'Fattening' foods--perceptions and misconceptions: a qualitative and quantitative exploration.

Abstract

Objective: To explore perceptions of 'fattening' foods and determine the prevalence of these perceptions among adults using qualitative and quantitative methods.

Design: Twelve tape-recorded focus group discussions with men (n = 6 groups) and women (n = 6 groups) facilitated by three of the authors (AT, CB, DCS), and a postal survey.

Subjects: Convenience sample of 62 adults ([greater than or equal to] 18 years) (focus groups) and 1500 adults randomly selected from the electoral roll of Victoria (47% response rate) (survey).

Setting: Community (focus group participants recruited from workplaces, community groups/organisations and sporting facilities).

Main outcome measures: Beliefs about 'fattening' foods.

Statistical analyses: Taped focus group discussions were transcribed, checked and analysed for common themes using qualitative methodology. For the postal survey, the prevalence of beliefs was described and compared by sex using Chi-square tests. Factor analyses were applied to the belief items. Scores from each factor were compared by sex, BMI and weight-control behaviour using ANCOVA, controlling for age and education.

Results: The focus groups revealed a range of factors that are considered when judging foods as 'fattening', including fat, sugar, salt and calorie content, fat type, label information, balance with physical activity, 'moderation', metabolism, fresh vs processed and 'extra' or 'unnecessary' additions to a meal/snack. In the survey, 30% of respondents believed that if they watch their fat intake, they can eat as much as they like. Only 28% considered it important to check the energy content of foods, although 71% considered foods high in sugar to be 'fattening'. Many respondents disagreed that no foods are 'fattening' if eaten in moderation (41%) or balanced with physical activity (27%). Few respondents (16%) agreed that foods labelled 'reduced fat', 'low fat', 'diet' and 'lite' are not 'fattening' and 55% considered saturated fat more 'fattening' than unsaturated fat.

Conclusion: There are limitations in the public's understandings of 'fattening foods' which are inconsistent with current dietary recommendations.

Key words: beliefs, adults, diet, dietary fat, qualitative, survey

Introduction

The amount and type of dietary fat consumed is considered to play a role in the aetiology of cardiovascular disease, some cancers and weight gain (1). In Australia, cardiovascular disease and cancer are the two leading contributors to the burden of disease (2). In addition, the prevalence of overweight and obesity has risen dramatically over the last 20 years, with 66% of men and 49% of women currently classified as overweight or obese (3). Between 1983 and 1995, energy intake increased among Australian men and women (4). While the proportion of energy derived from fat in the Australian diet appears to have decreased from 37% since 1983 (5), Australian adults still consume more fat than recommended, with average consumption in 1995 being 33% of total energy intake (3,5). Eating a diet low in fat, particularly saturated fat, eating a moderate amount of sugars and foods containing sugars, and preventing unhealthy weight gain are therefore three important dietary guidelines (1).

There is evidence to suggest that consumers are aware of messages to reduce fat intake and maintain a healthy body weight, with 29% of men and 42% of women reporting being on a fat-modified diet and one in four believing they should consume fewer foods high in fat in 1995 (3). Additionally, one in two Australians report taking steps to either reduce or maintain their weight (6), and avoiding specific foods, particularly foods considered to be high in fat, is a commonly reported weight-control strategy (6-8). Adolescents and young adults have been shown to consider some foods to be 'fattening' or to contribute to weight gain (9-12). Among young adults in the US, Sobal and Cassidy (9,13) have shown that 'dieting foods' and 'fattening foods' exist as conceptual, almost mutually exclusive groupings, with the category of 'dieting foods' characterised predominantly by being low in calories, fat and carbohydrates (13). Sobal and Cassidy (9) have found that US consumers considered particular foods in each of the four food groups to be 'fattening', and expressed concern that classifying foods as 'fattening' may lead to poor dietary choices.

Many adults have a poor understanding of sources of dietary fat. European and US research has reported that 13-21% of adults consuming a high-fat diet considered their diet to be low in fat (14-16). While consumers' relative ranking of foods according to fat content has been shown to be reasonable, the fat content of high-fat foods is generally underestimated, and the fat content of low-fat foods generally overestimated (17). Some studies have also found that consumers are confused about the fat content of foods containing different types of fat (18-20). Additionally, adults have been shown to hold 'dose-insensitive' views about fat and calories, with many holding the view that high fat or calorie foods are harmful regardless of the quantity consumed (i.e. consumption of a small amount of fat is considered as harmful as consumption of a large amount of fat) (21).

While consumers appear aware of the need to limit fat intake and attempts to reduce dietary fat intake are common, inaccurate views about the fat content of foods and how to reduce dietary fat intake may undermine efforts to consume less fat. Little is known about how adults classify foods as 'fattening'. The limited research that focussed on adults used small, non-representative samples of mainly young college women (9,10,13). More data are needed to provide insights into how consumers think about 'fattening' foods to better inform nutrition communications and clarify dietary misperceptions. The aims of this study were to explore adults' perceptions of 'fattening' foods and determine the prevalence of these perceptions among adults.

Method

Overview of study design

Qualitative and quantitative data collection techniques were used to examine perceptions of 'fattening' foods among adults. In Part 1, 12 focus group discussions were held with men and women aged 18 years or over. In Part 2, findings from the focus group discussions were used to develop a questionnaire to assess perceptions of 'fattening' foods, and adults whose names were randomly selected from the electoral roll were surveyed.

Part 1: Qualitative exploration

Procedure

Each participant completed a brief background questionnaire to assess socio-demographic variables, including age, country of birth, highest educational attainment and employment status. Details of height and weight, and weight-control status were also requested.

Each focus group session was tape-recorded and followed a detailed discussion guide. Each session began with general discussion about causes of overweight, methods of weight control and definitions of 'nutritious'. To stimulate discussion regarding how consumers conceptualised 'fattening' foods, various food items were presented to the groups, and participants were asked to consider whether that particular food was 'fattening' and to describe the reasoning on which they based their decision. Other items on the discussion schedule are not reported here. The discussion groups lasted 50 to 105 minutes.

Subjects

A convenience sample of men and women aged 18 years and over were recruited from workplaces, community groups, sporting facilities and community organisations to participate in focus group discussions about food, physical activity and weight control. The project was publicised through posters, emails, newsletters, newspaper advertisements and information sessions. Interested individuals telephoned the researchers to volunteer. While all participants were adults (aged 18 years or over), no other recruitment criteria was applied. Recruitment and the conduct of focus group discussions continued until no new information was generated by the focus group discussions (i.e. until saturation was achieved). All volunteers were offered a $15 shopping voucher to compensate for their time.

In total, six focus groups were conducted with women [all facilitated by a female (three facilitated by AT and the remaining three facilitated by CB)], and six groups were conducted with men [four of which were facilitated by a male (DCS) and two conducted by a female (CB)]. Consistency between facilitators was achieved by using a structured discussion guide (described below). In addition, the role of scribe was rotated between the facilitators. Therefore, each facilitator was aware of the techniques used by the other facilitators. Each group comprised five to seven participants.

Analysis

The focus group discussions were transcribed verbatim and checked against the tape recordings to ensure accuracy. A coding framework was developed by the first (AT) and second (CB) authors based on the discussion schedule and major themes that emerged during the discussions. The raw transcripts were coded by AT and CB independently. These coding decisions were compared and differences discussed before final coding decisions were made. There were only five instances where codes regarding conceptualisation of 'fattening' foods differed. In each of these cases, one of the coders had neglected to code a segment of text related to 'fattening' foods. These segments were subsequently coded. The transcripts were converted to The Ethnograph (Qualis Research Associates, Amherst, The Ethnograph, version 4.0 software package 1996) and the final codes were entered against the relevant segments of text. Segments of text relevant to 'fattening' foods were extracted from The Ethnograph and key findings summarised under major themes. The analyses were performed by AT, and CB provided feedback throughout the process.

Self-reported height and weight were used to calculate body mass index (BMI = kilograms/[metres.sup.2]), which was collapsed into four categories as recommended for the Australian population (22): Underweight - BMI < 20; Acceptable weight = 20 [greater than or equal to] BMI [less than or equal to] 25; Overweight = 25 > BMI [less than or equal to] 30; Obese = BMI > 30.

Part 2: Population survey

Procedure

A self-administered questionnaire was mailed to 1500 Australian citizens drawn randomly from the electoral roll for the state of Victoria, Australia (enrolment is compulsory for all Australian citizens aged 18 years or above). A postcard reminder was mailed to non-respondents two weeks after the initial mail-out and a reminder letter with a replacement questionnaire was mailed to remaining non-respondents after a further two weeks. On the basis of advice from a consultant biostatistician, and based on our previous survey research experience, a sample of this size was considered to be adequate for the purposes of this.

The questionnaire included the measures described below, in addition to a series of additional questions (not discussed in this paper) about physical activity, eating and weight control. Prior to administration, the questionnaire was pilot tested in a group of 15 adults aged 22 to 60 years. The questionnaire took approximately 20 minutes to complete and no changes were made to the measures described below.

Measures

Background: The questionnaire requested details of sex, age, country of birth, highest educational attainment and employment status. Age was collapsed into six categories: 18-29 years; 30-39 years; 40-49 years; 50-59 years; 60+ years.

Body mass index: Self-reported height and weight were used to calculate BMI, which was collapsed into four categories of underweight, acceptable weight, overweight and obese, as described earlier.

Weight-control behaviour: Weight-control behaviour was assessed by a single question asking respondents to nominate which of the following categories best describes them: actively doing things to try to gain weight; activity doing things to try to avoid weight gain; actively doing things to try to lose weight; not doing anything in particular for their weight.

Perceptions of 'fattening' foods: Key findings from Part 1 were used to develop questionnaire items to assess perceptions of the characteristics of 'fattening' foods. A series of 15 statements was designed based on the most common themes and the reported considerations that were used to decide if a food was 'fattening' (as presented in the Results section). Wherever possible, the phrasing and language used by participants in Part 1 were used or adapted, although some items were phrased in the negative. Respondents were asked to indicate whether they agree or disagree with the series of 15 statements (listed in Table 2), for which the following response categories were provided: disagree; unsure/neutral; agree.

Analysis

Women who were pregnant at the time of the survey or had been pregnant over the previous two years, and women for whom this data were missing, were excluded from all analyses (n = 34). Adults who reported actively trying to gain weight at the time of the survey were excluded (a further seven respondents) from analyses that included weight-control behaviour. This group was too small for comparative purposes and too distinct to be combined with any of the other three categories.

The prevalence of beliefs about 'fattening' foods was determined by frequency counts, and each belief was then examined by sex, and between categories of age, level of education, BMI and weight-control behaviour among men and women, using Pearson's [chi square] tests of significance.

Principal components factor analysis (with varimax rotation) was applied to the 15 items relating to perceptions of 'fattening' foods. Five factors were extracted. To assess internal consistency, Cronbach's alpha was computed for each factor. Items that loaded on the factors were scored as follows: disagree (-1); unsure/neutral (0); agree (1) (items that were worded as 'not fattening' were scored in reverse order). For each Factor, the scores for each item that loaded above 0.4 were summed and standardised. Accordingly, a higher score indicates a greater belief that the items that comprise the factor make foods 'fattening'. The standardised scores for each of the four factors were compared by sex using independent t-tests, and by BMI and weight-control behaviour within sex, controlling for age group and level of education, using analysis of covarianee (ANCOVA).

Results

Part 1: Qualitative exploration

Profile of participants

Sixty-two adults (45% women) took part in the focus group discussions. The mean age ([+ or -] SD) of participants was 42.3 ([+ or -] 14.5) years. Most were Australian-born (80%) and employed (73%), and had completed a university or tertiary qualification (43%), with 28% having completed up to high school and 30% having a technical or trade certificate or apprenticeship. Most participants were classified as either overweight (42%) or obese (26%). At the time of the focus group discussions, 21% were trying to maintain their weight and 39% were trying to lose weight.

Beliefs about 'fattening 'foods

In the introductory discussion regarding causes of weight gain and methods of weight control, participants spontaneously used the term 'fattening'. The types of foods participants spontaneously cited as being 'fattening' included cakes and slices, chocolates and chocolate bars, lamingtons, biscuits, sweets, doughnuts and snack foods (described as 'goodies'), deep-fried products (such as fish and chips, battered fish, dim sims and potato cakes) and takeaway foods (such as 'junk food', French fries, thick shakes and sundaes). Additional foods perceived to be 'fattening' included alcohol, 'processed' foods (pre-prepared foods, ready-made sauces, two-minute noodles), 'excess' food (such as chocolate, morning and afternoon tea), dairy products (milk, cheese, cream, yoghurt, ice cream, butter), margarine, and meat and meat products (e.g. meat pies, pasties, sausages, chicken, ham). Bread, cereals, avocado, peanuts, crisps and salty foods were also cited as fattening. When prompted to describe the kinds of factors that are considered when deciding whether a food or food group is 'fattening', a variety of themes emerged, as summarised below.

Fat content

The most frequently mentioned consideration in determining whether a food was 'fattening', and that which the majority of participants felt to be most important, was the fat content of the food. Most participants acknowledged that consuming some fat is healthy and that it is only when dietary fat is consumed in excess that it becomes 'fattening'. Some participants recognised that there are two types of dietary fat. These were referred to in many different ways: good and bad fat; healthy kind of fat and unhealthy kind of fat; natural fat and processed fat; not concentrated and concentrated fat; and saturated and unsaturated fat. However, participants were divided over whether the two types of fat are equally as 'fattening'.

Balance

Some participants raised the concepts of 'balance' and 'moderation', while others believed that it is important to consider physical activity levels, when deciding whether a food is 'fattening'. In general, participants believed that different foods offer different benefits and that the diet should therefore contain 'good' and 'bad' foods, although only a small amount of 'bad' foods. Moderation and balance were referred to most often in the context of limiting the amount of a food consumed, rather than choosing a wide variety of foods.

Sugar and calories

Sugar content was also commonly mentioned as a consideration when determining whether a food is 'fattening'. Most participants believed that sugar is 'fattening'. Many specified, however, that sugar is 'fattening' only when consumed in excess. There was a view that sugar found in fruit is not as 'fattening' as sugar from other sources, as it can be 'burned' quickly. Some participants did not believe sugar is 'fattening', but acknowledged that sugar contains calories. Although infrequently mentioned, some participants reported checking the calorie or kilojoule content of food when judging whether it is 'fattening'.

Fresh versus processed foods

A popular consideration when deciding whether a food is 'fattening' was whether a food was fresh or processed. Processed or ready-made foods were considered 'fattening' as they have 'obviously been tampered with'. It was assumed that through processing, 'things', such as oils and preservatives, are added which make the food 'fattening'. Fresh, natural foods were not considered to be 'fattening' as nothing is added (with the exception of avocado and nuts).

Extra/unnecessary foods

Some participants also believed that 'extra' or 'unnecessary' foods are 'fattening'. For example, one participant commented, 'I've always been brought up with anything you add on, all your mayonnaise, your toppings, sauces, everything, it's that little bit extra that is the weight gaining.' Consistent with this, participants commented that certain foods could be 'fattening' depending on how they are eaten, the 'extras' that may be added, and how the food is prepared. Participants were divided over whether salt content is an important aspect of the concept of 'fattening' foods. Those that considered salt content to be important associated salt intake with water retention.

Labels

A common theme in considering whether a food is 'fattening' was the use of food labels, primarily to determine the fat content of food. Most participants reported using labels to compare products, while others used them to gauge absolute fat levels. They commented that different types of fats, not understanding labels and lack of labels on certain products make using labels difficult. Some participants were wary of labels, as statements such as 'lite' may have different meanings that do not relate to fat.

Many participants mentioned the National Heart Foundation's tick of approval, but perceptions of its legitimacy varied (a product labelling program where 'approved' foods are relatively low in saturated fat and salt, and high in fibre). Some felt the tick was a good indication of whether a food was 'fattening'. Others were concerned that payment for the endorsement by food companies jeopardises the objectivity of the assessment, and that the tick may relate to specific nutrients at the expense of others. Other participants felt the tick is misleading. One commented, 'I think that's misleading [the tick] in a lot of cases as well. They might say well that's the best of a bad bunch, but to me that's still really bad.'

Some participants reported that they base their decision regarding 'fattening' foods on information sources other than labels. These included weight watchers literature (described as a 'little Bible'), fat counter books, friends and the media.

Part 2: Population survey

Profile of respondents

A response rate of 47% was achieved (based on the return of 681 completed questionnaires, and 42 questionnaires confirmed as addressed to persons unable to be contacted). A profile of survey respondents is presented in Table 1. The final sample consisted of similar proportions of men and women, and respondents were evenly distributed across age groups. Most respondents were Australian-born, employed full or part time, and well educated, although 40% of women had not completed high school. Compared to census data, the profile of respondents was generally similar to the population of Victoria, although the sample consisted of a slightly greater proportion of women in the workforce (23,24).

Similar to nationally representative data from the National Health Survey (25), 47% of respondents were classified as overweight or obese, with greater proportions of men classified as overweight compared to women, and few respondents classified as underweight (Table 1). About half of the respondents were not trying to control their weight in any way at the time of the survey, more than a quarter were trying to avoid gaining weight and almost one in five were actively attempting to reduce their weight. Very few respondents were trying to increase their weight. Greater proportions of women than men were actively trying to reduce or maintain their weight.

Beliefs about 'fattening' foods

Responses to the 15 statements about 'fattening' foods are presented for the total sample and according to sex in Table 2. Although most respondents generally considered the 'fattening' potential of foods to be mediated by 'moderation' in food habits (45%) and balance with physical activity (58%), a high proportion considered foods to be 'fattening' regardless of the amount consumed (41%), and one in four believed that foods can be 'fattening' even if balanced with physical activity (27%).

Most respondents believed that sugar content (71%) and fat content (79%) are indications of a 'fattening' food. Almost one in three respondents believed that quantity of food intake is unimportant, as long as fat intake is limited. Consistent with this, just over half of the respondents did not consider the energy content of foods an important consideration when deciding if a food is 'fattening'. Most respondents (67%) agreed that 'a good way to tell if a food is 'fattening' is to check the label', however comparatively few (16%) did not consider foods labeled 'reduced fat', 'low fat', 'diet' and 'lite' to be 'fattening', and 55% considered saturated fat to be more 'fattening' than unsaturated fat. Almost three in four (73%) disagreed that the degree of processing decreases the likelihood that a food will be 'fattening'. Consistent with this, 60% agreed that fresh, natural foods are not 'fattening'.

Compared to men, women held stronger beliefs about 'fattening' foods. Greater proportions of women believed that foods high in fat, sugar or salt are 'fattening' (Table 2). Greater proportions of women compared to men also believed that unnecessary additions to food make them 'fattening', that saturated fat is more 'fattening' than unsaturated fat, that fresh, natural foods are not 'fattening', and that using food labels is a good method of checking fat content, and disagreed that foods labeled 'reduced fat' 'low fat', 'diet' or 'lite' are not 'fattening'.

Factor analysis of beliefs about 'fattening' foods

Principal components factor analysis of the 15 items assessing perceptions of 'fattening' foods resulted in five factors that accounted for 49.9% of variance. The items that loaded on each of these factors, and Cronbach's alpha ([alpha]) for each of these factors, are shown in Table 2. The first factor, 'unnecessary additions', was comprised of four items and accounted for 14.8% of variance. The second factor, 'fat', included four items and accounted for 10.0% of variance. The third factor, 'balance/moderation', was comprised of two items (9.2% variance) and the fourth factor, 'energy', included two items (8.6% variance). The final factor, 'metabolism', comprised three items (7.4% variance). Scores were not computed for the 'metabolism' factor due to low internal consistency, and one item ('as long as I watch my fat intake, I can eat as much as I like') was not included in the summary score for 'energy' as the inclusion of this item lowered the internal consistency of the factor.

Standardised scores for each of the four 'fattening' foods factors are shown in Table 3. The scores for each factor were negative among men and positive among women. For men, the highest scores were obtained for the 'unnecessary additions' and 'balance/moderation' factors. For women, the highest score was obtained for the 'fat' factor. Women's scores were significantly higher than men's scores for the 'fat' (F = 12.32, df = 609, P < 0.001) and 'energy' (F = 7.61, df= 605, P = 0.006) factors.

After controlling for age and education, there was little variation in standardised scores on any factors according to BMI status among men or women. However, overweight or obese women had higher scores than other women on the balance/moderation factor. That is, overweight or obese women held stronger beliefs that foods are 'fattening' regardless of whether consumed in moderation or if balanced with physical activity. Some differences in beliefs about 'fattening' foods were found according to weight-control behaviour. Compared to other men, men who were trying to lose weight held stronger beliefs that unnecessary additions to food are 'fattening', and men who were trying to avoid weight gain held stronger beliefs about the role of fat in 'fattening' foods. Women who were trying to avoid weight gain held stronger beliefs than other women that unnecessary additions to food are 'fattening'.

Discussion

This study used qualitative techniques to explore men and women's beliefs about the concept of 'fattening' foods and quantitative techniques to measure the prevalence of these beliefs. Consistent with previous work among college students (9,13), this study found that adults use the term 'fattening' to describe foods. While we acknowledge that perceptions of food are more complex than the simple categorisation of food as 'fattening' or 'non-fattening', other studies have also found categorisation of foods into dichotomies such as either 'healthy' or 'junk' foods(11), or 'good' or 'not good' (21,26), and others have shown that 'fattening' aspects of food are considerations in food selection (10-12).

In Part 1, a variety of factors emerged as considerations for judging foods as 'fattening' or not fattening, including fat, sugar, salt and calorie content, fat type, label information, balance with physical activity, 'moderation', metabolism, the notion of fresh versus processed food, and 'extra' or 'unnecessary' additions to a meal/snack. Some of these beliefs were also evident among survey respondents in Part 2. The survey findings suggest that the role of dietary fat in weight control is well known, but the link between sugar, fat and energy is not well understood, especially considering that many adults in the survey believed that as long as they watch their fat intake, they can eat as much as they like. Further, more than half believed that it is not important to check the energy content of foods, even though foods with high sugar content were considered 'fattening'. This confused view is perhaps not surprising, since dietary guidelines and other nutrition communications place most emphasis on limiting fat and sugar in the diets of Australians, with little mention of total energy intake (1). The results indicate a poor understanding of relationships between macronutrients and total energy content.

Importantly, many participants recognised that although a food may be 'fattening', its potential to be stored is limited, or can be mediated by balancing food intake, and also by engaging in physical activity. This is consistent with findings by Paxton, Sculthorpe and Gibbons (27), where 87% of their sample believed that eating a more balanced diet is a somewhat effective method of weight loss (27). The types of foods our focus group participants considered to be 'fattening' include core foods such as dairy foods, and meat and meat products, and are consistent with previous research (9,19). Given that exclusion of these foods from the diet may result in nutritional deficiencies, the view that all foods should be consumed to achieve a 'balance' is encouraging. Consistent with the finding by Rozin et al. (21) that adults hold 'dose-insensitive' views about fat and calories, it is noteworthy that a high proportion of adults considered foods to be 'fattening', regardless of the amount consumed or the amount of physical activity undertaken.

Three other clear findings regarding consumers' perceptions of 'fattening' foods are apparent from this study. Firstly, foods that are 'extra' or 'unnecessary' are considered to be 'fattening', while foods that are natural or fresh are not. Furthermore, most respondents disagreed that degree of processing decreases the likelihood of a food being 'fattening'. Findings from the qualitative component of the study suggest that this may be because of a belief that processed foods are tampered with, diminishing their goodness or health quality, and fats, additives, preservatives and chemicals are added to processed foods. Together, these findings indicate that for food to be considered non-fattening, and therefore 'allowed' or not avoided, it must be basic, and prepared without addition of non-essential ingredients. Such a perception may lead to the commonly held view that dieting takes the pleasure out of meals and that foods low in fat lack taste (6). Sobal and Cassidy (13) have also found that the concept of 'dieting' foods is linked with the notion of goodness, and that dieting foods are considered so due to their 'nutritious', 'natural', 'fresh', 'healthy' and 'pure' nature. By contrast, this suggests that foods considered 'fattening' are unhealthy.

Secondly, most adults disagreed that foods labelled as 'reduced fat', 'low fat, 'diet' or 'lite' are not 'fattening'. This is encouraging, particularly as there is speculation in the US that fat-modified food products may give people license to overeat (28). Thirdly, more than half of the adults in this study considered that saturated fat is more 'fattening' than unsaturated fat. This is consistent with previous research that has found that consumers have a poor understanding about different types of fats (19,20). The labelling of fats in the focus groups as 'good fat', 'healthy fat' and "natural fat' (as opposed to 'bad fat', 'unhealthy fat' and 'processed fat') goes some way to explaining this finding, and may be due to media interest in the beneficial role of particular fats in cardiovascular function.

It is possible that use of the electoral roll to randomly select a sample for the postal survey may have resulted in biases. While it is compulsory for all Australian citizens aged 18 years and above to register, this method of sampling excludes Australian residents and new arrivals, and address details may not be up to date. However, while only a modest response rate was obtained in the postal survey, the profile of respondents was similar to census data from the Victorian population (23,24). Additionally, the prevalence of overweight and obesity was comparable to national data based on self-report (25) and the prevalence of weight-control behaviour was consistent with previous reports (6). However, it is possible that the results of this study may reflect extreme views about food and eating. Adults that volunteered to take part in the focus group discussions, and those who responded to the survey, may have been more interested in these issues than the general public. Therefore, the terminology used in the focus groups and the subsequent questions developed for the survey may not have been representative, and the results of the survey may be biased. In addition, the internal consistency of the four 'fattening' foods factors was moderate ([alpha] = 0.4-0.6). Nevertheless, this study identified a range of factors that consumers take into account when considering whether a food is 'fattening'. Although fat content was the primary determinant of a 'fattening' food, many consumers did not consider energy content to be important and did not appear to understand the concept of energy balance. Furthermore, some consumers considered factors other than fat and energy to be features of 'fattening' foods.

This study has highlighted inconsistencies between lay and professional views about dietary fat that may have important implications for policy development and nutrition communication. Effective nutrition communication about dietary fat and weight control requires an awareness of lay beliefs and misconceptions. Nutritionists need to consider lay views about 'fattening' foods to avoid sending mixed messages and reinforcing inappropriate views about food and eating.

Table 1. Profile of survey respondents--by sex

                                            Total
                                          sample (a)
                              Sex (%)        (%)

                           Men    Women

Overall (n)               (305)   (318)     (647)
                           49.0    51.0      100
Age groups (n)            (304)   (318)     (646)

18-29                      17.8    18.6      18.3
30-39                      18.1    18.6      18.1
40-49                      18.8    23.0      21.1
50-59                      20.4    17.6      18.7
60+                        25.0    22.3      23.8
                          (305)   (318)     (646)
Country of birth (n)

Australia                  73.1    79.6      76.0
Other                      26.9    20.4      24.0

Education level (n) (b)   (304)   (318)     (644)

High school or less        28.0    40.3      34.2
Technical/trade            44.4    28.3      36.5
  certificate/
  apprenticeship
University/tertiary        27.6    31.4      29.3
  qualification

Employment (n)            (305)   (318)     (643)

Employed                   70.5    63.5      66.4
                          (300)   (307)     (627)
BMI category (n) (b)

Underweight                 2.7     9.1       5.7
Acceptable weight          43.7    51.5      47.2
Overweight                 41.7    26.4      34.3
Obese                      12.0    13.0      12.8

Weight-control            (301)   (311)     (635)
  behaviour (n) (b)

Trying to gain weight       5.0     1.6       3.3
Trying to avoid weight     23.3    30.2      26.3
  gain
Trying to lose weight      12.6    22.2      18.1
Doing nothing in           59.1    46.0      52.3
  particular for weight

(a) Total number (n) is greater than the sum of
male and females, as the sex of some participants
is unknown.

(b) p < 0.001: [chi square] test of significance
between men and women.

Table 2. Beliefs about 'fattening' foods among
survey respondents--overall and by sex

                                                     Response (%)

                                               Dis-    Unsure/
Statement                       Sex       n    agree   neutral   Agree

Unnecessary additions (a)
([alpha] = 0.6)

Foods that are high in       Total (c)   634   46.2     32.3     21.5
salt are 'fattening'         Men         300   43.3     37.7     19.0
                             Women *     311   48.2     27.7     24.1

It is the things that you    Total (c)   638   23.8     22.6     53.6
add to food, the 'extra'     Men         301   28.6     25.9     45.5
or 'unnecessary' things,     Women ***   314   20.4     19.4     60.2
that make foods
'fattening'

Because things are added     Total (c)   635   28.3     39.7     32.0
to processed foods, they     Men         297   27.6     39.4     33.0
become 'fattening'           Women       315   28.6     40.0     31.4

Fresh, natural foods are     Total (c)   639   23.5     16.1     60.4
not 'fattening'              Men         302   22.5     20.9     56.6
                             Women *     313   24.5     12.4     63.1

Balance/Moderation (a)
([alpha] = 0.6)

No foods are 'fattening'     Total (c)   640   40.9     13.8     45.3
if you eat them in           Men         302   45.0     15.2     39.7
moderation                   Women *     315   37.1     13.0     49.8

No foods are 'fattening'     Total (c)   639   26.6     15.0     58.4
if you balance what you      Men         303   26.1     17.5     56.4
eat with the amount of       Women       313   27.5     13.1     59.4
activity you do

Fat (a) ([alpha] = 0.4)

A good way to tell if a      Total (c)   639   20.2     12.4     67.4
food is 'fattening' is to    Men         302   24.5     16.2     59.3
check the label              Women ***   314   16.2      9.2     74.5

It is foods with a high      Total (c)   639    8.1     12.5     79.3
fat content that are         Men         302    8.6     15.9     75.5
'fattening'                  Women *     315    8.3      9.2     82.5

Energy (a) ([alpha] = 0.4)

As long as I watch my fat    Total (c)   642   54.7     15.3     30.1
intake, I can eat as much    Men         304   49.3     17.8     32.9
as I like (b)                Women *     315   60.3     12.4     27.3

The more processed a food    Total (c)   639   72.9     16.9     10.2
is, the less likely          Men         303   70.0     19.5     10.6
it is to be 'fattening'      Women       313   77.0     13.4      9.6

Foods labelled 'reduced-     Total (c)   638   62.9     21.5     15.7
fat', 'low fat',             Men         300    8.0     39.0     53.0
can burn fat                 Women       314    5.4     35.7     58.9

Metabolism (a)
([alpha] = 0.2)

When deciding if a food is   Total (c)   641   54.1     17.6     28.2
'fattening', it is           Men         303   49.8     20.8     29.4
important to check how       Women       315   58.1     14.6     27.3
many calories it has

Saturated fat is more        Total (c)   636   14.6     30.7     54.7
'fattening' than             Men         301   11.6     36.9     51.5
unsaturated fat              Women **    313   17.6     24.3     58.1

Your body can burn sugar     Total (c)   637    6.9     37.2     55.9
quicker than it can          Men         300    8.0     39.0     53.0
burn fat                     Women       314    5.4     35.7     58.9

* P < 0.05, ** P < 0.01, *** P [less than or equal to] 0.001:
[chi square] test of significance between men and women.

(a) Factors identified from factor analysis; [alpha] = Cronbach's
alpha; items listed under 'metabolism' are not considered a robust
factor, as [alpha] for these items was < 0.4.

(b) Although this item loaded on the 'energy' component, it was not
included in the final score as inclusion of this item lowered internal
consistency ([alpha] < 0.4).

(c) Total 'n' is greater than the sum of male and females, as the
sex of some participants is unknown.

Table 3. Mean standardised scores (z-scores) on four components of
'fattening' foods derived from factor analysis, for men and women
within weight status and weight-control behaviour groups (a)

                                           Weight status (c)

                                             Under/
                                           acceptable    Overweight/
                           Overall (b)       weight         obese

Men (n)                       (300)           (138)         (158)

'Unnecessary additions'   -0.03 (1.0)      -1.14 (l.0)    0.06 (l.0)
'Fat'                     -0.17 (1.1) **   -0.15 (1.0)   -0.19 (1.1)
'Balance/moderation'      -0.04 (l.0)      -0.07 (l.0)   -0.01 (1.0)
'Energy'                  -0.15 (1.0) *    -0.15 (l.0)   -0.13 (l.0)

Women (n)                     (313)           (183)         (119)

'Unnecessary additions'    0.09 (1.0)       0.04 (l.0)    0.17 (l.0)
'Fat'                      0.11 (1.0) **    0.05 (1.0)    0.19 (1.0)
'Balance/moderation'       0.09 (1.0)      -0.03 (l.0)    0.31 (1.0)
'Energy'                   0.06 (1.0) *     0.11 (0.9)    0.00 (1.0)

                                                      Weight-control
                             Weight status (c)        behaviour (c)

                             F-statistic,             Trying to avoid
                                  df        P-value     weight gain

Men (n)                                                    (69)

'Unnecessary additions'       1.77, 284      0.185      -0.11 (1.1)
'Fat'                         0.21, 295      0.645       0.10 (0.9)
'Balance/moderation'          0.09, 294      0.761      -0.19 (0.9)
'Energy'                      0.19, 292      0.663      -0.13 (l.0)

Women (n)                                                  (94)

'Unnecessary additions'       0.32, 294      0.575      0.24 (l.0)
'Fat'                         1.52, 301      0.218      0.18 (0.9)
'Balance/moderation'          5.59, 300      0.019      0.01 (1.0)
'Energy'                      0.23, 300      0.635      0.02 (l.0)

                          Weight-control behaviour (c)

                          Trying to      No weight
                          lose weight     control

Men (n)                      (38)          (176)

'Unnecessary additions'    0.29 (l.1)   -0.07 (1.0)
'Fat'                     -0.02 (0.9)   -0.30 (1.1)
'Balance/moderation'      -0.21 (1.2)    0.07 (1.0)
'Energy'                   0.14 (0.9)   -0.15 (1.0)

Women (n)                    (69)          (141)

'Unnecessary additions'    0.14 (0.9)   -0.08 (1.0)
'Fat'                      0.18 (0.9)    0.04 (l.0)
'Balance/moderation'       0.17 (l.0)    0.11 (1.0)
'Energy'                   0.28 (0.9)   -0.03 (l.0)

                          Weight-control
                          behaviour (c)

                            F, df     P-value

Men (n)

'Unnecessary additions'   3.32, 271    0.038
'Fat'                     4.29, 281    0.015
'Balance/moderation'      2.63, 281    0.074
'Energy'                  1.06, 278    0.348

Women (n)

'Unnecessary additions'   5.58, 295    0.004
'Fat'                     0.89, 302    0.411
'Balance/moderation'      0.98, 301    0.375
'Energy'                  1.63, 278    0.197

(a) A higher score (range of -1.0 to 2.0) reflects a greater belief
that the issues assessed in these items are considered to make
foods 'fattening'.

(b) Independent t-test between men and women: * P [less than or
equal to] 0.006; ** P [less than or equal to] 0.001.

(c) Analysis of co-variance (ANCOVA): Controlled by age
group and level of education.

Acknowledgement

David Crawford is supported by a National Health and Medical Research Council/National Heart Foundation Career Development Award.

References

(1.) National Health and Medical Research Council. Dietary Guidelines for Australians. Canberra: Australian Government Publishing Service; 1992.

(2.) Mathers C, Vos T, Stevenson C. The burden of disease and injury in Australia. Canberra: Australian Institute of Health and Welfare; 1999.

(3.) Australian Bureau of Statistics. National Nutrition Survey: Selected Highlights, Australia 1995. Catalogue No 4802.0. Canberra: Australian Government Publishing Service; 1997.

(4.) Cook T, Rutishauser IHE, Allsopp R. The bridging study--comparing results from the 1983, 1985 and 1995 Australian national nutrition surveys. Canberra: Commonwealth of Australia; 2001.

(5.) Cashel KM, Greenfield H. Principal sources of dietary fat in Australia: evidence from apparent consumption data and the national survey of adults. Br J Nutr 1994;71:753-73.

(6.) Timperio A, Cameron-Smith D, Burns C, Crawford D. The public's response to the obesity epidemic in Australia: weight concerns and weight control practices of men and women. Public Health Nutr 2000;3:417-24.

(7.) Crawford D, Owen N, Broom D, Worcester M, Oliver G. Weight-control practices of adults in a rural community. Aust N Z J Public Health 1998;22:73-9.

(8.) Neumark-Sztainer D, Sherwood NE, French SA, Jeffery RW. Weight control behaviors among adult men and women: Cause for concern? Obesity Res 1999;7:179-88.

(9.) Sobal J, Cassidy CM. University students' perceptions of fattening and dieting foods in relationship to the four food groups. Nutr Res 1990;10:145-54.

(10.) Worsley A, Leitch D. Students' perceptions of favourite and disliked foods. J Hum Nutr 1981;35:173-87.

(11.) Chapman G, Maclean H. "Junk Food" and "Healthy Food": Meanings of food in adolescent women's culture. J Nutr Education 1993;25:108-13.

(12.) Contento IR, Michela JL, Williams SS. Adoleseent food choice criteria: Role of weight and dieting status. Appetite 1995;25:51-76.

(13.) Sobal J, Cassidy CM. Dieting foods: Conceptualizations and explanations. Ecology of Food and Nutrition 1987;20:89-96.

(14.) Brug J, van Assema P, Kok G, Lenderink T, Glanz K. Self-rated dietary fat intake: Association with objective assessment of fat, psychosocial factors, and intention to change. J Nutr Education 1994;26:218.

(15.) Glanz K, Brug J, van Assema P. Are awareness of dietary fat and actual fat consumption associated?--A Dutch-American comparison. Eur J Clin Nutr 1997;51:542-7.

(16.) Lloyd HM, Paisley CM, Mela DJ. Changing to a low fat diet: attitudes and beliefs of UK consumers. Eur J Clin Nutr 1993;47:361-73.

(17.) Mela DJ. Consumer estimates of the percentage energy from fat in common foods. Eur J Clin Nutr 1993;47:735-40.

(18.) Reid DJ, Hendricks SM. Consumer understanding and use of fat and cholesterol information on food labels. Can J Public Health 1994;85:334-7.

(19.) Wiseman J. Beliefs about food components, foods, fat and heart disease in New Zealand. British Food J 1994;96:14-9.

(20.) Reid DJ, Conrad SA, Hendricks SM. Tracking nutrition trends, 1989-1994: An update on Canadians' attitudes, knowledge and reported actions. Can J Public Health 1996-87:113-8.

(21.) Rozin P, Ashmore M, Markwith M. Lay American conceptions of nutrition: Dose insensitivity, categorical thinking, contagion, and the monotonic mind. Health Psychol 1996;15:438-47.

(22.) National Health and Medical Research Council. Acting on Australia's weight: a strategic plan for the prevention of overweight and obesity. Canberra: Australian Government Publishing Service; 1997.

(23.) Australian Bureau of Statistics. Labour Force, Victoria 1996. Catalogue No 6202.2. Canberra: Australian Government Publishing Service; 1996.

(24.) Australian Bureau of Statistics. Census of Population and Housing: Selected Social and Housing Characteristics for Statistical Local Areas, Victoria 1996. Catalogue No 2015.2. Canberra: Australian Government Publishing Service; 1996.

(25.) Australian Bureau of Statistics. National Health Survey: Summary Results, Australian States and Territories. Catalogue No 4368.0. Canberra: Australian Government Publishing Service; 1997.

(26.) Murcott A. Talking of good food: An empirical study of women's conceptualizations. Food and Foodways 1993;5:305-18.

(27.) Paxton SJ, Sculthorpe A, Gibbons K. Weight-loss strategies and beliefs in high and low socioeconomic areas of Melbourne. Aust J Public Health 1994;18:412-7.

(28.) Sigman-Grant M. Can you have your low-fat cake and eat it too? The role of fat-modified products. J Am Diet Assoc 1997:97:76S-81S.

School of Health Sciences, Deakin University, Melbourne

A. Timperio, BAppSc(Hons), PhD, Lecturer 

C. Burns, BSc, DipDiet, PhD, Lecturer

D. Cameron-Smith, BSc(Hons), PhD, Senior Lecturer 

D. Crawford, BSc, PhD, Associate Professor

Correspondence: A. Timperio, Centre for Physical Activity and Nutrition Research, Deakin University, 221 Burwood Highway, Burwood VIC 3125. Email: [e-mail]