Nutrition Knowledge Is Associated with Greater Weight Loss in Obese and Overweight Low-Income Mothers

RESEARCH Current Research

Continuing Education Questionnaire, page 77 Meets Learning Need Codes 4000, 4010, 4180, and 6000

Nutrition Knowledge Is Associated with Greater Weight Loss in Obese and Overweight LowIncome Mothers DEBORAH M. KLOHE-LEHMAN, PhD, RD; JEANNE FREELAND-GRAVES, PhD, RD; EDWARD R. ANDERSON, PhD; TODD MCDOWELL; KRISTINE K. CLARKE, PhD, MPH, RD; HENRY HANSS-NUSS; GUOWEN CAI, PhD; DIVYA PURI; TRACEY J. MILANI, PhD

D. M. Klohe-Lehman is child nutrition outreach program specialist at the Texas Department of Agriculture, Austin; at the time of the study, she was a graduate student in the Division of Nutritional Sciences, The University of Texas at Austin. J. Freeland-Graves is the Bess Heflin Centennial Professor, Division of Nutritional Sciences, The University of Texas at Austin. E. R. Anderson is associate professor, Division of Human Development and Family Sciences, The University of Texas at Austin. T. McDowell is a science teacher at Galindo Elementary in Austin, TX. K. K. Clarke is a postdoctoral research fellow at the Center for Pediatric Nutrition Research, Department of Pediatrics, The University of Utah, Salt Lake City; at the time of the study, she was a graduate student in the Division of Nutritional Sciences, The University of Texas at Austin. H. Hanss-Nuss is a doctoral student in the Division of Nutritional Sciences, The University of Texas at Austin. G. Cai is a postdoctoral scientist, Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX; at the time of the study, she was a graduate student in the Division of Nutritional Sciences, The University of Texas at Austin. D. Puri is an early childhood educator at a New York public school; at the time of the study, she was an undergraduate student in the Division of Nutritional Sciences, The University of Texas at Austin. T. J. Milani is a Clinical Project Manager at MedLogix Communications, LLC, San Diego, CA; at the time of the study, she was a graduate student in the Division of Nutritional Sciences, The University of Texas at Austin. Address correspondence to: Jeanne Freeland-Graves, PhD, RD, The Bess Hefflin Centennial Professor, Division of Nutritional Sciences, 1 University Station, A2700, The University of Texas at Austin, Austin, TX 78712. E-mail: [email protected] Copyright © 2006 by the American Dietetic Association. 0002-8223/06/10601-0006$32.00/0 doi: 10.1016/j.jada.2005.09.047

© 2006 by the American Dietetic Association

ABSTRACT Objective To examine if greater nutrition knowledge vs gains in knowledge promote more successful weight loss in low-income, overweight and obese mothers with young children. Design A convenience sample of mothers and their children were measured for height and weight; mothers completed demographic and nutrition knowledge questionnaires pre- and postintervention. Subjects/setting Participants (N⫽141) were recruited from government and public health clinics and elementary schools. Inclusion criteria for mothers were: family income ⬍200% federal poverty level; overweight/obese; and Hispanic, African-American, or white race/ethnicity. Intervention Eight weekly weight-loss classes emphasizing diet, physical activity, and behavior modification based on Social Cognitive Theory were administered to mothers. Main outcome measures Improvements in maternal nutrition knowledge and weight loss. Statistical analyses performed Paired-samples t tests, repeated measures analysis of variance, analysis of covariance, Pearson correlations, and ␹2 statistics. Results Nutrition knowledge of mothers increased in all areas. Participants with weight loss ⱖ2.27 kg (responders) had greater knowledge than those who did not; however, the actual net gain was similar for those who lost and did not lose weight. Weight gainers only improved in two areas on the test, whereas weight-loss responders increased knowledge in all six. Responders appeared more cognizant of diet, weight loss, and health information. Conclusions Weight-management programs should include a strong component of nutrition education to alleviate knowledge inequalities and promote more effective weight control. In low-income mothers, greater initial knowledge may be more predictive of weight loss than gains in knowledge during an intervention. J Am Diet Assoc. 2006;106:65-75.

Journal of the AMERICAN DIETETIC ASSOCIATION

65

O

besity rates continue to rise in the United States, particularly among minorities (1). Among women aged 20 to 39 years, 70% of African Americans, 62% of Hispanics, and 49% of whites are estimated to be overweight or obese (body mass index [BMI] ⱖ25) (2). Obesity is more pronounced in low socioeconomic (3) and food-insecure populations (4). For example, the rate of overweight for adult Mexican-American women is about 13% higher for those living below vs above the poverty line (5). In addition, the risk of becoming overweight increases by 60% to 110% for women who have had at least one live birth (6). Finally, postpartum weight retention may be associated with method of infant feeding. Kac and colleagues (7) reported that breastfeeding may protect against weight retention, but the rates of this practice are typically less in low socioeconomic populations (8). The pervasiveness of obesity, particularly in low-income, minority mothers, suggests the need for interventions tailored for this population. One component of an effective weight-loss program should be the improvement of nutrition knowledge. Lack of knowledge may contribute to the higher rates of obesity in economically disadvantaged people. For example, low-income caretakers with children were less likely to know about diet– disease associations, to use food labels, or to have low-fat eating habits than those with higher incomes (9). Thus, weightmanagement curricula for those at lower socioeconomic levels should be designed to increase awareness of basic nutrition and diet– disease links. Improving nutrition knowledge is an important tool for stimulating dietary behavior that promotes weight loss. Examples are: increased fat avoidance (10) and low-fat dietary patterns (11-14), adoption of more healthful cooking methods (15), improved skill at label reading (16) and meal planning (17), and reduced consumption of highenergy and high-fat foods (14,17-19). In addition, Smith and colleagues (20) found that nutrition knowledge was predictive of dietary change and that those making two or more dietary changes had greater nutrition knowledge scores. However, others have not observed an association between nutrition knowledge and more healthful diet practices (21-23) or weight control (24,25). These conflicting results suggest knowledge does not necessarily lead to behavior change.

Lack of knowledge may contribute to the higher rates of obesity in economically disadvantaged people. Although the relationship between knowledge and behavior remains debatable, a number of studies have reported greater nutrition knowledge and weight loss following a nutrition education intervention (26-31). Of these, only two studies were conducted in low-income women (28,29). None of the studies reported if gains in nutrition knowledge were related to successful weight loss, except the study by Jeffery and Wing (30). In 177 adults (mean age 37 years) participating in a behavioral weight-loss intervention, increased nutrition knowledge

66

January 2006 Volume 106 Number 1

at 18 and 30 months was related to weight loss (30). However, this population was not made up of low-income or minority mothers, who may respond differently to nutrition education interventions. In contrast, four other studies reported increases in nutrition knowledge but no weight loss (10,32-34). This suggests that improving nutrition awareness may not always lead to better weight control. Although knowledge is clearly a necessary component, the type of curriculum, the focus of the intervention (ie, weight loss vs reduction of dietary fat/cholesterol for cardiovascular disease), the effectiveness of the instructor, and the learning motivation of the participants may significantly influence its efficacy. The purpose of our study was to examine if greater nutrition knowledge vs gains in knowledge promote more successful weight loss in low-income, overweight and obese mothers with young children. In addition, differences in maternal knowledge by demographics and weight status were evaluated. METHODS Experimental Design Low-income, Hispanic, African-American, and white mothers (N⫽141) participated in an 8-week nutrition and physical activity weight-loss intervention. Mothers and children were measured for height and weight, and mothers completed demographic and nutrition knowledge questionnaires at weeks 0 and 8 (pre- and postintervention). Self-monitoring of food intake and physical activity were assessed weekly with optional 7-day food records and pedometers worn for 3-day intervals, respectively. Subjects Participants were overweight or obese (BMI ⱖ25) mothers of children aged 8 months to 12 years, aged ⱖ18 years, had a combined family income ⬍200% of the federal poverty level, and were literate in English or Spanish. Subjects were primarily recruited via flyers posted at Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) clinics and public health clinics. In addition, children at two local low-income elementary schools that requested the weight-loss program for mothers were given flyers to take home with a phone number to call for more information. Applicants called the number on the flier and were enrolled in the study if they met inclusion criteria. No differences in ethnicity or socioeconomic status were observed between mothers recruited from schools or WIC clinics. Mothers were informed of the benefits and risks of the study and informed consent was obtained. The Institutional Review Board of The University of Texas at Austin approved the study. Calculation of Weight Status Height and weight of the mothers was recorded to assess BMI changes following the weight-loss program. These parameters were observed in the children to explore the relationship between child’s weight status and maternal nutrition knowledge. A stadiometer (Perspective Enterprises, Portage, MI) was used to determine height of mothers and children ⱖ2

years. Mothers and children who could stand unaided on a scale were weighed using a calibrated electronic scale (Model HS-100-A, Fairbanks Scales, St Johnsbury, VT) without shoes or heavy clothing. Mothers with a child unable to stand alone on the scale held the child, and the mother’s weight was subtracted out to obtain the child’s weight. Recumbant length of children younger than age 2 years was measured on a flat surface with a tape measure affixed. The head of the child was positioned perpendicular to, and up against, a wall and the ankles bent with feet at a 90° angle to the leg. Two trained professionals aided in the measurement. A mark was placed on the tape measure at the bottom of the heel. BMI was calculated as kg/m2. Mothers were divided into two weight categories for comparison: responders (those able to lose 2.27 kg or more) and nonresponders (those losing less than 2.27 kg). A 2.27-kg weight loss was chosen for responder criterion because the mean weight loss for the sample was 2.45 kg, and this classification allowed for a fairly equal division. To examine the knowledge characteristics of those participants who gained weight vs those losing 2.27 kg or more, the responders who gained weight were further categorized as weight gainers. In children aged 2 years or older, BMI-for-age percentiles were estimated using growth charts from the Centers for Disease Control and Prevention (35). Underweight was classified as less than the fifth percentile, normal weight as at or above the fifth percentile and ⬍85th percentile, at risk for overweight as ⱖ85th percentile, and overweight/obesity as ⱖ95th percentile. In children younger than age 2 years, weight for length was used instead of BMI for age, with children below the fifth percentile labeled as underweight, between the fifth and 95th percentiles as normal weight, and ⬎95th percentile as overweight. Demographics Questionnaire A 33-item general questionnaire was used to obtain information on age, income, education level, primary language, parity, childbirth, employment status, job satisfaction, and family problems. Additional questions asked about current practices to lose weight, like skipping meals, and about previous attendance at childbirth classes. Nutrition Knowledge Test A 25-item test (final version) measured nutrition knowledge of mothers at weeks 0 (pre-) and 8 (postintervention). This questionnaire contained 17 multiple-choice and eight true/false questions and took approximately 15 minutes to complete. The nutrition knowledge score was calculated by summing the total number of correct responses out of 25 (maximum number possible). A 30-item test (original version) was developed specifically for this weight-loss program. None of the questions were adapted from an existing instrument. An expert panel of five dietetics professionals identified questions (20 multiple choice and 10 true/false) that evaluated understanding of concepts taught in the weight-loss classes. The experts identified seven areas of interest (weight

loss, heart disease, prenatal nutrition, child nutrition, the Food Guide Pyramid, food sources and functions of macronutrients, and food sources and functions of vitamins and minerals) and collapsed the related questions into one of the seven areas for ease of analysis. The items were reviewed for readability and item difficulty, appropriateness, bias, and correctness. An additional eight experts rated the individual questions for relevance to the areas of interest. Acceptable agreement (70% to 100%) was achieved for all questions except one, yielding a content validity index of 96%. This item was removed from the questionnaire because it also performed poorly in the reliability analysis. The heart disease questions were removed because of their poor content validity index and/or reliability values. Thus, the final version contained only six areas of interest. The 30-item test was pilot tested for readability and item difficulty in a sample of 92 low-income mothers with a similar triethnic breakdown as the study population. This pilot resulted in the removal of two questions because of their high reading level. The pilot also provided data for the reliability and validity analyses of the questionnaire. Kuder Richardson’s KR-20 was used to determine internal consistency reliability of the 30-item test because it is preferred for use with dichotomous variables (ie, correct vs incorrect). The KR-20 for the initial 30-item test was 0.56; thus three questions on heart disease, one on alcoholic content of beverages, and one on olestra’s side effects were removed from the test to improve reliability. The final 25-item test administered had six areas of interest with a content validity index of 100% and a KR-20 of 0.6. This value is considered a good scale because KR-20 produces more conservative estimates than Cronbach’s ␣ (36). Discriminate validity was established by comparing scores of a sample of mothers (n⫽92) to scores of nutritional science doctoral degree students (n⫽11). An independent samples t test revealed that advanced nutrition graduate students performed significantly better than the mothers, with mean scores of 92% and 56%, respectively (P⬍.001). This disparity indicated that the questionnaire could successfully differentiate between those with lower and higher levels of knowledge. The questionnaire also was translated and back translated into Spanish and evaluated for readability by two different bilingual speakers. Only seven participants needed the Spanish version of the knowledge test. Self-Monitoring with Weekly Food Records and Pedometers Self-monitoring techniques included maintaining food and pedometer records. Mothers were strongly encouraged, but not required, to complete 7-day food records and wear pedometers each week (model AE170, Accusplit, San Jose, CA) for 3 days (2 weekdays and 1 weekend day) to monitor their food intake and physical activity. The total number of completed food records and/or pedometer records was tallied, and this number was used as a score for self-monitoring, with more recalls/records indicating greater use of self-monitoring techniques.

January 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION

67

Weight-Loss Program Weight-loss classes for mothers consisted of dietary, activity, and behavior changes. A series of eight weekly classes were taught by registered dietitians. Classes in Spanish were taught bilingually and separately from those in English. In the 2-hour classes a variety of topics were discussed. The specific class topics selected for this intervention were derived from other strategies proven effective in weight-loss programs. Examples include: weight loss and exercise goals, personalized meal plans and menus, energy content of various foods, types/sources/ health effects of fats, low-fat cooking demonstrations, child nutrition, stress reduction, emotional eating, tips for eating healthfully when short on time, and techniques for maintenance of weight loss. Classes for mothers of 1to 3-year-olds and elementary school children were identical, except for information on offering low-fat milk and types of physical activities encouraged. To achieve desired behavior changes, the weight-loss program was based on concepts from Social Cognitive Theory because it is generally regarded as a successful method for promoting health behavior change (37,38). Social Cognitive Theory explains how people acquire and maintain certain behavior patterns, while also providing a framework for the design, implementation, and evaluation of interventions. According to this theory, health behaviors can be attributed to a dynamic interaction between a person’s environment, behaviors, and cognitions. Change in any one of these areas may alter health behaviors (39). This intervention was designed to improve knowledge (cognition) of general health and nutrition and facilitate practice of behaviors (behavioral capability). It also stimulated mothers to change their eating environment, thereby promoting healthful changes in diet and activity for weight loss. Social Cognitive Theory constructs that were incorporated into the curriculum included enhancing knowledge and skills needed to lose weight (behavioral capabilities); increasing confidence and overcoming barriers to healthful eating and exercise (self-efficacy); behavior role modeling by peers (observational learning); goal-setting, self-monitoring, and decision making (self-control); providing rewards and incentives for meeting goals (reinforcements); improving the healthfulness of foods served in the home (environment); correcting misperceptions and improving the anticipated outcomes of weight loss (expectations); enhancing the value placed on weight loss (expectancies); and learning strategies to deal with emotional stimuli and stress (emotional coping responses). The dietary component was based on healthful, wellbalanced eating plans that promoted a daily 500-kcal deficit for weight loss. The majority of mothers had prestudy calculated energy needs of 1,800 to 2,000 kcal/day; thus, mothers were instructed to consume no more than 1,200 to 1,500 kcal/day, depending on their weight and activity level. Participants were given the option of either reducing their energy intake via diet changes, increasing energy expenditure through activity, or a combination of both. Mothers were provided with specific strategies to improve diet and exercise to result in a 500 kcal/day reduction. These included following a meal plan or use of the exchange lists system, lists of energy content of foods, more healthful food substitutions, and encouraging daily

68

January 2006 Volume 106 Number 1

vigorous walking. A discussion of health benefits of foods/ nutrients and development of meal-planning skills using the Food Guide Pyramid (40) and the Food Guide Pyramid for Young Children (41) were additional topics. The role of physical activity in weight loss was discussed, as were techniques to increase the amount and type of exercise. Mothers also were encouraged to increase kilocalories burned per day (as shown on their pedometer) by 350, or to obtain 10,000 steps per day, to facilitate weight loss. Mothers also were given examples of activities to do together with their child. Further details of the intervention are provided in a report by Klohe-Lehman and colleagues (Klohe-Lehman DM, Clarke KK, Cai G, Voruganti VS, Milani T, Hanss-Nuss H, Proffitt JM, Bohman TM, Freeland-Graves JH. Low-income, overweight/obese mothers act as agents of change to improve food choices and fat habits in their 1-3 year old children. Manuscript submitted for publication.). Statistical Analysis Data were analyzed using SPSS (version 11.5, 2002, SPSS Inc, Chicago, IL). Frequencies and means were computed for descriptive purposes. Paired-samples t tests examined overall changes in scores pre- and postintervention. Repeated measures analysis of variance assessed differences in pre and post scores by demographic/ anthropometric variables and weight category. Initial statistical models using analysis of covariance were formulated that included pretest scores, ethnicity, education level, and income as fixed factors and covariates. Only pretest scores significantly affected posttest scores; therefore, posttest scores were controlled for differences in pretest scores only. The adjusted posttest scores calculated by analysis of covariance are listed alongside the unadjusted scores. ␹2 statistics were used to determine individual posttest questions that were significantly associated with successful weight loss. For example, contingency tables computed the number of weight-loss responders, nonresponders, and gainers who scored correctly on each question. From these values, the percentages of responders, nonresponders, and gainers who scored correctly could be derived. Pearson correlations examined relationships between test scores and continuous demographic data (ie, BMI and weight loss). RESULTS The 141 participants were a triethnic sample of mothers (67% Hispanic, 18% African American, and 15% white) (Table 1). Of these, 21% were overweight and 79% were obese. Age (mean age 28 years) and BMI (mean BMI 35) of the mothers and number of children per household (mean two children) did not vary according to ethnicity. The majority of participants reported incomes between $15,000 and $29,999 (46%). The most frequent education level attained was partial college for African Americans (40%) and whites (38%); for Hispanics, high school graduate (33%). Ages of children ranged from 8 months to 12 years; with whites slightly younger than Hispanics. In this sample, 28% of children were classified as overweight (20%) or at risk (8%). Sample characteristics of mothers and children by

Table 1. Sample characteristics of mothers and children recruited from government and public health clinics and elementary schools (N⫽141) by ethnicity and weight-loss category of mothers Ethnicity Characteristic

Weight-Loss Category

African American

Hispanic

White

Responders

Nonresponders

Weight gainers

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ mean⫾standard deviation ™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ 3 Mothers Age (y) Body mass index No. of children Children Age (y) Overweight/at riska Healthful weighta a

28.2⫾5.9 35.7⫾6.5 2.4⫾1.3

28.6⫾6.9 33.9⫾7.2 2.0⫾0.9

3.2⫾2.9y frequency 26 of 95 69 of 95

28.5⫾6.2 33.0⫾7.5 2.1⫾1.2 1.5⫾0.9y

2.5⫾1.6 % 27 73

frequency 9 of 25 16 of 25

27.9⫾5.7 34.9⫾6.5 2.2⫾1.3y

% 36 64

frequency 4 of 21 17 of 21

28.7⫾6.4 35.0⫾7.1 2.4⫾1.2

2.1⫾1.6z % 19 81

frequency 17 of 66 49 of 66

29.3⫾7.1 34.7⫾7.0 2.9⫾1.1y

3.4⫾3.0z % 26 74

frequency 22 of 75 53 of 75

3.3⫾2.8 % 29 71

frequency 7 of 21 14 of 21

% 33 67

According to growth charts from the Centers for Disease Control and Prevention (35). Means with the same superscript within a row are significantly different (Tukey post hoc test, P⬍.05).

yz

Table 2. Nutrition knowledge scores of mothers recruited from government and public health clinics and elementary schools before and after the weight-loss intervention Score (%) Category

n

Pretest

Posttesta

Adjusted posttestb

All Ethnicity African American White Hispanic Education level Less than junior high Partial high school High school graduate Partial college College/graduate degree Method of infant feeding Breast milk Formula Combination of formula and breast milk Body mass index categories 25.0-29.9 30.0-34.9 ⱖ35.0 Child’s weight status Normal weight Overweight or at risk

141

60

72

—

25 21 95

56u 72uv 60v

76 80u 72u

76 76 72

12 20 43 48 18

52uvw 56xy 60wz 64vx 68uyz

60uvwx 72uy 76v 76wy 80x

64uvwx 72u 76v 76w 76x

29 49 63

68uv 56u 60v

76u 72u 76

72 72 76

30 48 63

64 60 60

80 72 72

76 72 76

102 39

56 64

72 76

72 76

a

Posttest scores for all subgroups higher than at pretest (P⬍.05). Adjusted for differences in pretest scores (analysis of covariance test with pretest as a covariate). uvwxyz Means with the same superscript within a column are significantly different (Bonferroni adjustment, P⬍.05). b

weight-loss category of mothers are presented in Table 1. Responders (ⱖ2.27 kg weight loss) did not differ in education, income, or any other demographic factors from nonresponders (⬍2.27 kg weight loss) or weight gainers.

However, a larger percentage of whites (76%) were responders than Hispanics (43%) and African Americans (36%) (P⬍.05), and nonresponders were more likely to be living without a spouse or partner (P⬍.05) and to have

January 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION

69

Table 3. Nutrition knowledge scores of mothers recruited from government and public health clinics and elementary schools according to areas of interest Score (%) Responders Area of interest All areas Pretest Posttest Prenatal nutrition Pretest Posttest Child nutrition Pretest Posttest Vitamins and minerals Pretest Posttest Macronutrients Pretest Posttest Weight loss Pretest Posttest Food Guide Pyramida Pretest Posttest

Nonresponders

All (Nⴝ141)

>2.27 kg weight loss (nⴝ66)

<2.27 kg weight loss (nⴝ75)

Weight gainers (nⴝ21)

60 72***

64yz 80***yz

56y 68***y

52z 64***z

71 84***

78yz 88*y

65y 80***

55z 64y

69 82***

74yz 87***yz

64y 77***y

59z 73**z

66 71**

67 75**y

64 68

58 62y

50 66***

52 73***yz

49 60**y

43 56z

51 67***

54y 74***yz

47 61***y

42y 48z

44 67***

53yz 73***

36y 62***

29z 60**

a

See reference 40. Means with the same superscript within a row are significantly different (Bonferroni adjustment, P⬍.05). *P⬍.05 for significant increase at posttest. **P⬍.01 for significant increase at posttest. ***P⬍.001 for significant increase at posttest.

yz

children who were older (P⬍.01). Weight gainers had more children in the household (P⬍.05), weighed less prepregnancy (P⬍.05), were more likely to skip meals (P⬍.05), and less likely to be white (P⬍.05). Nutrition Knowledge Changes in nutrition knowledge before and after the intervention are shown in Table 2. Whites scored significantly higher than African Americans and Hispanics at pretest, but only higher than Hispanics at posttest. However, after adjusting the posttest scores for differences of pretest, ethnicity was no longer significant. Within ethnicities, African Americans had the lowest pretest scores, followed by Hispanics, then whites; however, whites only increased their nutrition knowledge by 11%, compared with a 20% and 36% enhancement in Hispanics and African Americans, respectively (P⬍.05, between African Americans and whites/Hispanics). Scores increased as years of formal education increased. At pre- and posttest, those with less than a junior high school education, with incomes ⬍$15,000, and having Spanish as the primary language scored lower than those with higher education, income (P⬍.01), and speaking English (P⬍.01).

70

January 2006 Volume 106 Number 1

Mothers who breastfed their child during infancy had significantly higher pre- and posttest scores than those who fed formula. No significant differences were observed pre- or posttest by BMI category; between mothers of healthful-weight children vs overweight or at-risk children; or age of children. Mothers who skipped meals had lower pretest scores (56%) than those who did not miss meals (64%, P⬍.05); by posttest this difference was no longer significant. Mothers who had attended childbirth classes had higher pretest scores (68%) than those who did not (60%, P⬍.01); at posttest both groups scored similarly. In addition, pre- and posttest knowledge was correlated negatively with number of children in the household (r⫽ – 0.23, P⬍.01), and posttest score was correlated positively with class attendance (r⫽0.24, P⬍.01) and selfmonitoring with the pedometer (r⫽0.27, P⬍.01) and food records (r⫽0.26, P⬍.01). Virtually all participants reported some form of stimulus control; thus, its relationship to knowledge could not be identified. Setting goals and rewarding for achievements was unrelated to knowledge. At pretest, responders scored 14% and 23% higher than nonresponders and weight gainers, respectively, and continued to do so at posttest (18% and 25%, respectively) (Table 3). Women who lost more weight had higher pre-

Figure. Weight loss by mothers recruited from government and public health centers and elementary schools correlated with posttest knowledge score. test (r⫽0.31, P⬍.001) and posttest (r⫽0.34, P ⬍.001) knowledge scores than nonresponders (see the Figure). In addition, pounds lost were correlated positively with posttest knowledge of weight loss (r⫽0.25, P⬍.01), macronutrients (r⫽0.25, P⬍.01), child nutrition (r⫽0.21, P⬍.05), and prenatal nutrition (r⫽0.21, P⬍.05). Mothers improved in all six areas of interest covered in the test (Table 3). At pretest, mothers were most cognizant about prenatal nutrition, child nutrition, and micronutrients; at posttest, these topics remained the strongest. The weakest sections (macronutrients, weight loss, Food Guide Pyramid) improved the most at posttest. Weight gainers increased significantly only in two areas at posttest: child nutrition and the Food Guide Pyramid. Posttest knowledge of responders was significantly higher than weight gainers in all categories, except the Food Guide Pyramid. Table 4 displays individual posttest questions that were associated with successful weight loss. Responders were more likely to answer correctly on questions related to sodium, snacks for children, folic acid, food labels, and energy and carbohydrates in foods than weight gainers. DISCUSSION Successful weight loss was associated with greater nutrition knowledge in this study. Those mothers able to lose ⱖ2.27 kg scored higher on both pre- and posttests than those who did not following the program. However, both responders and nonresponders to the intervention gained in knowledge. Responders appeared to be more cognizant of information about diet and health that is presented in public health or WIC clinics, as well as information related to weight loss. Overall, the weakest areas were weight loss, the Food Guide Pyramid, and macronutrients.

A number of weight-management interventions have reported increased nutrition knowledge with weight loss (26-31). These include cholesterol education programs to diminish risk for coronary heart disease in 49 men and 51 women (31) and 145 worksite employees (27), and weight management for 64 older African Americans (26). However, differences in knowledge by weight loss responders and nonresponders were not reported. Of these, only one examined the degree of association between knowledge and weight reduction (30). In a behavioral weight-loss program, Jeffery and Wing (30) found increased knowledge in 177 slightly older (mean age 37 years) subjects with obesity at the end of 18 months and the 30-month follow-up. Weight loss was correlated positively with increases in knowledge at both time points. This finding differed in our study, in which overall greater knowledge, not gains, were related to weight loss. The only studies examining nutrition knowledge and weight loss in low-income women were by Domel and colleagues (28,29) in 31 African-American and 34 Hispanic, low-income, overweight women. Higher weight change was found in those receiving nutrition education. After 11 weeks, weight loss averaged 1.41 kg and 4.09 kg in African American and Hispanics, respectively, and significant improvements in nutrition knowledge were seen (18% and 26%). The increase in knowledge in Hispanics is similar to our study (20%). However, AfricanAmerican women in our research had the greatest knowledge gains (36%) compared with other ethnic groups.

Social Cognitive Theory appears to be an effective theoretical framework for designing weight-loss interventions for low-income mothers. Others have not documented weight loss with increased nutrition knowledge. These interventions included dietary fat modification in 351 low-income participants (34), worksite cholesterol reduction in 272 men (33), a heart-health initiative in 434 low-literate Latino adults (10), and fat and Food Guide Pyramid education for 365 Samoans (32). Thus, cognition does not necessarily translate into behavior changes. It should be emphasized that factors other than nutrition knowledge may contribute to successful weight reduction in this population (ie, psychosocial factors and intrinsic/extrinsic motivations). Some examples are social support, depression (42), stress, self-efficacy (43), body image (44), and attitudes (45) toward nutrition and weight loss that will be presented in future articles. In our investigation the degree of obesity did not influence baseline nutrition knowledge or subsequent weight loss. Burns and colleagues (25) also reported that knowledge scores were comparable for obese and overweight subjects, but these subjects scored higher than subjects with a healthful weight. In addition, Allison and colleagues (24) observed that obesity among African-American women was not related to nutrition knowledge be-

January 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION

71

Table 4. Nutrition knowledge posttest questions associated with successful weight loss in mothers recruited from government and public health clinics and elementary schools (N⫽141)a Percentage with a Correct Answer Responders

Nonresponders

>2.27 kg weight loss (nⴝ66)

<2.27 kg weight loss (nⴝ75)

Weight gainers (nⴝ21)

Canned soups, ham, and pickles are very high in: a. thiamin b. fiber c. zinc d. vitamin E e. sodium

94yz

80y

71z

Which of the following is the best snack for a small child? a. hot dog slice b. nuts c. popcorn d. whole grapes e. sliced cheese

86yz

71y

57z

85y

73

48y

85y

72

57y

73y

60

29y

Which food below does NOT contain carbohydrates? a. apple b. Coke c. hamburger meat d. milk e. peas

61yz

43y

33z

A pound of fat equals about ____ kcal. a. 500 b. 1,500 c. 2,500 d. 3,500

56yz

36y

24z

Question

Not enough ____ in the diet has been linked to birth defects involving the brain and spine (spina bifida). a. calcium b. folic acid c. iron d. vitamin A e. vitamin C A food that is labeled “low-fat” always has fewer calories than the regular version of the food. (False) If you were to decrease your food intake by 500 kcal/d, how many pounds would you expect to lose in a week? a. 1-2 b. 2-3 c. 3-4 d. 4-5 e. 5-6

a

e. 4,500

Correct answers are shown in boldface. Means with the same superscript within a row are significantly different (␹2, P⬍.05).

yz

cause the women with obesity correctly answered broad questions regarding health and diet. In our study, knowledge was not indicative of better initial weight control because pretest scores of the healthful-weight (data not presented), overweight, and obese participants did not differ significantly. Demographic characteristics of our low-income population are comparable to previous investigations. Similarities include higher pretest scores in whites vs other ethnicities (46-48), and in those with more formal education (46,49,50) and income (9,50,51). Nutrition knowledge is reported to be greater among those who speak English as their primary language (49); this finding also was present in this study. In addition, our results compare favorably to others who found higher nutrition knowledge levels in households with fewer children (9,49,52). Breastfeeding, not skipping meals, and attending childbirth classes were all related positively to knowledge and were not associated with education, income, or weight status in this study. We have not identified literature that examined the relationship between nutrition knowledge and these behaviors. However, breastfeeding was associated with higher levels of formal education in 2,223 families in Canada (53), and skipping meals was greater

72

January 2006 Volume 106 Number 1

among those with lower education and in the overweight subjects in a sample of 1,431 adults trying to lose weight (54). In our sample, 90% of mothers participated in WIC. Nutrition knowledge of the participants was not significantly different from nonparticipants at either time point, and both groups increased their knowledge scores postintervention. Improvements in health-related knowledge of WIC enrollees has been observed after completion of an intervention designed to enhance acquired immunodeficiency syndrome/human immunodeficiency virus knowledge (55) and fruit and vegetable consumption (56,57). In addition, Havas and colleagues (58) found that awareness of the number of servings of fruits and vegetables required per day was predictive of fruit and vegetable intake in a study of 3,122 WIC participants. Thus, education of WIC mothers can promote specific health behaviors. In contrast, nutrition knowledge in this group may be lacking in some areas. For example, Gupta and colleagues (59) reported that identification of iron-rich foods was poor, and WIC mothers of children with anemia scored similarly to WIC participants with children without anemia on questions related to good food sources of

iron. These discrepancies indicate that nutrition knowledge in WIC mothers can be improved. Maternal knowledge of nutrition was related inversely to child obesity in 189 Mexican-American and 188 white 3- to 5-year-old children whose mothers participated in WIC (47). We did not observe such a relationship; however, the children in our study were somewhat younger (mean age 2.8 years). A limitation of our study was the period of intervention. Yet, limited funding and the highly transient nature and chaotic lifestyle of these low-income, primarily minority women precluded a longer program. These women had great difficulty with financial resources, social support, depression, transportation, and child care. Efforts to provide an extended program were not successful. A follow-up at 6 months was conducted; however, these data are not presented because nutrition knowledge was not evaluated. Use of self-monitoring, but not stimulus control or contingency management (goal setting and rewards), was greater in those with higher nutrition knowledge. Although no studies examining these relationships could be identified, participants most likely to improve their knowledge may have higher levels of motivation and may be more likely to apply techniques (ie, self-monitoring) for successful weight loss. CONCLUSIONS The positive association of nutrition knowledge with weight loss seen in this study suggests these programs should include a strong component of nutrition education to alleviate knowledge inequalities and promote more effective weight control. The majority of participants improved their knowledge scores as a result of the intervention. However, those with greater initial levels of nutrition knowledge were more successful at weight reduction. Spanish-speaking mothers and those with less than a junior high school education had the smallest improvements in their nutrition knowledge. Further segmentation of low-income mothers based on language and education level in weight management interventions may facilitate knowledge improvements. This population of primarily WIC mothers had the greatest knowledge of prenatal and child nutrition, indicating that the WIC programs are successful in this area of their curriculum. But information about the Food Guide Pyramid, weight loss, energy nutrients, and vitamins/minerals needs reinforcement in this low-income population. Consequently, interventions should target these areas to improve knowledge and promote weight loss. Social Cognitive Theory appears to be an effective theoretical framework for designing weight-loss interventions for low-income mothers. This theory emphasizes the use of multiple avenues for behavior change, including environment, skill, and personal change, that can be easily tailored to this population’s needs. This research was supported by grants from the Texas Higher Education Coordinating Board (grant no. UTA#00-377) and the RGK Foundation.

References 1. Flegal KM, Carroll MD, Ogden CL, Johnson CL. Prevalence and trends in obesity among US adults, 1999-2000. JAMA. 2002;288:1723-1727. 2. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA. 2004;291:2847-2850. 3. Zhang Q, Wang Y. Trends in the association between obesity and socioeconomic status in US adults: 1971 to 2000. Obes Res. 2004;12:1622-1632. 4. Townsend MS, Peerson J, Love B, Achterberg C, Murphy SP. Food insecurity is positively related to overweight in women. J Nutr. 2001;131:1738-1745. 5. American Obesity Association. AOA facts sheets: Obesity in minority populations. Available at: http:// www.obesity.org/subs/fastfacts/Obesity_Minority_ Pop.shtml. Accessed October 6, 2005. 6. Williamson DF, Madans J, Pamuk E, Flegal KM, Kendrick JS, Serdula MK. A prospective study of childbearing and 10-year weight gain in US white women 25 to 45 years of age. Int J Obes. 1994;18:561569. 7. Kac G, Benicio MH, Velasquez-Melendez G, Valente JG, Struchiner CJ. Breastfeeding and postpartum weight retention in a cohort of Brazilian women. Am J Clin Nutr. 2004;79:487-493. 8. Li R, Darling N, Maurice E, Barker L, GrummerStrawn LM. Breastfeeding rates in the United States by characteristics of the child, mother, or family: The 2002 National Immunization Survey. Pediatrics. 2005;115:E31-E37. 9. Morton JF, Guthrie JF. Diet-related knowledge, attitudes, and practices of low-income individuals with children in the household. Fam Econ Nutr Rev. 1997; 10:2-14. 10. Elder JP, Candelaria JI, Woodruff SI, Criqui MH, Talavera GA, Rupp JW. Results of Language for Health: Cardiovascular disease nutrition education for Latino English-as-a-second-language students. Health Educ Behav. 2000;27:50-63. 11. Auslander W, Haire-Joshu D, Houston C, Williams JH, Krebill H. Res Soc Work Prac. 2000;10:78-97. 12. Brinberg D, Axelson AL, Price S. Changing food knowledge, food choice, and dietary fiber consumption by using tailored messages. Appetite. 2000;35:3543. 13. Resnicow K, Wallace DC, Jackson A, Digirolamo A, Odom E, Wang T, Dudley WN, Davis M, Mitchell D, Baranowski T. Dietary change through African American churches: Baseline results and program description of the Eat for Life Trial. J Cancer Educ. 2000;15:156-163. 14. Wardle J, Parmenter K, Waller J. Nutrition knowledge and food intake. Appetite. 2000;34:269-275. 15. Kennedy AL, Hunt C, Hodgson P. Nutrition education program based on EFNEP for low-income women in the United Kingdom: “Friends with Food.” J Nutr Educ. 1998;30:89-99. 16. Kessler H, Wunderlich SM. Relationship between use of food labels and nutrition knowledge of people with diabetes. Diabetes Educ. 1999;25:549-559. 17. Arnold CG, Sobal J. Food practices and nutrition

January 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION

73

18.

19.

20. 21. 22.

23. 24.

25. 26.

27.

28. 29. 30.

31.

32.

33.

74

knowledge after graduation from the Expanded Food and Nutrition Education Program (EFNEP). J Nutr Educ. 2000;32:130-138. Colavito EA, Guthrie JF, Hertzler AA, Webb RE. Relationship of diet-health attitudes and nutrition knowledge of household meal planners to the fat and fiber intakes of meal planners and preschoolers. J Nutr Educ. 1996;28:321-328. Dallongeville J, Marécaux N, Cottel D, Bingham A, Amouyel P. Association between nutrition knowledge and nutritional intake in middle-aged men from Northern France. Public Health Nutr. 2000;4:27-33. Smith AM, Baghurst K, Owen N. Socioeconomic status and personal characteristics as predictors of dietary change. J Nutr Educ. 1995;27:173-181. Shepherd A, Stockley M. Nutrition knowledge, attitudes, and fat consumption. J Am Diet Assoc. 1987; 87:615-619. Stafleu A, Van SW, DE Graaf C, Burema J, Hautvast J. Nutrition knowledge and attitudes towards highfat foods and low-fat alternatives in three generations of women. Eur J Clin Nutr. 1996;50:33-41. Trexler ML, Sargent R. Assessment of nutrition risk knowledge and its relationship to the dietary practices of adolescents. J Nutr Educ. 1993;25:337-344. Allison DB, Kanders BS, Osage GD, Faith MS, Heymsfield SB, Heber D, Foreyt JP, Elashoff RM, Blackburn GL. Weight-related attitudes and beliefs of obese African-American women. J Nutr Educ. 1995;27:18-23. Burns CM, Richman R, Caterson ID. Nutrition knowledge in the obese and overweight. Int J Obes. 1987;11:485-492. Agurs-Collins TD, Kumanyika SK, Ten Have TR, Adams-Campbell LL. A randomized controlled trial of weight reduction and exercise for diabetes management in older African American subjects. Diabetes Care. 1997;20:1503-1511. Bruno R, Arnold C, Jacobson L, Winick M, Wynder E. Randomized controlled trial of a nonpharmacologic cholesterol reduction program at the worksite. Prev Med. 1983;12:523-532. Domel SB, Alford BB, Cattlett HN, Gench BE. Weight control for black women. J Am Diet Assoc. 1992;92:346-348. Domel SB, Alford BB, Cattlett HN, Rodriguez ML, Barbara EG. A pilot weight control program for Hispanic women. J Am Diet Assoc. 1992;92:1270-1271. Jeffery RW, Wing RR. Long-term effects of interventions for weight loss using food provision and monetary incentives. J Consult Clin Psychol. 1995;63:793796. Rhodes KS, Bookstein LC, Aaronson LS, Mercer NM, Orringer CE. Intensive nutrition counseling enhances outcomes of National Cholesterol Education Program dietary therapy. J Am Diet Assoc. 1996;96: 1003-1010. Bell AC, Swinburn BA, Amosa H, Scragg RK. A nutrition and exercise intervention program for controlling weight in Samoan communities in New Zealand. Int J Obes. 2001;25:920-927. Braeckman L, DeBacquer D, Maes L, DeBacker G.

January 2006 Volume 106 Number 1

34.

35.

36. 37. 38. 39. 40.

41.

42.

43.

44.

45.

46.

47. 48.

49.

Effects of a low-intensity worksite-based nutrition intervention. Occup Med. 1999;49:549-555. Howard-Pitney B, Winkleby MA, Albright CL, Bruce B, Fortmann SP. The Stanford Nutrition Action Program: A dietary fat intervention for low-literacy adults. Am J Public Health. 1997;87:1971-1976. Kuczmarski RJ, Ogden CL, Guo SS, GrummerStrawn LM, Flegal KM, Mei Z, Wei R, Curtin LR, Roche AF, Johnson CL. 2000 CDC Growth Charts for the United States: Methods and development. National Center for Health Statistics. Vital Health Stat. 2002;11:1-190. McDonald RP. Test Theory: A Unified Treatment. Mahwah, NJ: Erlbaum; 1999. Thomas PR, ed. Weighing the Options: Criteria for Evaluating Weight-Management Programs. Washington, DC: National Academy Press; 1995. Foreyt JP, Poston WS. What is the role of cognitivebehavior therapy in patient management? Obes Res. 1998;6(suppl 1):S18-S22. Bandura A. Social Learning Theory. Englewood Cliffs, NJ: Prentice Hall; 1977. The Food Guide Pyramid. Washington, DC: US Department of Agriculture, Center for Nutrition Policy and Promotion; 1996. Home and Garden Bulletin No. 252. US Department of Agriculture, Center for Nutrition Policy and Promotion. The Food Guide Pyramid for Young Children. Available at: http://www.usda.gov/ cnpp/KidsPyra/PyrBook.pdf. Accessed October 6, 2005. Voruganti VS, Freeland-Graves JH, Proffitt JM, Cai G, Clarke KK, Klohe DM, Milani T, Walker LO. Depression and weight loss in low-income women [abstract]. FASEB J. 2002;16:A1018. Clarke KK, Klohe DM, Cai G, Proffitt JM, Voruganti VS, Milani T, Hans-Nuss H, Walker LO, FreelandGraves JH. The role of self-efficacy in weight management for low-income mothers [abstract]. FASEB J. 2002;16:A611. Klohe DM, Clarke KK, Cai G, Freeland-Graves JH, Milani T, Hanss-Nuss H, Proffitt JM, Walker LO. Body image and eating behavior assessment in obese, low-income women [abstract]. FASEB J. 2001;15: A1087. Clarke KK, Klohe DM, Hanss-Nuss H, Voruganti VS, Proffitt JM, Milani T, Freeland-Graves JH. The influence of nutrition attitudes and beliefs on the food choices of low-income women of young children [abstract]. J Am Diet Assoc. 2003;103(suppl 1):A80. Sapp SG, Jensen HH. Reliability and validity of nutrition knowledge and diet-health awareness tests developed from the 1989-1991 diet and health knowledge surveys. J Nutr Educ. 1997;29:63-72. Sherman JB, Alexander MA, Dean AH, Kim M. Obesity in Mexican-American and Anglo children. Prog Cardiovasc Nurs. 1995;10:27-34. Winkelby MA, Albright CL, Howard-Pitney B, Lin J, Fortmann SP. Hispanic/white differences in dietary fat intake among low-educated adults and children. Prev Med. 1994;23:465-473. Boulanger PM, Pérez-Escamilla R, Himmelgreen D, Segura-Millán S, Haldeman L. Determinants of nu-

50. 51.

52.

53.

54.

trition knowledge among low-income, Latino caretakers in Hartford, Conn. J Am Diet Assoc. 2002;102: 978-981. Parmenter K, Waller J, Wardle J. Demographic variation in nutrition knowledge in England. Health Educ Res. 2000;15:163-174. Harnack L, Block G, Subar A, Lane S. Cancer prevention-related nutrition knowledge, beliefs, and attitudes of US adults: 1992 NHIS cancer epidemiology supplement. J Nutr Educ. 1998;30:131-138. Ivanovic D, Castro CG, Ivanovic R. Food and nutrition knowledge of school-age children’s mothers from elementary and high school from different socioeconomic levels. Arch Latinoamer Nutr. 1997;47:248255. Dubois L, Girard M. Social determinants of initiation, duration and exclusivity of breastfeeding at the population level: The results of the Longitudinal Study of Child Development in Quebec. Can J Public Health. 2003;94:300-305. Levy AS, Heaton AW. Weight control practices of US adults trying to lose weight. Ann Intern Med. 1993; 119:661-666.

55. Ashworth CS, DuRant RH, Gaillard G, Rountree J. An experimental evaluation of an AIDS educational intervention for WIC mothers. AIDS Educ Prev. 1994;6:154-162. 56. Havas S, Anliker J, Damron D, Langenberg P, Ballesteros M, Feldman R. Final results of the Maryland WIC 5-A-Day promotion program. Am J Public Health. 1998;88:1161-1167. 57. Langenberg P, Ballesteros M, Feldman R, Damron D, Anliker J, Havas S. Psychosocial factors and intervention-associated changes in those factors as correlates of change in fruit and vegetable consumption in the Maryland WIC 5-A-Day Promotion Program. Ann Behav Med. 2000;22:307-315. 58. Havas S, Treiman K, Langenberg P, Ballesteros M, Anliker J, Damron D, Feldman R. Factors associated with fruit and vegetable consumption among women participating in WIC. J Am Diet Assoc. 1998;98:11411148. 59. Gupta S, Venkateswaran R, Gorenflo DW, Eyler AE. Childhood iron deficiency anemia, maternal nutritional knowledge, and maternal feeding practices in a high-risk population. Prev Med. 1999;29:152-156.

January 2006 ● Journal of the AMERICAN DIETETIC ASSOCIATION

75