AJCN North Carolina Research Campus
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.
Agricola
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.
American Journal of Clinical Nutrition, Vol. 70, No. 6, 965-975, December 1999
© 1999 American Society for Clinical Nutrition

Original Research Communications

Weight-loss attempts and risk of major weight gain: a prospective study in Finnish adults1,2,3

Maarit Korkeila, Aila Rissanen, Jaakko Kaprio, Thorkild IA Sørensen and Markku Koskenvuo

1 From the Department of Public Health, University of Helsinki; the Department of Psychiatry, Helsinki Central University Hospital; the Danish Epidemiology Science Centre, Institute of Preventive Medicine, Copenhagen University Hospital; the Department of Public Health, University of Turku, Finland; and the Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki.

2 Supported by The Academy of Finland (grants 38332 and 42044) and the Doctoral Programs in Public Health, University of Helsinki.

3 Address reprint requests to J Kaprio, Department of Public Health, PO Box 41, 00014 University of Helsinki, Helsinki, Finland. E-mail: jaakko.kaprio{at}helsinki.fi.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Background: The effects of weight-loss attempts on long-term weight gain remain unclear.

Objective: The objective was to study prospectively how attempts to lose weight relate to future risk of major weight gain (>10 kg) and whether familial factors affect this relation.

Design: Participants in the Finnish Twin Cohort (3536 men and 4193 women aged 18–54 y at baseline) were followed up for 6–15 y. The role of familial factors was studied in 1705 twin pairs in this cohort who were discordant for weight-loss attempts at baseline. Baseline (1975) and follow-up (1981 and 1990) data—including weight, weight-loss attempts (dieting), and selected confounders—were obtained via mailed questionnaires.

Results: Average weight gain was at most weakly associated with weight-loss attempts. The risk of major weight gain for subjects attempting to lose weight at baseline was greatest among initially young (18–29 y) men (over 6 and 15 y, respectively—odds ratios: 2.01 and 1.74; 95% CI: 1.13, 3.57 and 1.11, 2.75) and middle-aged (30–54 y) women (over 6 and 15 y, respectively—2.43 and 1.52; 1.33, 4.42 and 1.06, 2.22) and persisted after potential confounders were controlled for. These risks decreased and became nonsignificant in the pairwise twin analysis, suggesting that the relation between dieting and subsequent major weight gain may also have a familial component.

Conclusions: Weight-loss attempts may be associated with subsequent major weight gain, even when several potential confounders are controlled for. Genetic and familial factors may contribute to this association.

Key Words: Weight-loss attempts • weight gain • Finland • obesity • Finnish Twin Cohort • humans


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Weight gain with age is well documented in several population studies (1–7). There has been a global increase in the prevalence of obesity over recent decades and it is a growing public health problem in affluent societies (2, 8–13). Attempts to lose weight are common; {approx}30% of US adults reported recent weight-loss attempts (14, 15). Because any achieved weight loss is seldom maintained for prolonged periods (16–20), dieting becomes an integral part of the lifestyle of some individuals (21). Weight gain has been shown to be associated with an increased risk of diabetes (22) and hypertension (23).

Obesity and relative body weight, as gauged by body mass index (BMI; in kg/m2), are partly heritable (24–28), although weight change in adulthood may be less dependent on genetic factors (29, 30) and more on gene-environment interactions (31, 32). Smoking cessation, physical inactivity, low education and socioeconomic status, and parity have been identified as risk factors for major weight gain in both cross-sectional and prospective studies (33–42). However, most prospective studies have relied on a single follow-up measurement of weight change, leaving its stability unexplored. Additionally, few studies have examined the potential confounding effect of several predictors simultaneously.

In view of the commonness of weight-loss attempts, surprisingly little is known about the long-term effects of such attempts on weight. In the prospective study by French et al (40), a dieting history at baseline predicted weight gain in 3552 American men and women over a 2-y follow-up period. In another recent prospective study of 19478 healthy American male health professionals aged 40–75 y, frequent weight-loss attempts (dieting) were related to weight gain (43).

The aim of this study was to investigate in an adult population the long-term effects of dieting on weight development over follow-up periods of 6 and 15 y while several pertinent putative confounding factors were controlled for. Data on twin pairs discordant for weight-loss attempts were analyzed to determine whether the association between weight-loss attempts and major weight gain was independent of familial factors.


   SUBJECTS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
The Finnish Twin Cohort
The present study is based on data from the Finnish Twin Cohort compiled in 1974. The cohort included 13888 monozygotic or same-sexed dizygotic twin pairs born before 1958 (44, 45). Individuals were mailed nearly similar questionnaires in 1975 and 1981 and the response rates were 89% and 84%, respectively. An additional questionnaire was mailed in 1990 to twin pairs born between 1930 and 1957 who were both alive and resident in Finland in 1987 (n = 8938 twin pairs and 17876 individuals); the response rate was 77.3% (46).

The study population
In the present study, one member of each twin pair from the initial sample of 15458 twins was randomly selected for the analyses of weight changes over 6 y (Figure 1Go). These 7729 subjects (twin A sample: 3536 unrelated men and 4193 unrelated women at baseline) were aged 18–54 y in 1975 and provided complete weight data on the 1975 and 1981 surveys and complete height data on at least one of these surveys. For the 15-y follow-up analyses, we also included randomly selected subjects who had responded to the 1990 questionnaire (2152 men and 2721 women). Members of each twin pair who were not included in the twin A sample formed the twin B sample, which was used for replication analyses. All analyses were carried out for 2 age groups (18–29 and 30–54 y at baseline) separately because their patterns of weight gain differed significantly (26). When divided by age and sex, 4 approximately equal-sized groups were formed (Table 1Go). Subjects aged 18–29 y at baseline are referred to as young and those aged 30–54 y as older. Subjects were aged 24–60 y at the first follow-up in 1981. At the second follow-up in 1990, subjects were aged 33–60 y, which was the maximum inclusion age for the 1990 survey.



View larger version (13K):
[in this window]
[in a new window]
  		FIGURE 1. Description of the sample of twin individuals and twin pairs in The Finnish Twin Cohort Study.

 

View this table:
[in this window]
[in a new window]
  		TABLE 1. Reported weight-loss attempts (dieting), BMI at baseline, and weight changes over follow-up periods of 6 and 15 y in the Finnish Twin Cohort1
 
Women who were pregnant at the time of the baseline questionnaire or during the time between the baseline and follow-up questionnaires were excluded from some analyses (n = 1195 for the 6-y follow-up and n = 1191 for the 15-y follow-up. The exclusion of women with a change in parity between the 1975 and 1981 questionnaires was based on data of deliveries from the national hospital discharge registry with use of time-related criteria given in detail elsewhere (26). Women with a change in parity between the 1981 and 1990 surveys were identified by a question on family structure in the 1990 questionnaire or from hospital data. After these exclusions, the remaining sample consisted of 2998 unrelated women over the shorter (6 y) and 1530 unrelated women over the longer (15 y) follow-up.

Determination of weight change
In the questionnaires, weight was given in kilograms and height in centimeters, which, when necessary, were rounded to the nearest whole number. BMI was used as the measure of excess body weight, and weight change was the difference between the self-reported weights in 1975 and 1981 or 1990. A gain of >10 kg was regarded as major weight gain and was the outcome measure in the analyses.

The validity of self-reported weights in 1990 was ascertained in a random sample of twins living in greater Helsinki in 1994–1995 (47); clinical measures for 100 men aged 52.5 ± 6.5 y (x ± SD) and 125 women aged 50.4 ± 7.9 y who reported their heights and weights in the 1990 questionnaire were directly measured as part of a clinical examination. BMI values computed from the self-reported and measured values were in good agreement, with a correlation coefficient of 0.89 (Figure 2Go). The regression equation was BMImeasured = 1.8 + 0.98BMIself-reported (1990) (SE: 0.03).



View larger version (14K):
[in this window]
[in a new window]
  		FIGURE 2. Self-reported body mass index (BMI) in 1990 compared with measured BMI in 1994–1995 in 225 twins in The Finnish Twin Cohort.

 
Lifestyle and socioeconomic factors
The following baseline (1975) variables, which were used to assess lifestyle and socioeconomic status, were studied as predictors of major weight gain.

Weight-loss attempts
Subjects were asked whether they were currently trying to lose excess weight. Those who reported that they were trying to lose weight are referred to as dieters; 64% of the men and 80% of the women reported restricting food intake voluntarily or changing their diet (with or without exercise) as their chosen method of weight loss (44). The follow-up questionnaires in 1981 and 1990 did not include questions on current weight-loss attempts.

Smoking
Respondents were asked for a detailed smoking history at baseline. They were classified as nonsmokers or as occasional, former, or current cigarette smokers (48). Current smoking quantity was based on the number of cigarettes smoked daily and categorized as 1–14 or >=15 cigarettes/d.

Alcohol use
The frequency (days per month), quantity, and type of alcohol used at baseline were rated by the responses to 6 questions about the current use of alcoholic beverages (beer, wine, and spirits). Quantities were then converted into grams of ethanol to give the monthly consumption of pure alcohol (49). The subjects were classified as abstainers (no use), occasional users (1–150 g/mo), moderate users (151–399 g/mo), and heavy users (>=400 g/mo).

Education level
On the basis of schools and higher-level institutions attended, the respondents' education level was classified as primary (<=6 y), secondary (7–11 y), or tertiary (>=12 y).

Social class and marital status
The subjects were grouped by their occupation according to the 1970 classification of the Finnish Central Statistical Office. The 5 social-class categories were upper- and lower-grade professionals, skilled and unskilled workers, and farmers. Students, housewives, and unemployed and pensioned persons of unknown occupation were not classified by social class. The respondents were classified as unmarried, married, divorced, or widowed. Those remarried or living together were considered married. Detailed descriptions of these variables are presented elsewhere (45, 50).

Energy expenditure at leisure and work
Physical activity was estimated separately for work and leisure time. Intensity of physical activity was expressed as MET values (metabolic rate divided by resting metabolic rate in J/mo) (51). Calculations of leisure time activity were based on 7 questions about 3 components of physical activity: duration, intensity, and monthly frequency. Physical activity at work was determined as the product of work strenuousness (intensity) x the number of working hours/d (8) x the number of working days/mo (51). The obtained leisure and work indexes were divided into tertiles by sex for further analyses.

Statistical methods
The statistical analyses were performed by using BMDP software (version 7.0; BMDP Statistical Software, Inc, Los Angeles). Analysis of covariance was used to compare mean BMI at baseline, adjusted for initial age in different groups. Odds ratios (ORs) for gaining >10 kg were assessed by logistic regression analysis (52). Weight changes were adjusted for age and BMI at baseline by entering these terms as continuous variables in models, and the ORs over the 2 follow-up periods (6 and 15 y) were computed. Significance was estimated by the 95% CI. In addition to baseline age and BMI, potential confounders (smoking, alcohol use, educational level, social class, marital status, and energy expenditure at leisure and at work) were included in the logistic regression equation to test the independence of the effects of attempts to lose weight over 6 and 15 y. Analyses of the future risk of gaining >10 kg after weight-loss attempts over 6 and 15 y were also replicated in a second sample (twin B) (Figure 1Go).

A discordant pair analysis based on data of both members of all pairs was carried out to assess the effect of familial factors on the relation of weight-loss attempts to weight gain (Figure 3Go). Of all pairs initially in the study, either both or one twin in 2258 pairs reported weight-loss attempts.



View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 3. Design for analysis of matched-pair data nested within a cohort study. B, number of twin pairs discordant for weight-loss attempts and weight gain in which the dieting twin gained >10 kg; C, number of twin pairs discordant for dieting and weight gain in which the nondieting twin gained >10 kg. Relative risk of disease associated with exposure = number of twin pairs in cell B/number of twin pairs in cell C.

 
Tetrachoric correlations were computed to address the question of whether the report of weight-loss attempts was more concordant among monozygotic than among dizygotic twins. Tetrachoric correlations are one of the most powerful measures of twin concordance, provided the variable under study is unidimensional with a normal distribution (53). A tetrachoric correlation is the correlation that best approximates the cell probabilities from a 2 x 2 contingency table (weight-loss attempt/no weight-loss attempt in twin 1 compared with the same in twin 2). The pairs discordant for weight-loss attempts were the basis for the analyses in which the distribution of pairwise weight changes adjusted for initial age and BMI was examined in relation to weight-loss attempts. ORs for major weight gain by weight-loss attempt status in the discordant pairs were computed by conditional logistic regression with the PECAN program (EPICURE; Hirosoft International Corporation, Seattle). These ORs give the risk for major weight gain when the weight-loss attempt status of the subject is compared with that of their cotwin and his or her weight gain data. Thus, familial effects are controlled for by the inclusion of partial (for dizygotic pairs) or complete (for monozygotic pairs) adjustment for genetic background similarity in addition to shared (nongenetic) family effects.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Reported weight-loss attempts and initial weight
At baseline, 7.6% of men and 20.8% of women aged 18–29 y and 18.8% of men and 26.8% of women aged 30–54 y reported weight-loss attempts. These subjects had significantly higher initial BMIs at baseline than did others in the same sex and age groups. Mean weight changes over 6 y did not differ significantly by reported weight-loss attempts at baseline (Table 1Go). Mean weight change over 15 y, however, was higher among dieting than among nondieting subjects in 3 of 4 groups; however, these differences were significant only in older women and were nearly significant in younger men (P = 0.07). Mean initial BMIs and weight changes for women with no change in parity were not significantly different from those of all women.

The frequency of dieting and mean BMIs at baseline were assessed in different categories of potential confounders. Dieting was more frequent in former smokers than in nonsmokers and current smokers aged 30–54 y as well as in young women who were heavy smokers than in light smokers (35.3% in subjects smoking >=15 cigarettes/d; P = 0.016). Subjects with the highest alcohol consumption reported dieting more often than did others [10.5% (P = 0.009) in young men and 37.2% (P = 0.008) in older women]. Dieting was clearly more common in men of the upper social classes than in those of lower social classes and there were only minor differences by educational or marital status. High energy expenditure during leisure time and heavy physical work were both associated with a high likelihood of dieting, especially among young subjects (data not shown).

Young male current smokers, particularly the heavy smokers (>=15 cigarettes/d), and older male former smokers had the highest BMIs. Men who drank heavily and older nondrinking women had the highest baseline BMIs. BMIs and education level were inversely related (P < 0.01) in all groups and upper-grade professionals had lower BMIs than did subjects in other social classes. Married women had higher BMIs than did single women, whereas no significant difference by marital status was seen in men. Low leisure time physical activity was related to the highest baseline BMIs in older women, whereas work physical activity was positively related to baseline BMI in all groups (Table 2Go).


View this table:
[in this window]
[in a new window]
  		TABLE 2. BMI by levels of potential confounders (smoking status and dose, alcohol use, education level, social class, marital status, and leisure and working time energy expenditure) in the Finnish Twin Cohort at baseline (1975)
 
Weight-loss attempts and the risk of major weight gain
The ORs for major weight gain according to dieting status (yes or no) at 2 strata of baseline BMI (<25 and >=25), adjusted for baseline age and BMI, and potential confounders are presented for men in Table 3Go and for women in Table 4Go. The Breslow-Day test (54) for heterogeneity of ORs over 4 BMI strata showed no significant differences in any of the sex and age groups nor between subjects with a height above the mean and those with a height below the mean at baseline over both follow-up periods.


View this table:
[in this window]
[in a new window]
  		TABLE 3. Weight-loss attempts at baseline (1975) and subsequent risk of major weight gain (>10 kg) in men by BMI at baseline over follow-up periods of 6 (1975–1981) and 15 (1975–1990) y in the Finnish Twin Cohort1
 

View this table:
[in this window]
[in a new window]
  		TABLE 4. Weight-loss attempt at baseline (1975) and subsequent risk of major weight gain (>10 kg) in all women and in women with no change in parity by BMI at baseline over follow-up periods of 6 (1975–1981) and 15 (1975–1990) y in the Finnish Twin Cohort 19751
 
Overall, there was a tendency for an increased risk of major weight gain in dieting men and women, although not all results were significant. Dieting at baseline was a consistent predictor of major weight gain in young men at both follow-up periods (over 6 and 15 y, respectively—ORs: 2.01 and 1.74; 95% CI: 1.13, 3.57 and 1.11, 2.75), which persisted even after the adjustments for age, BMI, smoking, alcohol use, education level, and social class at baseline. Over 6 y, the risk of major weight gain increased in initially normal-weight (BMI < 25) dieting men (OR: 2.81; 95% CI: 1.47, 5.38), although no increase in risk was observed in young initially overweight (BMI > 25) dieting men (OR: 0.80; 95% CI: 0.25, 2.56). Over 15 y, the risk was significantly elevated in older men (OR: 2.10; 95% CI: 1.27, 3.48).

Among young women, the risk of major weight gain tended to be higher in dieters than in nondieters over 6 y in both initial-weight groups, although these differences were not significant. Older women who were dieting at baseline had a significantly higher risk of major weight gain over both follow-up periods (over 6 and 15 y, respectively—ORs: 2.43 and 1.53; 95% CI: 1.33, 4.42 and 1.06, 2.22). The ORs decreased only slightly when potential confounders were controlled for, although one previously significant result became nonsignificant. Exclusion of women with a change in parity during the follow-up periods or adjustment for baseline age only did not change the point estimates substantially.

The ORs for 6-y weight changes were reanalyzed by using the same sample for 6- and 15-y weight changes. Compared with the larger sample of all subjects available at 6 y, the results differed significantly in one group only, older women. Among these older women, the OR decreased from 2.43 to 1.89. In the replicating analyses, where the other randomly selected twin was studied, the results for women were fully replicated, whereas the results among the men showed mostly weaker, though consistent, relations (data not shown).

Pairwise analyses
There were no significant differences among individuals in the prevalence of dieting (weight-loss attempts) by zygosity: 14.1% of monozygotic and 12.3% of dizygotic men and 26.2% of monozygotic and 24.6% of dizygotic women reported weight-loss attempts at baseline. Pairs concordant (n = 553) and discordant (n = 1705) for dieting in the entire sample by sex, age, and zygosity are shown in Table 5Go. Most pairs were discordant for dieting at baseline. The nondieting member had a consistently lower BMI in all sex, zygosity, and age groups at baseline (P < 0.0001). The difference in BMI was maintained over follow-up: 23.8 (95% CI: 23.7, 24.0) in nondieting and 25.2 (95% CI: 25.06, 25.41) in dieting pair members in 1981. This difference was significant in both monozygotic and dizygotic pairs.


View this table:
[in this window]
[in a new window]
  		TABLE 5. Weight-loss attempts in Finnish twin pairs: pairwise status and tetrachoric correlations and 95% CIs by sex and zygosity
 
The tetrachoric correlations (r) for weight-loss attempts were r = 0.60 (95% CI: 0.51, 0.70) for monozygotic and r = 0.36 (95% CI: 0.27, 0.45) for dizygotic men and r = 0.59 (95% CI: 0.51, 0.66) for monozygotic and r = 0.34 (95% CI: 0.28, 0.41) for dizygotic women (Table 5Go). This pattern of correlations suggests that dieting behavior aggregates within pairs. Furthermore, because the monozygotic correlations were almost twice as great as the dizygotic correlations, genetic factors appear to have accounted for most of this aggregation.

Altogether, 59 twin (13 men and 46 women) pairs discordant for dieting at baseline were also discordant for major weight gain over the 6-y follow-up period; the corresponding number of pairs over the 15-y follow-up was 211 (68 men and 143 women). The mean pairwise differences in weight change did not differ significantly by zygosity. The matched-pair analyses in major weight gain and dieting discordant pairs of twins were carried out as depicted in Figure 3Go. For men, the ORs for major weight gain over 6 y were 3.33 (95% CI: 0.86, 18.9) for monozygotic and 1.19 (95% CI: 0.64, 2.24) for dizygotic pairs. Compared with men, the ORs for major weight gain in all women were lower (monozygotic and dizygotic women, respectively—ORs: 0.93 and 0.56; 95% CI: 0.40, 2.13 and 0.32, 0.94), indeed significantly so for dizygotic women. Exclusion of women with a change in parity during the subsequent 6 y slightly decreased the ORs (ORs: 0.63 and 0.38 in monozygotic and dizygotic women, respectively). The ORs for major weight gain over 15 y were all nonsignificant compared with the shorter follow-up (monozygotic and dizygotic men, respectively) in both men (ORs: 0.64 and 0.67; 95% CI: 0.21, 1.80 and 0.36, 1.21) and women (ORs: 0.73 and 0.69; 95% CI: 0.36, 1.45 and 0.46, 1.04).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The aim of this prospective study was to determine the effects of weight-loss attempts on major weight gain over 6 and 15 y. The results showed that reported weight-loss attempts were related to major weight gain in adult Finns after several important confounders were controlled for. The risk of major weight gain increased, although not always significantly so, in nearly all groups of initially normal-weight subjects who were attempting to lose weight at baseline. In contrast, the history of weight-loss attempts was not consistently associated with increased risk of major weight gain among initially overweight subjects, except for older women, in whom the risk for major weight gain increased markedly. Additionally, average weight gain was at most only weakly associated with weight-loss attempts, which most probably reflects the fact that some subjects can successfully lose weight. Our findings agree with prospective information showing that dieting behavior may be related to the risk of weight gain (40, 43). Further evidence for the link between dieting and weight gain was provided by an observational study in 4278 young US adults, in whom self-reported weight-loss attempts were associated with weight gain over 2 y (55).

The particular strength of the present study was that there was long-term follow-up information on weight after 2 distinct intervals (6 and 15 y). There was also extensive baseline information on potential confounders. Unfortunately, information about weight-loss attempts was obtained only at baseline, and we only had indirect means of estimating the reliability of this item. The disparity between an observed and full correlation (of 1.0) in monozygotic twins is due to differences in individual unshared experiences between them as well as to measurement error, the magnitude of which was unknown. Thus, the true reliability of the question about weight-loss attempts may well have exceeded the observed correlation (0.60), which is likely to represent the lowest limit of reliability of the question on weight-loss attempts. However, the validity of the question, ie, whether the reporting of weight-loss attempts actually reflects actual behavior, is unknown.

The cutoff point for major weight gain in this study (10 kg) was, of course, arbitrary. We chose it mainly because it is large enough to exclude random variation. It is possible that such a large value may mask some predictors with weaker effects. The risk of major weight gain by weight-loss-attempt status did not depend on height, but showed some variation with initial BMI. Self-reported data were used in the present study, which have been suggested to provide sufficient accuracy for population-based studies (56–59). The level of discrepancy in self-reported weights compared with measured weights ascertained in a subsample of twins from this cohort showed that the internal validity was good. The ORs for weight gain were also computed for twin individuals excluded from the preliminary analyses (twin B sample). The trends for risk were similar in both samples (twin A and twin B samples), suggesting little effect of random variation.

Several explanations for the connection between weight-loss attempts and risk of gaining weight can be suggested. Subjects whose body weight is increasing may try to counter this by dieting, with some short-term success, but they frequently fail to control body weight in the long run. This typical pattern is congruent with our finding of an association between weight-loss attempts and subsequent weight gain among those with normal weight at baseline. The poor success in weight maintenance after dieting predisposes individuals to the vicious cycle of frequent dieting attempts and weight regain (18–20). The relation between weight cycling and subsequent weight gain is well described in the literature (60, 61). Part of the weight gain occurring in young adults may be regarded as physiologic, and is likely to occur independently of attempts to lose weight. Smoking, alcohol use, education level, and marital status were regarded as potential confounders in this study. Weight-loss attempts were more frequent among heavy consumers of tobacco and alcohol and among women with high socioeconomic status, which agrees with the findings of previous studies (60, 62, 63).

Predisposition to gain weight, which is partly genetically determined (24, 26, 29), might be greater among dieting than among nondieting subjects and is possibly strong enough to override any weight-loss attempts. The tetrachoric correlations of twin pairs suggest that weight-loss attempts aggregate in families and that part of this aggregation may be inherited; this familial aggregation may reflect the more similar BMIs of monozygotic than of dizygotic pairs. However, a formal bivariate twin analysis is needed to test this hypothesis. Weight-loss attempts were related to subsequent major weight gain in individuals, whereas this relation was not found after adjustment for familial factors based on the ORs of twin pairs discordant for both weight gain and dieting. It is possible that there are genetic pleiotropic effects influencing both weight gain and dieting behavior.

We conclude that weight-loss attempts appear to increase the risk of long-term major weight gain in adults. The results indicate, in accordance with previous literature, that persons whose weight is currently reduced through dieting are at risk of regaining weight. These findings are based on undefined weight-loss attempts and do not rule out the potential successes and benefits of structured weight-loss programs aimed at permanent changes in health behavior. The twin-pair analyses suggest that much of the observed relation between weight-loss attempts and major weight gain can be attributed to a familial predisposition to gain weight, which ultimately overwhelms even ambitious weight-loss attempts. However, any such genetic effects do not signify that attempts to modify the environment in a healthier direction are likely to fail.


   ACKNOWLEDGMENTS
 
We thank Mikko Lehtovirta for contributing the results on the reliability of BMI measurements, Kauko Heikkilä for managing the database, and Richard Burton for revising the English.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 SUBJECTS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Borkan GA, Sparow D, Wisniewski C, Vokonas PS. Body weight and coronary disease risk: patterns of risk factor associated with long-term weight change. Am J Epidemiol 1986;124:410–9.[Abstract/Free Full Text]
  2. Rissanen A, Heliövaara M, Aromaa A. Overweight and anthropometric changes in adulthood: a prospective study of 17,000 Finns. Int J Obes 1988;12:391–401.[Medline]
  3. Colditz GA, Willett WC, Stampfer MJ, London SJ, Segal MR, Speizer FE. Patterns of weight change and their relation to diet in a cohort of healthy women. Am J Clin Nutr 1990;51:1100–5.[Abstract/Free Full Text]
  4. Heitmann BL. Body fat in the adult Danish population aged 35–65 years. An epidemiological study. Int J Obes 1991;15:535–45.[Medline]
  5. Korkeila M, Kaprio J, Rissanen A, Koskenvuo M. Weight changes among 13 097 adult Finns over six years. In: Ailhaud G, Guy-Grand B, Lafontan M, Ricquier D, eds. Obesity in Europe 91. London: John Libbey, 1991:129–33.
  6. 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;150:665–72.[Abstract]
  7. Grinker JA, Tucker K, Vokonas PS, Rush D. Overweight and leanness in adulthood: prospective study of male participants in the Normative Ageing Study. Int J Obes Relat Metab Disord 1996;20:561–9.[Medline]
  8. Kuczmarski RJ, Flegal KM, Campbell SM, Johnson CL. Increasing prevalence of overweight among US adults—The National Health and Nutrition Examination Surveys, 1960 to 1991. JAMA 1994; 272:205–11.[Abstract]
  9. Shah M, Hannan PJ, Jeffery RW. Secular trend in body mass index in the adult population of three communities from the upper mid-western part of the USA: the Minnesota Heart Health Program. Int J Obes 1991;15:499–503.[Medline]
  10. Kuskowska-Wolk A, Bergström R. Trends in body mass index and prevalence of obesity in Swedish women 1980–89. J Epidemiol Community Health 1993;47:195–9.[Abstract]
  11. Pietinen P, Vartiainen E, Männistö S. Trends in body mass index and obesity among adults in Finland from 1972 to 1992. Int J Obes Relat Metab Disord 1996;20:114–20.[Medline]
  12. Seidell JC. Obesity in Europe: scaling an epidemic. Int J Obes Relat Metab Disord 1995;19:S1–4.
  13. Yanai M, Kon A, Kumasaka K, Kawano K. Body mass index variations by age and sex, and prevalence of overweight in Japanese adults. Int J Obes Relat Metab Disord 1997;21:484–8.[Medline]
  14. Williamson DF, Serdula MK, Anda RF, Levy A, Byers T. Weight loss attempts in adults: goals, duration, and rate of weight loss. Am J Public Health 1992;82:1251–7.[Abstract/Free Full Text]
  15. Serdula MK, Williamson DF, Anda RF, Levy A, Heaton A, Byers T. Weight control practices in adults: results of a multistate telephone survey. Am J Public Health 1994;84:1821–4.[Abstract/Free Full Text]
  16. Stalonas PM, Perri MG, Kerzner AB. Do behavioral treatments of obesity last? A five-year follow-up investigation. Addict Behav 1984;9:175–83.[Medline]
  17. Jordan HA, Canavan AJ, Steer RA. Patterns of weight change: the interval 6 to 10 years after initial weight loss in a cognitive-behavioral treatment program. Psychol Rep 1985;57:195–203.[Medline]
  18. Kramer FM, Jeffery RW, Forster JL, Snell MK. Long-term follow-up of behavioral treatment for obesity: patterns of weight regain among men and women. Int J Obes 1989;13:123–36.[Medline]
  19. Fitzwater SL, Weinsier RL, Woolridge NH, Birch R, Liu C, Bartolucci AA. Evaluation of long-term weight changes after a multidisciplinary weight control program. J Am Diet Assoc 1991;91:421–6, 29.[Medline]
  20. Wadden TA, Frey DL. A multicenter evaluation of a proprietary weight loss program for the treatment of marked obesity: a five-year follow-up. Int J Eat Disord 1997;22:203–12.[Medline]
  21. Haus G, Hoerr SL, Mavis B, Robison J. Key modifiable factors in weight maintenance: fat intake, exercise, and weight cycling. J Am Diet Assoc 1994;94:409–13.[Medline]
  22. Holbrook TL, Barrett-Connor E, Wingard DL. The association of lifetime weight and weight control patterns with diabetes among men and women in an adult community. Int J Obes 1989;13:723–9.[Medline]
  23. Yong LC, Kuller LH, Rutan G, Bunker C. Longitudinal study of blood pressure: changes and determinants from adolescence to middle age. The Dormont High School follow-up study, 1957–1963 to 1989–1990. Am J Epidemiol 1993;138:973–83.[Abstract/Free Full Text]
  24. Stunkard AJ, Foch TT, Hrubec Z. A twin study on human obesity. JAMA 1986;256:51–4.[Abstract]
  25. Sörensen TIA, Price AR, Stunkard AJ, Schulsinger F. Genetics of obesity in adult adoptees and their biological siblings. BMJ 1989;298:87–90.
  26. Korkeila M, Kaprio J, Rissanen A, Koskenvuo M. Effects of gender and age on the heritability of body mass index. Int J Obes 1991; 15:647–54.[Medline]
  27. Harris JR, Tambs K, Magnus P. Sex-specific effects for body mass index in the Norwegian twin panel. Genet Epidemiol 1995; 12:251–65.[Medline]
  28. Herskind AM, McGue M, Sörensen TIA, Harvard B. Sex and age-specific assessment of genetic and environmental influences on body mass index in twins. Int J Obes Relat Metab Disord 1996;20:106–13.[Medline]
  29. Fabsitz RR, Carmelli D, Hewitt JK. Evidence for independent genetic influences on obesity in middle age. Int J Obes Relat Metab Disord 1992;16:657–66.[Medline]
  30. Korkeila M, Kaprio J, Rissanen A, Koskenvuo M. Consistency and change of body mass index and weight. A study on 5967 adult Finnish twin pairs. Int J Obes Relat Metab Disord 1995;19:310–7.[Medline]
  31. Heitmann BL, Lissner L, Sörensen TIA, Bengtsson C. Dietary fat intake and weight gain in women genetically predisposed for obesity. Am J Clin Nutr 1995;61:1213–7.[Abstract/Free Full Text]
  32. Heitmann BL, Kaprio J, Harris JB, Rissanen A, Korkeila M, Koskenvuo M. Are genetic determinants of weight gain modified by leisure-time physical activity? A prospective study of Finnish twins. Am J Clin Nutr 1997;66:672–8.[Abstract/Free Full Text]
  33. Kahn HS, Williamson DF. The contributions of income, education and changing marital status to weight change among US men. Int J Obes 1990;14:1057–68.[Medline]
  34. Kant AK, Schatzkin A, Graubard BI, Ballard-Barbash R. Frequency of eating occasions and weight change in the NHANES I epidemiologic follow-up study. Int J Obes Relat Metab Disord 1995;19:468–74.[Medline]
  35. Ching PLYH, Willett WC, Rimm EB, Colditz GA, Gortmaker SL, Stampfer MJ. Activity level and risk of overweight in male health professionals. Am J Public Health 1996;86:25–30.[Abstract/Free Full Text]
  36. Haapanen N, Miilunpalo S, Pasanen M, Oja P, Vuori I. Association between leisure time physical activity and 10-year body mass change among working-aged men and women. Int J Obes Relat Metab Disord 1997;21:288–96.[Medline]
  37. Liu S, Serdula MK, Williamson DF, Mokdada AH, Byers T. A prospective study of alcohol intake and change in body weight among US adults. Am J Epidemiol 1994;140:912–20.[Abstract/Free Full Text]
  38. Klesges RC, Klesges LM, Haddock CK, Eck LH. A longitudinal analysis of the impact of dietary intake and physical activity on weight change in adults. Am J Clin Nutr 1992;55:818–22.[Abstract/Free Full Text]
  39. Rissanen A, Heliövaara M, Knekt P, Reunanen A, Aromaa A. Determinants of weight gain and overweight in adult Finns. Eur J Clin Nutr 1991;45:419–30.[Medline]
  40. French SA, Jeffery RW, Forster JL, McGovern PG, Kelder SH, Baxter JE. Predictors of weight change over two years among a population of working adults: the Healthy Worker Project. Int J Obes Relat Metab Disord 1994;18:145–54.[Medline]
  41. Brown JE, Kaye SA, Folsom AR. Parity-related weight change in women. Int J Obes Relat Metab Disord 1992;16:627–31.[Medline]
  42. Harris HE, Ellison GTH, Holliday M, Lucassen E. The impact of pregnancy on the long-term weight gain of primiparous women in England. Int J Obes Relat Metab Disord 1997;21:747–55.[Medline]
  43. Coakley EH, Rimm EB, Colditz G, Kawachi I, Willett W. Predictors of weight change in men: results from The Health Professionals Follow-Up Study. Int J Obes Relat Metab Disord 1998;22:89–96.[Medline]
  44. Kaprio J, Sarna S, Koskenvuo M, Rantasalo I. The Finnish Twin Registry: formation and compilation, questionnaire study, zygosity determination procedures and research program. Prog Clin Biol Res 1978;24B:179–84.
  45. Kaprio J, Sarna S, Koskenvuo M, Rantasalo I. The Finnish Twin Registry: baseline characteristics. Section II. History of symptoms and illnesses, use of drugs, physical characteristics, smoking, alcohol and physical activity. Helsinki: University of Helsinki Press, 1978.
  46. Hublin C, Kaprio J, Partinen M, et al. The prevalence of narcolepsy: an epidemiological study of the Finnish Twin Cohort. Ann Neurol 1994;35:709–16.[Medline]
  47. Harrela M, Koistinen H, Kaprio J, et al. Genetic and environmental components of interindividual variation in circulating levels of IGF-I, IGF-II, IGFBP-1, and IGFBP-3. J Clin Invest 1996;98:2612–5.[Medline]
  48. Kaprio J, Koskenvuo M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. J Clin Epidemiol 1988;41:139–50.[Medline]
  49. Romanov K, Rose RJ, Kaprio J, Koskenvuo M, Langinvainio H, Sarna S. Self-reported alcohol use. A longitudinal study of 12 994 adults. In: Lindros KO, Ylikahri R, Kiianmaa K, eds. Advances in biomedical alcohol research. Alcohol and alcoholism. Oxford, United Kingdom: Pergamon Journals Ltd, 1987:619–23.
  50. Koskenvuo M, Langinvainio H, Kaprio J, Rantasalo I, Sarna S. The Finnish Twin Registry: baseline characteristics. Section III. Occupational and psychosocial factors. Helsinki: Kansanterveystieteen Julkaisuja, 1979.
  51. Fogelholm M, Kaprio J, Sarna S. Healthy lifestyles of former Finnish world class athletes. Med Sci Sports Exerc 1994;26:224–9.[Medline]
  52. Kleinbaum DG. Logistic regression. A self-learning text. New York: Springer Verlag, 1994.
  53. Chmura Kraemer H. What is the ‘right’ statistical measure of twin concordance (or diagnostic reliability and validity)? Arch Gen Psychiatry 1997;54:1121–4.[Medline]
  54. Breslow NE, Day NE. In: Statistical methods in cancer research. Vol 1. The analysis of case-control studies: Lyon, France: International Agency for the Research on Cancer, 1980.
  55. Bild DE, Sholinsky P, Smith DE, Lewis CE, Hardin JM, Burke GL. Correlates and predictors of weight loss in young adults: the CARDIA study. Int J Obes Relat Metab Disord 1996;20:47–55.[Medline]
  56. Perry GS, Byers TE, Mokdad AH, Serdula MK, Williamson WF. The validity of self reports of past body weights by U.S. adults. Epidemiology 1995;6:61–6.[Medline]
  57. Kuskowska-Wolk A, Karlsson P, Stolt M, Rössner S. The predictive validity of body mass index based on self-reported weight and height. Int J Obes 1989;13:441–53.[Medline]
  58. Fortenberry JD. Reliability of adolescents' reports of height and weight. J Adolesc Health Care 1992;13:114–7.
  59. Casey VA, Dwyer JT, Berkey CS, Coleman KA, Gardner J, Valadian I. Long-term memory of body weight and past weight satisfaction: a longitudinal follow-up study. Am J Clin Nutr 1991;53:1493–8.[Abstract/Free Full Text]
  60. Iribarren C, Sharp DS, Burchfiel CM, Petrovitch H. Association of weight loss and weight fluctuation with mortality among Japanese American men. N Engl J Med 1995;333:686–92.[Abstract/Free Full Text]
  61. French SA, Jeffery RW, Folsom AR, McGovern P, Williamson DF. Weight loss maintenance in young adulthood: prevalence and correlations with health behavior and disease in a population-based sample of women aged 55–69 years. Int J Obes Relat Metab Disord 1996;20:303–10.[Medline]
  62. Jeffery RW, French SA, Forster JL, Spry VM. Socioeconomic status differences in health behaviors related to obesity: The Healthy Worker Project. Int J Obes 1991;15:689–96.[Medline]
  63. Drewnowski A, Kurth CL, Krahn DD. Body weight and dieting in adolescence: impact of socioeconomic status. Int J Eating Disord 1994;16:61–5.[Medline]
Received for publication June 17, 1998. Accepted for publication May 10, 1999.




This article has been cited by other articles:


Home page
 Arch Pediatr Adolesc MedHome page
D. L. Paineau, F. Beaufils, A. Boulier, D.-A. Cassuto, J. Chwalow, P. Combris, C. Couet, B. Jouret, L. Lafay, M. Laville, et al.
Family Dietary Coaching to Improve Nutritional Intakes and Body Weight Control: A Randomized Controlled Trial
Arch Pediatr Adolesc Med, January 1, 2008; 162(1): 34 - 43.
[Abstract] [Full Text] [PDF]

Home page
 J. Am. Coll. Nutr.Home page
H. Tamim, N. Dumit, A. Terro, R. Al-Hourany, D. Sinno, F. Seif, S. Steitieh, and U. Musharrafieh
Weight Control Measures among University Students in a Developing Country: A Cultural Association or a Risk Behavior
J. Am. Coll. Nutr., October 1, 2004; 23(5): 391 - 396.
[Abstract] [Full Text] [PDF]

Home page
 Scand J Public HealthHome page
M. Hernelahti, U. M. Kujala, and J. Kaprio
Stability and change of volume and intensity of physical activity as predictors of hypertension
Scand J Public Health, August 1, 2004; 32(4): 303 - 309.
[Abstract] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.
Agricola
Right arrow Articles by Korkeila, M.
Right arrow Articles by Koskenvuo, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS