IHP;
87
and although significant difference was seen in the change of HDL-C in CSP, it improved more in more fit subjects than less fit subjects (data not shown). The results showed that food intake change patterns derived by cluster analysis were useful, although there were some unfit food groups. In this study, there were four subjects with an increase in body weight after the program. The mean weight gain of these four subjects (weight gain group) was 0.9 ± 0.8 kg (range 0.3—2.0 kg). The weight gain group was older, had a lower body weight and a greater number of walking steps than the weight loss group (n = 98) at baseline. There was no difference in energy intake. After the program, the change in energy intake was not different between the groups. However, the weight gain group decreased the number of walking steps while the weight loss group increased it. Therefore, this might be associated with the increasing of body weight (data not shown). This study has some limitations. This study used a dietary record because it is a gold standard. There are several studies which used a 7-day dietary record [16,19], but in this study, only a 2-day dietary record was used, so habitual food intake is not reflected. However, many studies of dietary patterns use food frequency questionnaires [9—12,15,17,18] and the kinds of foods they can investigate are limited. On the other hand, the current study investigated all of the foods which subjects ate. Physical activity was assessed only by the number of walking steps, thus the effect of exercise may not be well evaluated. However, the study subjects performed the same exercise during the program. Therefore, we evaluated the change in physical activity by step count data. The subjects in this study were also limited. The subjects were only women and most of them were middle aged, so other food intake change patterns may be derived in case of men or other age brackets. Thus, more studies are needed using a various subjects. In summary, this study examined the correlation between food intake change patterns and body weight loss, and three food intake patterns were derived. The mean body weight decreased in all of the food intake change patterns. Among the three patterns, body weight loss was significantly greater in IHP, which decreased intake of sugars, oils, beans except soybeans and meats, and increased intake of fruits and seafood, than DJP or CSP, although the change in energy intake was not significantly different. In conclusion, nutrition education and assessment of food intake change patterns in addition to energy intake could be useful for effective weight loss.
88
M. Nakade et al. Appendix A (Continued)
Acknowledgement Contibutions. A part of this study was supported by a grant from the Foundation for Total Health Promotion (Title: Daily Dietary Patterns and Obesity; Principal investigator: Jung Su LEE). We would like to express our thanks to the study participants and to the Koto Health Promotion Center for their cooperation.
22 Food groups
28 Food groups
Examples of foods in the groups
Meats
Meats
Beef, pork, poultry, processed meat
Dairy products
Regular-fat dairy
Regular-fat milk and dairy products Low-fat milk and dairy products
Low-fat dairy
Appendix A. Food groups used in the food intake change pattern analysis in the study 22 Food groups
28 Food groups
Examples of foods in the groups
Rice
Rice
Rice
Wheat
Bread, wheat flour Noodles
Other grains Nuts and seeds
Other grains Nuts and seeds
Potatoes
Potatoes
Sugars and sweeteners Sweets
Sugars and sweeteners Sweets
Bread, wheat flour Soba, udon noodles, pasta, Chinese noodles Cereals Chestnut, sesame, other nuts and seeds Sweet potato, white potato Sugar, honey
Oils
Butter Vegetable oils
Butter Vegetable oils, margarine
Miso
Miso
Miso (soybean paste)
Beans
Soybeans
Soybean and soybean products except miso Adzuki beans, kidney beans
Beans except soybeans Fruits Brightly colored vegetables
Pickles Mushrooms
Fruits Brightly colored vegetables Other vegetables Pickles Mushrooms
Seaweed
Seaweed
Seafood
Seafood
Other vegetables
Cake, cookies
Fruits, fruit juice Carrots, spinach
Lettuce, onions Pickles Shiitake mushroom Laver, wakame seaweed Fish, shellfish, processed fish
Eggs
Eggs
Eggs, processed eggs
Beverages
Alcohol Other beverages
Sake, beer, wine Coffee, tea, green tea
Seasonings
Seasonings
Soy sauce, dressing without oil Mayonnaise, dressing with oil
Mayonnaise
References [1] Formiguera X, Canton A. Obesity: epidemiology and clinical aspects. Best Pract Res Clin Gastroenterol 2004;18: 1125—46. [2] World Health Organization. Diet, nutrition and the prevention of chronic diseases. Report of a joint WHO/FAO expert consultation, WHO Technical Report Series No. 916; 2003. Geneva. [3] Sowers JR. Obesity as a cardiovascular risk factor. Am J Med 2003;115(Suppl.):37S—41S. [4] Ministry of Health, Labour and Welfare. The results of the national nutrition survey in Japan, 2002. Japan: Daiichi Publishers; 2004 [in Japanese]. [5] Japan Society for the Study of Obesity. Manual for the treatment of obesity. 2nd ed. Tokyo: Ishiyaku Publishers; 2003 [in Japanese]. [6] Tetsuguchi M, Abe K, Mimura K. Effects of the exercise and nutritional prescription on body composition in middle age obese woman. J Educ Health Sci 2000;46:1012—9 [in Japanese]. [7] Imaizumi T, Asami T. Effect of weight reduction program on visceral and subcutaneous fat tissues in white collar male workers. Bull Phys Fitness Res Inst 1995;89:24—31 [in Japanese]. [8] Andersson I, Lennernas M, Rossner S. Meal pattern and risk factor evaluation in one-year completers of a weight reduction program for obese men—–the ‘Gustaf’ study. J Intern Med 2000;247:30—8. [9] Maskarinec G, Novotny R, Tasaki K. Dietary patterns are associated with body mass index in multiethnic women. J Nutr 2000;130:3068—72. [10] Gittelsohn J, Wolever TM, Harris SB, Harris-Giraldo R, Hanley AJ, Zinman B. Specific patterns of food consumption and preparation are associated with diabetes and obesity in a Native Canadian community. J Nutr 1998;128: 541—7.
Correlation between food intake change patterns and body weight loss [11] Sichieri R. Dietary patterns and their associations with obesity in the Brazilian city of Rio de Janeiro. Obes Res 2002;10:42—8. [12] Wirfalt AK, Jeffery RW. Using cluster analysis to examine dietary patterns: nutrient intakes, gender, and weight status differ across food pattern clusters. J Am Diet Assoc 1997;97:272—9. [13] Wirfalt E, Hedblad B, Gullberg B, Mattisson I, Andren C, Rosander U, et al. Food patterns and components of the metabolic syndrome in men and women: a cross-sectional study within the Malmo Diet and Cancer cohort. Am J Epidemiol 2001;154:1150—9. [14] Ledikwe JH, Smiciklas-Wright H, Mitchell DC, Miller CK, Jensen GL. Dietary patterns of rural older adults are associated with weight and nutritional status. J Am Geriatr Soc 2004;52:589—95. [15] Lin H, Bermudez OI, Tucker KL. Dietary patterns of Hispanic elders are associated with acculturation and obesity. J Nutr 2003;133:3651—7. [16] Newby PK, Muller D, Hallfrisch J, Qiao N, Andres R, Tucker KL. Dietary patterns and changes in body mass index and waist circumference in adults. Am J Clin Nutr 2003;77:1417—25. [17] Quatromoni PA, Copenhafer DL, D’Agostino RB, Millen BE. Dietary patterns predict the development of overweight in women: The Framingham Nutrition Studies. J Am Diet Assoc 2002;102:1239—46. [18] Togo P, Osler M, Sorensen TI, Heitmann BL. A longitudinal study of food intake patterns and obesity in adult Danish men and women. Int J Obes Relat Metab Disord 2004;28:583—93. [19] Newby PK, Muller D, Hallfrisch J, Andres R, Tucker KL. Food patterns measured by factor analysis and anthropometric changes in adults. Am J Clin Nutr 2004;80:504—13. [20] Hu FB, Rimm EB, Stampfer MJ, Ascherio A, Spiegelman D, Willett WC. Prospective study of major dietary patterns and risk of coronary heart disease in men. Am J Clin Nutr 2000;72:912—21. [21] Fung TT, Willett WC, Stampfer MJ, Manson JE, Hu FB. Dietary patterns and the risk of coronary heart disease in women. Arch Intern Med 2001;161:1857—62. [22] Fung T, Hu FB, Fuchs C, Giovannucci E, Hunter DJ, Stampfer MJ, et al. Major dietary patterns and the risk of colorectal cancer in women. Arch Intern Med 2003;163:309—14.
89
[23] Slattery ML, Boucher KM, Caan BJ, Potter JD, Ma KN. Eating patterns and risk of colon cancer. Am J Epidemiol 1998;148:4—16. [24] van Dam RM, Rimm EB, Willett WC, Stampfer MJ, Hu FB. Dietary patterns and risk for type 2 diabetes mellitus in U.S. men. Ann Intern Med 2002;136:201—9. [25] Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, et al. A clinical trial of the effects of dietary patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med 1997;336:1117—24. [26] de Lorgeril M, Salen P, Martin JL, Monjaud I, Delaye J, Mamelle N. Mediterranean diet, traditional risk factors, and the rate of cardiovascular complications after myocardial infarction: final report of the Lyon Diet Heart Study. Circulation 1999;99:779—85. [27] Okuda N, Okamura T, Kadowaki T, Tanaka T, Uesima H. Weight-control intervention in overweight subjects at high risk of cardiovascular disease: A trial of a public health practical training program in a medical school. Jpn J Public Health 2004;51:552—9 [in Japanese]. [28] Nisigami T, Imamura H, Moriwaki T, Yoshimura Y, Shimada R, Nishimura T, et al. Effects of dietary education and physical exercise on mildly obese and hypertensive women, vol. 24. Japan Society of Health Evaluation and Promotion; 1997. p. 375—81 [in Japanese]. [29] Nagamine S, Suzuki S. Anthropometry and body composition of Japanese young men and women. Hum Biol 1964;36:8—15. [30] Brozek J, Grande F, Anderson JT, Keys A. Densitometric analysis of body composition: revision of some quantitative assumption. Ann NY Acad Sci 1963;110:113—40. [31] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499—502. [32] Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 2002;13: 3—9. [33] Westerterp-Plantenga MS, Lejeune MP, Nijs I, van Ooijen M, Kovacs EM. High protein intake sustains weight maintenance after body weight loss in humans. Int J Obes Relat Metab Disord 2004;28:57—64. [34] de Jonge L, Bray GA. The thermic effect of food and obesity: a critical review. Obes Res 1997;5:622—3.
