OBJECTIVE: To investigate the prevalence and predictors of weight maintenance over time in a large sample of young Australian women.
DESIGN: This population study examined baseline and 4 y follow-up data from the cohort of young women participating in the Australian Longitudinal Study on Women's Health.
SUBJECTS: A total of 8726 young women aged 18–23 y at baseline.
MEASURES: Height, weight and body mass index (BMI); physical activity; time spent sitting; selected eating behaviours (eg dieting, disordered eating, takeaway food consumption); cigarette smoking, alcohol consumption; parity; and sociodemographic characteristics.
RESULTS: Only 44% of the women reported their BMI at follow-up to be within 5% of their baseline BMI (maintainers); 41% had gained weight and 15% had lost weight. Weight maintainers were more likely to be in managerial or professional occupations; to have never married; to be currently studying; and not to be mothers. Controlling for sociodemographic factors, weight maintainers were more likely to be in a healthy weight range at baseline, and to report that they spent less time sitting, and consumed less takeaway food, than women who gained weight.
CONCLUSIONS: Fewer than half the young women in this community sample maintained their weight over this 4 y period in their early twenties. Findings of widespread weight gain, particularly among those already overweight, suggest that early adulthood, which is a time of significant life changes for many women, may be an important time for implementing strategies to promote maintenance of healthy weight. Strategies which encourage decreased sitting time and less takeaway food consumption may be effective for encouraging weight maintenance at this life stage.
The adverse effects of overweight and obesity on health and longevity are well documented.1 Overweight and obesity increase risk of morbidity and/or mortality from numerous chronic conditions including type 2 diabetes mellitus, coronary heart disease, hypertension, hyperlipidaemia, osteoarthritis and certain cancers. In many developed countries, overweight and obesity have reached epidemic proportions, and now pose a significant threat to the health of populations throughout the world.
The prevalences of overweight and obesity have increased substantially over the past several decades.1 For example, national surveys conducted in the USA between 1962 and 1988–1994, show that the age-adjusted prevalence of overweight (body mass index (BMI) 25–30 kg/m2) increased from 48 to 59% in men, and from 39 to 59% in women during this period.2 Moreover, the prevalence of obesity (BMI>30 kg/m2) reportedly increased from 12% in 1991 to 18% in 1998.3 In Britain, the prevalence of clinical obesity doubled between 1980 and 1991, and continues to increase,4 with 17.5% males and 20% females obese in 1997.2 Similar trends have been observed in many European countries.2 In Australia, according to the 1995 National Nutrition Survey, almost one in five adults are obese, and a further 45% of men and 29% of women are overweight.5 While cross-sectional data suggest that the proportion of overweight or obese men increased by 23% between 1980 and 1989, the corresponding increase among women was 58%.6 There are however currently no published longitudinal data on weight change in the Australian population.
Efforts to manage this obesity epidemic are currently hindered by a lack of understanding of the aetiology of weight gain and obesity. The freely available nature of a diet high in energy and fat and increasingly sedentary lifestyles are thought to combine to increase the risk of obesity.1 The relative contributions of these behavioural factors are not well understood. The development of effective prevention and treatment strategies for obesity requires a better understanding of these, and other sociodemographic, behavioural and environmental, determinants of weight status and weight gain in populations.
Typically, studies investigating factors impacting on body weight have focused on the predictors of obesity.7 Very few studies have examined factors associated with the maintenance of a healthy weight. However, the identification and description of predictors of weight maintenance has important potential utility in deriving preventive interventions.8,9 While it has been suggested that weight maintenance over time is relatively uncommon,10,11 currently very little is known about what constitutes weight maintenance, or the strategies used for successful weight maintenance. The ability to maintain a steady weight is likely to be the product of the interaction of genetic predisposition and modifiable personal, behavioural and environmental factors.12 The identification of those modifiable characteristics of ‘weight maintainers’ (ie people who have successfully maintained a stable weight over time), particularly within the population groups identified as high risk, could assist in the development of strategies aimed at preventing weight gain, and is thus a promising avenue of research.
In one of the few studies to investigate predictors of weight maintenance, St Jeor et al8 reported that weight maintainers (defined as those having weight change of less than 5 lb over five annual measurements) comprised 19% of a sample of 385 healthy weight and obese men (n=205) and women (n=180) aged 20 y and above. Compared with non-maintainers, maintainers had lower BMI and body fat, and reported fewer dieting/weight loss attempts. No association was reported between maintainer status and exercise.8 However, the measure of exercise in this study (number of times participants had used exercise programs to lose weight) was problematic. Furthermore, the study assessed behaviours (eg exercise, diet) retrospectively. A subsequent study using the same sample9 defined maintainers as those whose Survey 1 weight was within 5 lb of their body weight 4 y later (50% men and 42% women). Results showed that maintainers were more likely to be in the healthy weight range at baseline and to have less annual weight fluctuation than either weight gainers or weight losers. Maintainers did not differ from the other two groups on measures of dietary fat intake or the number or duration of reported exercise sessions.9 However, the sample used in both studies was relatively small and non-representative, and there was no adjustment in the analyses for the potential contributions of non-modifiable risk factors for weight gain, such as socioeconomic status or parity.
Longitudinal data from the USA suggest that women are more likely than men to experience large weight gains or losses over a 10 y period, and women are twice as likely as men to experience major weight gain.11 Lower rates of participation in physical activity among women compared with men13 may contribute to these gender differences in risk of weight gain. Other factors which may contribute to women's increased risk of weight gain include their greater number of voluntary weight loss attempts or dieting, childbearing, smoking initiation and cessation, and alcohol consumption.11 Generally, however, data on weight change among women over time are scant.
Among women, the development of overweight and obesity is usually assumed to occur in mid-life. For example, in a large national cohort study of Australian women, the Australian Longitudinal Study on Women's Health (now referred to as the Women's Health Australia project, or WHA), more than 50% of the middle-aged women (40–45 y) were classified as overweight or obese, and only 7% were classified as ‘underweight’ at baseline in 1996.14 In contrast, in the younger cohort (18–23 y) almost 30% of the women were underweight, with 50% in the healthy weight range. These cross-sectional data suggest that there is likely to be considerable weight gain during young adulthood, and longitudinal data from the US confirm that young women are one of the most ‘at risk’ groups in terms of weight gain and obesity.11,15 Longitudinal follow-up of the young women participants in the WHA study provides the opportunity to explore factors associated with weight maintenance or change over time during young adulthood. This is particularly pertinent since young women are moving into their child-rearing years, a life-stage known to be associated with weight gain.16
The aim of this study was therefore to investigate the prevalence and predictors of weight maintenance over time in a large, representative sample of young Australian women. Specifically, the study aimed to quantify relationships between weight maintenance and behavioural factors including physical activity, sedentary behaviours, diet and selected eating behaviours, smoking, and alcohol consumption.
Data from 8726 women who are participants in the Australian Longitudinal Study on Women's Health (WHA) were used in this study. WHA involves three age cohorts of young, middle-aged and older women selected randomly from the Australian national health care database, which includes all women who are resident in Australia. Details of the recruitment methods and baseline surveys are described elsewhere.17 Since young women have been identified as a high-risk group for weight gain, the focus of this paper is on the young cohort (aged 18–23 y at baseline).
In 1996, 14 779 young women (41% of those invited to participate) completed a baseline survey (Survey 1), which assessed a broad range of women's health issues. Comparison with the 1996 national Census showed that the women were broadly representative of the female population in this age group, although tertiary-educated women were slightly over-represented.17 Of this group, 9657 (70% of the 13 826 Survey 1 respondents who had not subsequently withdrawn from the study, died or been lost to follow-up) completed a follow-up survey (Survey 2) 4 y later. The present analyses focus only on those women who completed both Survey 1 and 2. Comparison of these women with those who failed to respond to Survey 2 showed no consistent differences on area or state of residence, education, occupation, marital status, children living in the same house, or measures of self-rated physical or mental health.
Data from women who were pregnant at the time of either survey (n=694) were excluded from analyses. In addition, data from women who reported being diagnosed between Surveys 1 and 2 with a serious medical condition which might impact on their weight (HIV/AIDS or cancer; n=167), and women who reported requiring regular help with daily tasks due to a long-term illness or disability (n=96), were also excluded from analyses. Taking into account that some women reported two or more of these exclusion conditions, the total number of women whose data were excluded was 931, leaving 8726 women for whom data were analysed in this study.
Women were asked in both Surveys 1 and 2 to report their height and weight, from which BMI was calculated as weight (kg) divided by Survey 1 height (metres)2. BMI was categorized as ‘underweight’ (BMI<20 kg/m2), ‘healthy weight’ (BMI 20–25 kg/m2), ‘overweight’ (BMI> 25–30 kg/m2) and ‘obese’ (BMI>30 kg/m2), according to the Australian National Health and Medical Research Council classification system.18
A universal definition of weight maintenance has not been established. For the purposes of this study, women were classified as ‘maintainers’ if their BMI at Survey 2 was within 5% of their Survey 1 BMI. This criterion has been previously used to define stable weight in studies of correlates and health implications of weight change.10,19 Women who had a Survey 2 BMI more than 5% greater than their Survey 1 BMI were classified as ‘gainers’, and those with a Survey 2 BMI more than 5% below their Survey 1 BMI were categorized as ‘losers’.
Physical activity was assessed in Survey 1 with the items, ‘In a normal week, how many times do you engage in vigorous exercise lasting for 20 min or more? (exercise which makes you breathe harder or puff and pant, such as netball, squash, jogging, aerobics, vigorous swimming, etc)’ and, ‘In a normal week, how many times do you engage in less vigorous exercise lasting for 20 min or more? (exercise which does not make you breathe harder or puff and pant, like walking, gardening, swimming and lawn bowls)’. Responses for both questions were the same: never; once a week; two or three times a week; four, five or six times a week; once every day; more than once every day. A physical activity score was derived from responses to both exercise questions (PA score=(number of sessions of vigorous exercise * 5)+(number of sessions of less vigorous exercise * 3), and women were categorized as ‘none’ (<5), ‘low’ (5– <15), ‘moderate’ (15– <25) or ‘high’ (≥25) levels of physical activity.20
Sitting time was assessed in Survey 2 with the question ‘How many hours in total do you typically spend sitting down while doing things like visiting friends, driving, reading, watching television, or working at a desk or computer? On a usual week day; on a usual weekend day’ (open ended response). Responses to the weekday and weekend items were used to provide an estimate of total time spent sitting ((weekday hours * 5)+weekend hours * 2)), which was divided into tertiles for the present analyses. Women were categorized as ‘high sitting time’ (≥52 h per week sitting), ‘moderate sitting time’ (33– <52 h sitting) or low sitting time (<33 h sitting).
In Survey 1, dieting history was assessed with the questions, ‘Have you ever dieted to lose weight?’ and ‘How often have you gone on a diet (that is, limited how much you ate) in order to lose weight during the last year?’ (never; 1–4 times; 5–10 times; more than 10 times; I am always on a diet to lose weight). Respondents were then categorized as ‘never dieted’, ‘occasional dieters’ (1–4 times), or ‘frequent dieters’ (5 or more times or always).
Unhealthy weight loss behaviours such as binge eating, vomiting, laxative use, diuretics and fasting were assessed with the following questions: ‘Have there been times when you felt that you have eaten what other people would regard as an unusually large amount of food given the circumstances?’ (yes, in the past month; yes, more than one month ago; no) and ‘During these times of overeating, did you have a sense of having lost control over your eating, that is, feeling that you couldn't stop eating once you had started?’ (Yes/No). Respondents were categorized as either ‘current binge eating’ (reported binge eating, with a loss of control, in the past month) or ‘no binge eating’. Respondents were also asked, ‘Have you used any of the following to control your weight or shape?’ (vomited on purpose after eating; laxatives; diuretics; fasting (not eating food for at least a day). Respondents were categorized as ‘current restrictive disordered eating’ (DE; reported any of the four restrictive eating behaviours in the past month), or ‘no current restrictive DE’ (none of these behaviours in the past month).
Frequency of eating takeaway food (never, less than once a month, about once a month, about once a week, more than once a week, almost every day), was also assessed in Survey 1. Women were categorized as ‘never/rarely’ (never, less than once a month, or once a month); ‘occasionally’ (once a week) or ‘frequently’ (more than once a week) eating take away food.
Smoking and alcohol
In Survey 1 the women were asked, ‘Which of the following best describes your smoking status now?’ (I have never smoked; I used to smoke; I now smoke occasionally; I now smoke regularly). Respondents were classified as either ‘never smoked’, ‘ex-smokers’, or ‘current smokers’.
Alcohol consumption was assessed in Survey 1 with the questions ‘How often do you usually drink alcohol?’ (I never drink alcohol; I drink rarely; less than once week; on 1 or 2 days a week; on 3 or 4 days a week; on 5 or 6 days a week; every day); and ‘On a day when you drink alcohol, how many standard drinks do you usually have?’ (1 or 2 drinks per day; 3 or 4 drinks per day; 5–8 drinks per day; 9 or more drinks per day). Combinations of the frequency/quantity responses were used to categorize women as as ‘non-drinker or rarely drink’, ‘low intake’ (≤14 drinks/week), or ‘hazardous/ harmful’ drinkers (>14 drinks/week).21
In both surveys, women were asked how many times they had given birth to a child (never; once; twice; three times; four or more times; don't want to answer). Parity at Survey 1 was categorized as ‘no births’, ‘one birth’ or ‘two or more births’. Women were classified as ‘new mothers’ if they had given birth to a child between Survey 1 and 2.
Sociodemographic characteristics were derived from responses to the following Survey 1 items. Respondents were asked about their marital status; the age at which they left school (still at school; never attended school; 14 y or less; 15–16 y; 17–18 y; or 19 y or older); and whether they were currently attending an educational institution (no; yes, part-time student; yes, full-time student).
Main occupation was also assessed (manager or administrator; professional; para-professional; tradesperson; clerk; sales or personal service worker; machine operator or driver; manual worker; never had a paid job; or other). It was expected that a large proportion of women in this cohort would be studying, and hence the question on main occupation stated ‘If you are a student, circle the occupation you are studying for’.
Analyses were conducted using the SPSS version 10.0 statistical software package. Chi-square analyses and ANOVAs were used to investigate univariate differences in sociodemographic and behavioural factors between women who had maintained their weight and those whose weight had changed. A logistic regression model was then used to predict the likelihood of being a weight maintainer, compared with being a weight gainer, from behavioural variables. Only those variables found to differ between groups in univariate analyses were entered into the logistic regression model. A forced entry model was used and adjusted for those sociodemographic factors significantly associated with weight maintenance status in univariate analyses.
Sociodemographic and weight-related characteristics of the women were assessed at Survey 1, except for time sitting which was assessed at Survey 2. At Survey 1, the majority of the women were educated to at least 17–18 y, and 37% were still full-time students. Only 22% were currently married or living in de facto relationships, and 8% had at least one child. Only half the women were in the healthy weight category (47.0%), with approximately one-quarter categorized as ‘underweight’ (23.3%) or ‘overweight or obese’ (13.9 and 5.9%, respectively), after exclusion of 985 women (11%) who had height or weight missing in either survey and hence could not be classified. More than half the women were classified as ‘moderately’ or ‘highly’ active. At Survey 2, reported sitting time per week was quite high, with one-third of the women reporting spending more than 52 h per week sitting. More than 10% had given birth between Survey 1 and Survey 2.
At Survey 2, the proportion of women in each of the BMI categories had changed, with only 64% reporting a BMI≤25 (18.6% were underweight, and 45.4% had a BMI in the healthy range); 17.6% were overweight and 9.6% obese. The proportions of women categorized as ‘gainers’, ‘maintainers’ and ‘losers’ are shown in Table 1. Only 44% of the women were classified as maintainers, with 41% gainers and 15% losers. The distributions of sociodemographic factors by weight maintenance status are also shown in Table 1. Significant associations were found between weight maintenance status (ie gainer, maintainer or loser) and occupation, marital status, student status, parity and new motherhood. Weight maintainers were more likely to be in managerial or professional occupations, to have never married, to be currently studying, and to not be mothers (particularly new mothers).
Univariate comparisons of the Survey 1 behavioural and weight characteristics of weight maintainers, weight gainers and weight losers are presented in Table 2. In this Table, BMI is shown as both a continuous and categorical variable—both were significantly associated with weight maintenance status (P<0.01). Weight maintainers were more likely to have a BMI≤25 at baseline. With the exception of physical activity, all other behaviours were significantly associated with weight maintenance status in univariate analyses.
Those behaviours which were significantly associated with weight change status in univariate analyses were entered into a logistic regression analysis to predict weight maintenance. Since the primary focus of these analyses was the investigation of factors potentially protective against weight gain, weight losers were omitted from these analyses. The results are shown in Table 3. Unadjusted odds ratios show the likelihood of being a weight maintainer, as opposed to a weight gainer, for each of the variables entered into the regression model (see Table 3).
Initial analysis found that, compared with women in the healthy weight range, those who were underweight, overweight or obese were less likely to maintain weight. Similarly, compared with the ‘low sitting’ group, the women who reported moderate or high sitting time were less likely to be in the weight maintainers. Being a current smoker, having restrictive eating practices (P=0.06) and eating takeaway food more than rarely were also significantly associated with decreased likelihood of weight maintenance. Only low intake of alcohol was associated with increased likelihood of maintaining weight. After adjustment for occupation, marital status, student status, parity and new mothers (the sociodemographic variables shown in Table 2 to be associated with weight change), smoking (P=0.08), low alcohol intake and current restrictive eating practices (P=0.07) were no longer associated with weight maintenance.
Compared with women who were in the healthy weight range at Survey 1, those who were underweight, overweight or obese were 15, 28 and 31%, respectively, less likely to be weight maintainers. Similarly, compared with the women who spent least time sitting, those who reported moderate and high sitting times (eg 33– <52 h,>52 h) were 17–20% less likely to have maintained their weight, and women who reported eating takeaway food occasionally were 15% less likely to be weight maintainers than those who rarely or never ate takeaway food.
This study provides the first published longitudinal data on weight change in the Australian population, and the first analyses of national data on weight maintenance in a large population-based sample of young Australian women. Only 44% of the young women in this cohort maintained their weight within 5% of baseline values over the 4 y period when they were in their late teens and early twenties. Forty-one percent reported increased weight in this period and 15% reported weight loss. These proportions are similar to those reported by St Jeor and collegues,9 who found that 42% of adult women maintained their weight within a comparable range (5 lb, or about 3%) and 35% gained weight over a 4 y period.
Univariate analyses showed that women who were single, continued as students, were in (or studying for) more prestigious occupations, and who had no children (and particularly did not become new mothers in this period) were most likely to maintain their weight. The univariate analyses also suggested that weight maintainers have generally healthier profiles. They were more likely to have never smoked, to consume only low levels of alcohol, and to have reported no binge or restrictive eating practices. The ability to maintain weight may be a marker of other healthy behaviours, and may reflect a generally healthier lifestyle. This is consistent with previous findings that showed weight maintainers exhibited better biological health risk factors (eg serum cholesterol, systolic and diastolic blood pressure) than weight gainers.9
When these socio-demographic and health behaviours were taken into account, the factors associated most clearly with weight maintenance in the present study were initially healthy BMI, low sitting times and low consumption of takeaway food. Although St Jeor and collegues9 did not assess sitting time or takeaway food consumption, they also found those in the healthy weight range were more likely to maintain weight. While it is understandable that underweight women might choose to gain weight, the fact that those young women in the overweight and obese categories in our study were about 30% more likely to gain weight is of concern, as this may signal the onset of ‘creeping obesity’ in the early twenties. This finding is consistent with results of a recent population study using data from the Australian National Nutrition Survey in which weight gain was more commonly reported by those already overweight.22 Previously published cross-sectional data from this WHA cohort show significant associations between overweight and obesity and health problems such as hypertension, even at this young age.23 St Jeor and collegues9 also reported that weight gainers have more deleterious changes in health risk factors over time. These women hence face a doubly heightened health risk in that their initial BMI is above the range associated with good health, and this is compounded by the fact that they are also more likely to gain weight over time.
The finding that less frequent takeaway food consumption was associated with weight maintenance is consistent with findings of a recent US study of fast food use. French and colleagues24 reported that fast food use (defined as the number of meals per week eaten from fast food restaurants), which was particularly high among young adult women, was associated with higher body weight and increased weight gain over a 3 y period. That study also showed that increases in fast food use over time were linked with increased total energy intake and percentage fat intake, and with decreased physical activity.24 In light of evidence on the health outcomes associated with obesity, with high fat diets and with physical inactivity, our findings and those of previous studies have serious public health implications, and suggest that reducing takeaway/fast food consumption should be an important public health priority.
To our knowledge, this is the first time that estimates of sitting time have been used in a population-based cohort study. The women were asked to report time spent sitting during transport, work and leisure, and the data clearly show that women who report greater sitting time are less likely to maintain weight. It should be noted that sitting time was measured retrospectively at Survey 2, and hence no causality can be inferred. It is, however, unlikely that the higher sitting time of weight gainers is a consequence of gained weight rather than a potential contributing factor. Further studies should investigate the relationship between sitting and other sedentary behaviours and weight change and obesity in more detail. It is interesting, however, to note that physical activity (PA) was not associated with weight maintenance in this study. Despite the fact that the PA measure used here has shown good face validity in terms of its relationship with physical and mental health outcomes,20 it is possible that, because it relied only on reports of frequency (and not duration) of PA, this measure was not sufficiently sensitive to differentiate between the PA patterns of weight maintainers and gainers. Subsequent surveys with this cohort have used an improved measure of PA and future analyses will be able to examine this issue more closely once the data are available.
Several limitations of this study should be acknowledged. First, the surveys relied completely on self-reported data, which may be subject to recall or social desirability biases. In particular, body weight is frequently under-reported and height over-reported, and hence BMI derived from self-report surveys such as this is likely to under-estimate the true prevalence of overweight and obesity in the population.25 However this study investigated the prevalence of weight maintenance/gain rather than overweight/obesity per se, and it is reasonable to expect that any self-report bias of body weight would be relatively consistent over the study period, and hence would not impact substantially on the accuracy of categorizing weight maintainers/gainers. A second limitation was that data were analysed from two time points only, without considering interim changes or ‘weight cycling’ which may in fact be important in terms of longer term health outcomes.26,27,28 Third, the sitting time variable was assessed only at Survey 2. Finally, there was a relatively low response rate to Survey 2. This was not, however, considered to be problematic since Survey 2 responders were socio-demographically representative of baseline responders and of the general population. Comparison of the sample characteristics with a national sample of women in a comparable age group (19–24 y) showed that WHA responders were also similar to the general population on measures of BMI.29
The main strengths of the study were that, unlike previous studies in this area, it involved a large population sample, and included collection of data on a wide variety of potentially confounding demographic and health behaviour factors such as occupation, parity and smoking. The study was also innovative in its unique perspective on the public health problem of overweight, focusing on weight maintenance rather than weight gain.
The increasing prevalence of obesity highlights an urgent need for more innovative approaches to the prevention of weight gain.6,30 In the past we have tended to focus our efforts on weight reduction and the treatment of obesity, in middle-aged women who are already overweight or obese. The findings presented here, which show an increase in the proportion of young women with BMI>25 over this 4 y period, suggest that strategies for preventing overweight and obesity in women should begin earlier, because weight gain appears to be associated with key events in women's lives such as marriage and childbirth. The data presented here suggest that decreasing sitting time and encouraging less use of takeaway food might be effective strategies for encouraging weight maintenance at this life stage. These findings provide an important basis for future intervention efforts to prevent weight gain among high-risk groups in the population, through contributing to a better understanding of the determinants of weight maintenance.
World Health Organization. Obesity, preventing and managing the global epidemic Report of the WHO Consultation on Obesity WHO: Geneva 1997
Flegal KM . The obesity epidemic in children and adults: current evidence and research issues Med Sci Sports Exerc 1999 31: S509–S514.
Mokdad A, Serdula M, Dietz W, Bowman B, Marks J, Koplan J . The spread of the obesity epidemic in the United States, 1991–1998 JAMA 1999 282 (6): 1519–1522.
Bennett N, Dodd T, Flatley J, Freeth S, Bolling K . Health survey for England 1993 HMSO: London 1995
Australian Bureau of Statistics. National Nutrition Survey: User's Guide, 1995 Catalogue no. 4801.0 Australian Bureau of Statistics: Canberra 1998
Bennett SA, Magnus P . Trends in cardiovascular risk factors in Australia: results from the National Heart Foundation's Risk Factor Prevalence Study, 1980–1989 Med J Am 1994 161: 519–527.
Hill JO, Melanson EL . Overview of the determinants of overweight and obesity: current evidence and research issues Med Sci Sports Exerc 1999 31 (11): S515–S521.
St Jeor ST, Brunner RL, Harrington ME, Scott BJ, Cutter GR, Brownell KD, Dyer AR, Foreyt JP . Who are the weight maintainers?? Obes Res 1995 3: S249–S259.
St Jeor ST, Brunner RL, Harrington ME, Scott BJ, Daugherty SA, Cutter GR, Brownell KD, Dyer AR, Foreyt JP . A classification system to evaluate weight maintainers, gainers and losers J Am Diet Assoc 1997 97: 481–488.
French SA, Jeffery RW, Folsom AR, Williamson DF, Byers T . Weight variability in a population-based sample of older women: reliability and intercorrelation of measures Int J Obes Relat Metab Disord 1995 19: 22–29.
Williamson DF, Kahn HS, Remington PL, Anda RF . The 10-year incidence of overweight and major weight gain in US adults Arch Intern Med 1990 150: 665–672.
Pasman WJ, Saris WM, Westerterp-Plantanga MS . Predictors of weight maintenance Obes Res 1999 7: 43–50.
US Department of Health and Human Services. Physical activity and health: a report of the Surgeon General Centers for Disease Control: Atlanta, GA 1996
Brown W, Dobson A, Mishra G . What is a healthy weight range for middle aged women?? Int J Obes Relat Metab Disord 1998 22: 520–528.
Parham ES . The context of weight changes: factors associated with weight changes in adult women J Am Diet Assoc 1988 88: 1539–1544.
Williamson DF . Descriptive epidemiology of body weight and weight change in US adults Ann Intern Med 1993 119: 646–649.
Brown WJ, Bryson L, Byles JE, Dobson AJ, Lee C, Mishra G, Schofield M . Women's Health Australia: recruitment for a national longitudinal cohort study Women Health 1998 28: 23–40.
National Health, and Medical, Research Council . Acting on Australia's weight: a strategic plan for the prevention of overweight and obesity AGPS: Canberra 1997
Lewis CE, Smith DE, Wallace DD, Williams OD, Bild DE, Jacobs DR . Seven-year trends in body weight and associations with lifestyle and behavioral characteristics in black and white young adults: the CARDIA Study Am J Pub Health 1997 87: 635–642.
Brown WJ, Mishra G, Lee C, Bauman A . Leisure time physical activity in Australian women: relationships with well being and symptoms Res Q Exerc Sport 2000 71: 201–216.
Jonas HA, Dobson AJ, Brown WJ . Patterns of alcohol consumption in young Australian women: socio-demographic factors, health-related behaviours and physical health Aust NZ J Pub Health 2000 24: 185–191.
Jackson M, Ball K, Crawford D . Beliefs about the causes of weight change in the Australian population Int J Obes Relat Metab Disord 2001 25: 1512–1516.
Brown WJ, Mishra G, Kenardy J, Dobson AJ . Relationships between body mass index and well-being in young Australian women Int J Obesity Relat Metab Disord 2000 24: 1360–1368.
French SA, Harnack L, Jeffery RW . Fast food restaurant use among women in the Pound of Prevention study: dietary, behavioural and demographic correlates Int J Obes Relat Metab Disord 2000 24: 1353–1359.
Flood V, Webb K, Lazarus R, Pang G . Use of self-report to monitor overweight and obesity in populations: some issues for consideration Aust NZ J Pub Health 2000 24: 96–99.
Wing R . Weight cycling in humans: a review of the literature Ann Beh Med 1992 14: 113–119.
Kajioka T, Tsuzuku S, Shimokata H, Sato Y . Effects of intentional weight cycling on non-obese young women Metabolism 2002 51 (2): 149–154.
Olson MB, Kelsey SF, Bittner V, Reis SE, Reichek N, Handberg EM, Merz CN . Weight cycling and high-density lipoprotein cholesterol in women: evidence of an adverse effect: a report from the NHLBI-sponsored WISE study. Women's Ischemia Syndrome Evaluation Study Group J Am Coll Cardiol 2000 36 (5): 1565–1571.
Australian Bureau of Statistics. National Health Survey: Users Guide Catalogue no. 4363.0 Australian Bureau of Statistics: Canberra 1995
Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Kaplan JP . The continuing epidemic of obesity in the United States JAMA 2000 284 (13): 1605–1608.
The research on which this paper is based was conducted as part of the Australian Longitudinal Study on Women's Health (Women's Health Australia). We are grateful to the Australian Commonwealth Department of Health and Ageing for funding.
Kylie Ball is supported by a Public Health Postdoctoral Research Fellowship from the National Health and Medical Research Council, ID 136925. Dr David Crawford is supported by a Career Development Award funded jointly by the National Health and Medical Research Council and National Heart Foundation of Australia.
About this article
Examination of how food environment and psychological factors interact in their relationship with dietary behaviours: test of a cross-sectional model
International Journal of Behavioral Nutrition and Physical Activity (2019)
BMC Psychiatry (2018)
Socioeconomic Position and Reproduction: Findings from the Australian Longitudinal Study on Women’s Health
Maternal and Child Health Journal (2018)
Objectively-assessed physical activity and weight change in young adults: a randomized controlled trial
International Journal of Behavioral Nutrition and Physical Activity (2017)
An Evaluation of the Evidence Relating to Physical Inactivity, Sedentary Behavior, and Cancer Incidence and Mortality
Current Epidemiology Reports (2017)