Identifying the 'red flags' for unhealthy weight control among adolescents: findings from an item response theory analysis of a national survey.

Abstract

BACKGROUND: Weight control behaviors are common among young people and are associated with poor health outcomes. Yet clinicians rarely ask young people about their weight control; this may be due to uncertainty about which questions to ask, specifically around whether certain weight loss strategies are healthier or unhealthy or about what weight loss behaviors are more likely to lead to adverse outcomes. Thus, the aims of the current study are: to confirm, using item response theory analysis, that the underlying latent constructs of healthy and unhealthy weight control exist; to determine the 'red flag' weight loss behaviors that may discriminate unhealthy from healthy weight loss; to determine the relationships between healthy and unhealthy weight loss and mental health; and to examine how weight control may vary among demographic groups. METHODS: Data were collected as part of a national health and wellbeing survey of secondary school students in New Zealand (n = 9,107) in 2007. Item response theory analyses were conducted to determine the underlying constructs of weight control behaviors and the behaviors that discriminate unhealthy from healthy weight control. RESULTS: The current study confirms that there are two underlying constructs of weight loss behaviors which can be described as healthy and unhealthy weight control. Unhealthy weight control was positively correlated with depressive mood. Fasting and skipping meals for weight loss had the lowest item thresholds on the unhealthy weight control continuum, indicating that they act as 'red flags' and warrant further discussion in routine clinical assessments. CONCLUSIONS: Routine assessments of weight control strategies by clinicians are warranted, particularly for screening for meal skipping and fasting for weight loss as these behaviors appear to 'flag' behaviors that are associated with poor mental wellbeing.