Elsevier

Social Science & Medicine

Volume 70, Issue 8, April 2010, Pages 1194-1202
Social Science & Medicine

Individual, family, and area predictors of BMI and BMI change in an adult Norwegian population: Findings from the HUNT study

https://doi.org/10.1016/j.socscimed.2010.01.007Get rights and content

Abstract

The global obesity epidemic is a major public health concern and there is strong evidence that the drivers are varied and operate via diverse pathways. Taking a systems approach allows the contextual influences operating upon the individual to be identified and quantified. We adopt such a perspective in this study, where longitudinal data from a cohort of 24,966 settled individuals participating in two major health surveys, the Nord-Trøndelag Health Study (HUNT 1 and 2) in the county of Nord-Trøndelag, Norway, were used to investigate associations between individual, family and area characteristics and two outcomes: body mass index (BMI) at follow-up and BMI change over an 11 year period. Linear multilevel models were fitted, with individuals nested in 17,500 families, 447 wards and 24 municipalities. A range of putative individual, family, and area predictors were tested. We found both outcomes were strongly associated with individual characteristics, with higher BMIs generally being amongst males, unmarried participants, non-smokers, those of lower education and those undertaking physically demanding work but participating in less physical activity outside work. The characteristics of those in the sample exhibiting higher BMI gain were rather similar except that women gained more and those with no employment income gained less. Contextual influences were also found to be important: although just 1% of the unexplained variance was located on the neighbourhood and municipality levels respectively, and hence suggesting small environmental influences, between 10 and 13% could be attributed to families, highlighting the importance of the familial contextual environment. Rather little is known about the manner by which family influences may operate on bodyweight hence further work is needed to understand likely mechanisms and guide future interventions.

Introduction

The increasing prevalence of obesity has been described as an epidemic process (James, 2008) and is now a major driver behind rises in the prevalence of certain chronic diseases and disabilities worldwide. Recent projections from the World Health Organisation (WHO) estimate that globally 1.6 billion adults were overweight and at least 400 million were obese in 2005, with these figures expected to grow to 2.3 billion and 700 million by 2015 (WHO, 2006). In Europe current trends are expected to give rise to 150 million obese adults by 2010 (Branca, Nikogosian, & Lobstein, 2007). The annual rate of increase in childhood obesity in this region is a particular cause for concern, with the current prevalence being over 10 times higher than in 1970 and with 15 million children expected to be obese by 2015 (Branca et al., 2007). Indeed, some have forecasted that youths of today may, on average, live less healthy and possibly even shorter lives than their parents (Olshansky et al., 2005).

In a recent WHO publication, Branca et al. (2007, p. xiii) state that “obesity presents Europe with an unprecedented public health challenge that has been underestimated, poorly assessed and not fully accepted as a strategic governmental problem with substantial economic implications”. The authors further note that the prevention of obesity requires innovative environmental approaches. The term “obesogenic environment” refers to the role environmental factors may play in determining food intake and physical activity, both important determinants of bodyweight (Jones, Bentham, Foster, Hillsdon, & Panter, 2007). Swinburn and colleagues have defined the concept as “the sum of influences that the surroundings, opportunities, or conditions of life have on promoting obesity in individuals or populations” (Swinburn, Egger, & Raza, 1999, p. 564). These authors have further described the environment in terms of micro and macro components, where micro-environments are defined as settings that influence peoples' interactions (e.g. home, school, workplace, and neighbourhood) which are themselves influenced by macro-environments (e.g. the education and health system, government policy and society's attitudes and beliefs).

The concept of an obesogenic environment is grounded in a systems perspective where health related behaviour is contextualised in the environment within which it takes place. The advancement of the concept has been driven by the development of ecological models which suggest that weight related behaviours, such as food intake and physical activity, arise as the result of the combined action of psychosocial, demographic, as well as physical environmental processes (Diez Roux, 2007, Van Dyck et al., 2009).

Numerous different environmental factors at various geographical scales have been put forward as potential determinants of overweight and obesity (Black & Macinko, 2008). There is some evidence of an effect of income inequality, with Pickett and colleagues (Pickett, Kelly, Brunner, Lobstein, & Wilkinson, 2005) finding a positive association between income inequality and rates of obesity in 21 developed countries, whilst Holtgrave and Crosby (2006) found higher levels of social capital in US states to be associated with a lower prevalence of obesity. At the neighbourhood level, many studies in the US (e.g. Diez-Roux et al., 2000, Janssen et al., 2006, Rundle et al., 2008), UK (e.g. Moon, Quarendon, Barnard, Twigg, & Blyth, 2007) and in Canada (e.g. Ross et al., 2007) have reported that high material deprivation levels are associated with elevated adult obesity prevalence.

A large amount of research evidence on the role of the physical environment is available. A recent review of predominantly US studies found that the majority reported an association between some aspect of the neighbourhood built environment and obesity, with associations with features such as the walkability of neighbourhoods and the accessibility of greenspaces being found (Papas et al., 2007). In addition to objectively measured features, the findings of a recent meta-analysis support the view that perceptions of the neighbourhood environment, such as those regarding safety and the accessibility of destinations, are also important (Duncan, Spence, & Mummery, 2005).

A contextual unit rarely acknowledged in the obesity literature is the family or household. Yet here is compelling evidence in support of interventions at the parent or family level in paediatric obesity research (Zeller et al., 2007). Parental obesity has a strong predictive power in the development of child and adolescent obesity, arguably with a genetic component, but there are also studies suggesting that there are indeed modifiable determinants operating at this contextual level (Krahnstoever Davison, Francis, & Birch, 2005). These could include the availability of foods, and the provision of familial social support for physical activity, and weight management practices.

There is strong evidence that the drivers of the epidemic are varied and operate via diverse pathways. Systems approaches (see Bailey, 1994) to exploring health behaviour causation can be useful in such situations, as they allow the outcomes of individual actions to be examined within the social and environmental contexts within which the individual operates. We adopt such a perspective in this study, where the aim is to contribute to an understanding of how environments may directly and indirectly affect behaviour and how such behaviour is ultimately expressed in terms of bodyweight. This is done firstly by quantifying variation in body mass index (BMI) and changes in BMI associated with individual, family, and area characteristics. Secondly we try to explain how the composition of individuals in families and areas may account for this variation. Finally we test how contextual features like family and area socioeconomic status, and area deprivation and social capital may explain variability in the outcomes not associated with characteristics of individuals. The research is longitudinal, utilising anthropometric height and weight measurements from two time points amongst a large and settled adult population from the county of Nord-Trøndelag in Norway.

Section snippets

Data sources

The Nord-Trøndelag Health Study (HUNT) is one of the world's largest population health surveys. The first wave (HUNT 1) was carried out in 1984–1986. All residents of Nord-Trøndelag County in Norway aged 20 or above were invited to participate in the study which included a physical examination and self-completed questionnaires. Questionnaire 1 was attached to the invitation letter, and 74,977 (88.1%) of the 85,100 eligible persons completed and returned it when they attended. Of these, 74,332

Results

Table 1 describes the characteristics of the sample and the contextual variables. Mean BMI increased by nearly 2 kg/m2 during the 11 years of follow-up, and the standard deviation increased from 3.47 to 3.92. Mean age at baseline was 43 years. Males and females are equally represented and the majority were married. The variables measuring individual SES indicate that most individuals reported working in manual occupations and had a low educational level. Nevertheless, the majority did not

Discussion

In this study we have examined the correlates of BMI and BMI change in a large longitudinal sample of adult individuals. We found that both outcomes were associated with individual characteristics, with higher BMI most often found amongst males, unmarried participants, non-smokers, those of lower education and those undertaking physically demanding work but participating in less physical activity outside work. The characteristics of those in the sample exhibiting higher BMI gain were rather

Conclusion

In conclusion, we found that a statistically significant proportion of the variance in BMI and BMI change can be attributed to families, in addition to the individual. We were able to quantify the family components that determined our outcomes but insight from elsewhere does suggest that processes governing physical activity and nutrition in families are extremely complex (Kegler, Escoffery, Alcantara, Ballard, & Glanz, 2008). The small amounts of variance attributable to the ward and

Acknowledgements

The Nord-Trøndelag Health Study (The HUNT Study) is a collaboration between HUNT Research Centre (Faculty of Medicine, Norwegian University of Science and Technology NTNU), Nord-Trøndelag County Council and The Norwegian Institute of Public Health. The present study was supported by grants from the Faculty of Social Sciences and Technology Management, Norwegian University of Science and Technology and was carried out in 2009. The authors thank Stig H. Jørgensen, Steinar Krokstad, Johan Håkon

References (64)

  • G. Moon et al.

    Fat nation: deciphering the distinctive geographies of obesity in England

    Social Science & Medicine

    (2007)
  • A. Rundle et al.

    Personal and neighborhood socioeconomic status and indices of neighborhood walk-ability predict body mass index in New York City

    Social Science & Medicine

    (2008)
  • M. Stafford et al.

    Pathways to obesity: identifying local, modifiable determinants of physical activity and diet

    Social Science & Medicine

    (2007)
  • K. Sundquist et al.

    Linking social capital and self-rated health: a multilevel analysis of 11,175 men and women in Sweden

    Health and Place

    (2007)
  • B. Swinburn et al.

    Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity

    Preventive Medicine

    (1999)
  • D. Van Dyck et al.

    Neighbourhood walkability and its particular importance for adults with a preference for passive transport

    Health & Place

    (2009)
  • G. Veenstra et al.

    Who you know, where you live: social capital, neighbourhood and health

    Social Science & Medicine

    (2005)
  • K.D. Bailey

    Sociology and the new systems theory: Toward a theoretical synthesis

    (1994)
  • J.L. Black et al.

    Neighborhoods and obesity

    Nutrition Reviews

    (2008)
  • T. Blakely et al.

    Multilevel studies

  • T.A. Blakely et al.

    Ecological effects in multi-level studies

    Journal of Epidemiology and Community Health

    (2000)
  • F. Branca et al.

    The challenge of obesity in the WHO European region and the strategies for response

    (2007)
  • V. Burke et al.

    Family lifestyle and parental body mass index as predictors of body mass index in Australian children: a longitudinal study

    International Journal of Obesity

    (2001)
  • P. Congdon et al.

    A multi-level perspective on small-area health and mortality: a case study of England and Wales

    International Journal of Population Geography

    (1997)
  • A.V. Diez Roux

    Bringing context back into epidemiology: variables and fallacies in multilevel analysis

    American Journal of Public Health

    (1998)
  • A.V. Diez Roux

    Multilevel analysis in public health research

    Annual Review of Public Health

    (2000)
  • A.V. Diez Roux

    The study of group-level factors in epidemiology: rethinking variables, study designs, and analytical approaches

    Epidemiologic Reviews

    (2004)
  • M.J. Duncan et al.

    Perceived environment and physical activity: a meta-analysis of selected environmental characteristics

    International Journal of Behavioral Nutrition and Physical Activity

    (2005)
  • J.I. Elstad et al.

    Skewed income distribution and geographical mortality differences

    Tidsskrift for den Norske Laegeforening

    (2005)
  • K.L. Frohlich et al.

    A theorethical proposal for the relationship between context and disease

    Sociology of Health and Illness

    (2001)
  • K. Giskes et al.

    Household and food shopping environments: do they play a role in socioeconomic inequalities in fruit and vegetable consumption? A multilevel study among Dutch adults

    Journal of Epidemiology and Community Health

    (2009)
  • H. Goldstein

    Multilevel statistical models

    (2003)
  • Cited by (28)

    • Neighborhood deprivation and obesity: Sex-specific effects of cross-sectional, cumulative and residential trajectory indicators

      2022, Social Science and Medicine
      Citation Excerpt :

      Stress, the food environment, the availability of physical activity opportunities, social norms, and health care accessibility are among the biological pathways through which lower neighborhood socioeconomic status (SES) could be associated with obesity (Cetateanu and Jones, 2014; Diez Roux, 2003; Dulin-Keita et al., 2013; Kwarteng et al., 2017). Neighborhood SES is measured using a variety of indicators, such as neighborhood poverty level (Do and Zheng, 2017; Kwarteng et al., 2016; Lippert et al., 2017; Ludwig et al., 2011; Murray et al., 2010a, 2010b; Sheehan et al., 2017), average area income level (Joost et al., 2016) and composite indexes (Feng and Wilson, 2015; Kimbro et al., 2017; Stafford et al., 2010; Sund et al., 2010). Neighborhoods are also hypothesized to have differential effects on men and women (Lebel et al., 2010).

    • The impact of marriage and parenthood on male body mass index: Static and dynamic effects

      2017, Social Science and Medicine
      Citation Excerpt :

      Investigating BMI time path may also help explain mixed findings about BMI and marital status in cross-sectional studies. For instance, Noppa and Bengtsson (1980) and Sund et al. (2010) found that married individuals have lower BMI; Kittel et al. (1978), and Umberson et al. (2009) found no differences in BMI between married and non-married individuals; and yet many studies found higher BMI among married individuals (Ball et al., 2002; Hahn, 1993; Heineck, 2006; Klein, 2011; Sobal and Rauschenbach, 2003; Wilson, 2012; Sanz-de Galdeano, 2005). One of the possible reasons for these contradicting results is that these studies failed to consider individual and dynamic effects of marital status changes in a longitudinal setting by measuring the BMI impact of periods preceding and following marriage and divorce.

    • The impact of neighborhood on physical activity in the Jackson Heart Study

      2016, Preventive Medicine
      Citation Excerpt :

      Census tract was used to represent neighborhood, similar to other studies (Hickson et al., 2011; Krieger et al., 2003; Mujahid et al., 2008). Analyses were conducted using multi-level modeling to account for individual and neighborhood contributions to variance components, as well as clustering of nested data (Hearst et al., 2012; Sund et al., 2010). The two individual-level PA outcomes for this study were Active Living and Sports & Exercise (Sports).

    • Higher body mass index, less exercise, but healthier eating in married adults: Nine representative surveys across Europe

      2015, Social Science and Medicine
      Citation Excerpt :

      Relatedly, married couples also tend to have more financial resources (Averett et al., 2008) and are thus better able afford a healthier lifestyle (e.g., buying fresh produce or a gym membership). Echoing the conflicting predictions of the explanations reviewed above, empirical findings on the relation between BMI and marital status are mixed: Some cross-sectional studies have found that married individuals have a lower BMI (e.g., Noppa and Bengtsson, 1980; in a population sample of Swedish women; Sund et al., 2010; in a sample of Norwegian men and women); other studies have found no differences in the BMI of married and never married individuals (e.g., Kittel et al., 1978; in an industrial population of Belgian men; Umberson et al., 2009; in a U.S. national sample); and still other studies have observed married individuals to have a higher BMI (e.g., in a large Australian random sample, Ball et al., 2002; in a national US sample, Hahn, 1993; in representative German samples, Heineck, 2006, Klein, 2011; and in representative U.S. samples in which only married men—not women—had a higher BMI, Sobal et al., 1992; Wilson, 2012). Mixed results have also been obtained in longitudinal studies: Although studies examining weight changes across marital transitions (e.g., from being single to getting married) often show that either both partners (Averett et al., 2013; Meltzer et al., 2013) or women, in particular, gain weight, other studies have found no such regularity (Dinour et al., 2012 for a review).

    • Postmortem heart weight: Relation to body size and effects of cardiovascular disease and cancer

      2014, Cardiovascular Pathology
      Citation Excerpt :

      Further, while absolute heart weight decreased with higher age, adjustment for body weight and gender suggested a relative increase in heart weight. Consistent with global trends, the prevalence of obesity has increased in Norway during the past decades [23–25]. Overweight and obesity are associated with arterial hypertension and other conditions [26] that may affect heart weight.

    View all citing articles on Scopus
    View full text