Discussion

Discussion of the study results should be preceded by indicating certain interpretative limitations because of the cross-sectional nature of the study. The biological, socio-demographic and lifestyle factors were all measured via self-report, introducing the potential for socially desirable response patterns. Adding objective measures of physical activity, diet and other health-related behaviours may improve future research. The percentage of people who replied to the questions included in the questionnaire in 1983 was 70%, whereas only 36% of the children measured in 2000 had a questionnaire on their health behaviour completed by their parents. Despite the mentioned limitation, the obtained results, as based on a considerable number of observations conducted twice – first in 1983 and secondly in 2000 – can be, with a certain degree of approximation, treated as referring to the whole urban population. Apart from that, the strengths of this study were direct anthropometric measurements and several comprehensive, empirically derived social and lifestyle elements. The results of the performed tests should also be analysed within the context of the political transformation that took place in Poland at the end of the 20th century. Years of relative stability and an impression of prosperity based on loans from abroad (1970–1977) were followed by a deep economic crisis and collapse of the economy (1977–1989) accompanied by food shortages. After 1989, the process of post-socialist transition from a planned to a free-market economy was also accompanied by intensified psychosocial stress and feelings of insecurity, and this, as a result, affected the health of the Polish people (Watson, Reference Watson2006; Lipowicz, Reference Lipowicz2007). Higher differentiation of abdominal obesity risk factors in the 2000 cohort in comparison with the 1983 one may have resulted from these social and economic changes.

The results of the study indicate that girls whose parents were both obese had a higher risk of abdominal obesity than those with slim parents. The results refer both to the 1983 and 2000 series and are in conformity with the results of other studies showing that both maternal and paternal BMI can be associated with total and central adiposity in their offspring (Davey Smith et al., Reference Davey Smith, Steer, Leary and Ness2007, Labayen et al., Reference Labayen, Ruiz, Ortega, Loit, Harro, Veidebaum and Sjostrom2010; Vale et al., Reference Vale, Soares-Miranda, Santos, Moreira, Marques and Santos2011). An analysis of the data for boys showed that obesity in both parents increased the risk of obesity in the examined boys, but only in one series, i.e. in 2000.

According to the literature, the risk of obesity prevalence in children is higher in the case of mothers' obesity than fathers' obesity (Shrewsbury & Wardle, Reference Shrewsbury and Wardle2008; Whitaker et al., Reference Whitaker, Jarvis, Beeken, Boniface and Wardle2010), which is explained by the fact that the mothers spend more time with their children and prepare meals, which affects their families' eating habits. Similar dependencies were confirmed in the study, when abdominal obesity was assessed according to WC criterion. However, there are studies that indicate a reverse dependency, pointing at the fathers' role in arranging leisure time activities, which affects their families' physical condition (Fogelholm et al., Reference Fogelholm, Nuutinen, Pasanen, Myöhänen and Säätelä1999; Davison et al., Reference Davison, Cutting and Birch2003; Ferreira et al., Reference Ferreira, van der Horst, Wendel-Vos, Kremers, van Lenthe and Brug2006).

Fathers' abdominal obesity increased the risk of abdominal obesity in their sons, while mothers' abdominal obesity more than doubled the risk in their daughters. A similar model demonstrating that the risk of obesity prevalence in children was higher if their parent of the same sex was obese, has been described by Kivimaki et al. (Reference Kivimaki, Lawlor, Smith, Elovainio, Jokela and Keltikangas-Jarvinen2007), Lawlor et al. (Reference Lawlor, Smith, O'Callaghan, Alati, Mamun, Williams and Najman2007) and Perez-Pastor et al. (Reference Perez-Pastor, Metcalf, Hosking, Jeffery, Voss and Wilkin2009). The directions of the dependencies are not fully determined as some studies have shown that maternal BMI was more strongly associated with female offspring than male (Perez-Pastor et al., Reference Perez-Pastor, Metcalf, Hosking, Jeffery, Voss and Wilkin2009), while others, e.g. Leary et al. (Reference Leary, Davey Smith and Ness2010), have indicated that paternal associations are similar for both sexes.

Although the information on parents' body weight and height used to calculate BMI in the study came from questionnaire data, the accuracy of such data has been confirmed by other studies (Bolton Smith et al., Reference Bolton-Smith, Woodward, Tunstall-Pedoe and Morrison2000; Kuczmarski et al., Reference Kuczmarski, Kuczmarski and Najjar2001).

The results of the present study did not indicate a significantly increased risk of abdominal obesity in children whose mothers were older than 30 years when they delivered their child compared with those under 30. The results are consistent with those of studies of BMI in Polish (Mazur et al., Reference Mazur, Klimek and Małecka-Tendera2011) and Italian populations (Maffeis et al., Reference Maffeis, Micciolo, Must, Zaffanello and Pinelli1994). Many studies have shown that high birth weight of a child is an independent risk factor for overall obesity (Maffeis et al., Reference Maffeis, Micciolo, Must, Zaffanello and Pinelli1994; Mazur et al., Reference Mazur, Klimek, Telega, Hejda, Wdowiak and Małecka-Tendera2008) and high waist circumference (Ortega et al., Reference Ortega, Ruiz and Sjöström2007). Such dependencies were confirmed in this study for girls of birth weight ≥4000 g.

Of the socio-demographic factors studied, low parental education level increased the risk of abdominal obesity by two times for girls, but not for boys. Low parental education can be connected with unemployment, and thus worsening of family living conditions, incorrect diet, lack of educational and sport activities and psychosocial stress (Parizkova, Reference Parizkova2008; Shrewsbury & Wardle, Reference Shrewsbury and Wardle2008). Ferreira et al. (Reference Ferreira, van der Horst, Wendel-Vos, Kremers, van Lenthe and Brug2006) found that mothers' education level and family income emerged as independent correlates of adolescents' physical activity level. In the 2000 series a significant risk factor for abdominal obesity in boys was being an only child, which can be associated with an inactive way of spending leisure time (computer games, watching TV) and parents' excessive care (Deforche et al., Reference Deforche, De Bourdeaudhuij, Tanghe, Hills and De Bode2004). Similar results indicating an increased risk of obesity prevalence (BMI) in Polish children from families with two or fewer children were obtained by Mazur et al. (Reference Mazur, Klimek, Telega, Hejda, Wdowiak and Małecka-Tendera2008). In one study exploring family structure and physical activity those children who had an older sibling were more likely to spend time in physical activity of moderate intensity (Bagley et al., Reference Bagley, Salmon and Crawford2006).

Among the lifestyle factors studied, cigarette smoking by parents turned out to be a significant factor increasing the risk of abdominal obesity in their children. The risk increased in girls when either their father or mother smoked, whereas in boys a significantly higher risk occurred when their mothers were smokers. Smoking is commonly associated with more sedentary lifestyles (Kaczynski et al., Reference Kaczynski, Manske, Mannell and Grewal2008), usually shared by other family members. Studies on parental influences on tobacco use show that parents who have household rules against smoking and who do not smoke are more likely to have adolescents who do not smoke (O'Byrne et al., Reference O'Byrne, Haddock and Poston2002; Andersen et al., Reference Andersen, Leroux, Bricker, Rajan and Peterson2004). It has been suggested that paternal smoking during pregnancy is a risk factor for childhood obesity almost similar in amount to smoking by the mother (Leary et al., Reference Leary, Smith, Rogers, Reilly, Wells and Ness2006).

Higher frequency of meals consumed by children and adolescents indicated a lower risk of abdominal obesity but significance of the relationship was not confirmed statistically. Significant dependencies have been observed by, among others, Mota et al. (Reference Mota, Fidalgo, Silva, Ribeiro, Santos, Carvalho and Santos2008) with BMI and Barba et al. (Reference Barba, Troiano, Russo and Siani2006) with central fat deposition, showing that an additional meal reduces the probability of being at risk of obesity and central fat deposition in both boys and girls. Physical activity reduces excessive adipose tissue and many studies have demonstrated its influence on decreasing body fatness measures, including waist circumference in children (Klein-Platat et al., Reference Klein-Platat, Oujaa, Wagner, Haan, Arveiler, Schlienger and Simon2005; Ortega et al., Reference Ortega, Ruiz and Sjöström2007; Lehto et al., Reference Lehto, Ray, Lahti-Koski and Roos2011). In a large sample of 12-year-old French boys and girls, Klein-Platat et al. (Reference Klein-Platat, Oujaa, Wagner, Haan, Arveiler, Schlienger and Simon2005) reported that physical activity (>140 min/week outside school) remained independently associated with lower WC, taking into account BMI and sedentary activities (watching TV). The beneficial effect of physical activity on WC was greater in overweight adolescents compared with normal-weight ones (Klein-Platat et al., Reference Klein-Platat, Oujaa, Wagner, Haan, Arveiler, Schlienger and Simon2005). The results of the studies of Ortega et al. (Reference Ortega, Ruiz and Sjöström2007) or Dencker et al. (Reference Dencker, Thorsson, Karlsson, Lindén, Wollmer and Andersen2008) suggest that high levels of vigorous physical activity are beneficial for the prevention of abdominal obesity in the paediatric age group. Lack of undertaking leisure time physical activity by boys, but not by girls, turned out to be a significant risk factor for abdominal obesity. The risk increased by six times in boys if the criterion was lack of regular physical activity realised during organized classes in sport clubs. In the study of Mota et al. (Reference Mota, Fidalgo, Silva, Ribeiro, Santos, Carvalho and Santos2008) boys, but not girls, who were moderately active, were more likely to be of normal weight (BMI). Usually boys tend to be more active in all age groups and physical inactivity is more prevalent in obese girls than their male counterparts (Sallis et al., Reference Sallis, Prochaska and Taylor2000).

The WHtR is an anthropometric measure of central adiposity that has emerged as a better predictor of cardiovascular disease risk factors compared with BMI or WC alone in children and adolescents (Savva et al., Reference Savva, Tornaritis, Savva, Kourides, Panagi and Silikiotou2000; Hara et al., Reference Hara, Saito, Iwata, Okada and Harada2002) and adults (Ashwell et al., Reference Ashwell, Gunn and Gibson2012). The WHtR has been shown to be a simple, non-invasive and practical tool that correlates well with visceral fat (Ashwell & Hsieh, Reference Ashwell and Hsieh2005; Taylor et al., 2011). Furthermore, unlike BMI, WHtR is only weakly correlated with age, negating the need to express values relative to age and sex. However, recent papers have questioned the statistical validity of the WHtR due to the residual correlation with height (Tybor et al., Reference Tybor, Lichtenstein, Dallal and Must2008; Nambiar et al., Reference Nambiar, Truby, Abbott and Davies2009). In children aged 6–13 years WHtR will under-adjust for the effect of height, while in girls aged 14+ it will over-adjust for the effect of height (Tybor et al., Reference Tybor, Lichtenstein, Dallal and Must2008). In the present study the authors applied the border of 90th percentile of WHtR for proper age and sex to eliminate the influence of age. When the border of WHtR ≥50 was employed (data not included), age turned out to be a significant factor for abdominal obesity. However, the data of Taylor et al. (2011) demonstrated that the simple recommendation of having a waist less than half a person's height correctly discriminates between children and adolescents with low and high levels of central body fat more than 90% of the time, regardless of the DXA (dual X-ray absorptometry) referent method that was chosen.

Summarizing, higher differentiation of abdominal obesity risk factors in the 2000 cohort in comparison with the 1983 cohort may have resulted from the social and economic changes taking place in Poland at the end of the 20th century. It should be emphasized that prophylactic activities should be aimed at children with obese parents, those with parents with a low level of education, as well as children from families leading unhealthy lifestyles (smoking habits, lack of physical activity). The issue of norms and cut-off points for the application of the WHtR index to estimate abdominal obesity requires further studies.