Are computer and cell phone use associated with body mass index and overweight? A population study among twin adolescents

Prevalence of overweight is rapidly increasing among children and adolescents. As defined by the International Obesity Task Force criteria, the prevalence of obesity is estimated to be about 10% throughout the world among 5–17-year-olds and dramatically higher in developed countries, reaching about 30% in the Americas [1]. Genetic predisposition to obesity is high, with the heritability of BMI estimated to be 50–90% [2]. However, also environmental factors are important: while genes may predispose a child to weight gain, environmental factors can foster development of obesity.

Increasingly sedentary leisure activities are considered one of the main reasons for the obesity epidemic [1]. According to a meta-analysis based on 52 independent samples [3], there is a statistically significant relationship between TV viewing and body fatness among 3–18-year-olds. Conversely, there is an inverse association between sports/physical activity and weight status in cross-sectional studies [4–8].

Although sedentary leisure time activities involving information technology use have spread rapidly among children and adolescents, the association of weight status with computer and cell phone use has not been thoroughly investigated. Computer use was positively associated with BMI [9] and overweight [10] among adolescent girls but not among boys. Video/computer games were not significantly associated with overweight among children and teenagers [3, 10], nor was video game/computer time among children [11]. One study found a positive association between electronic games and weight status among boys and girls [12] and another among girls only [13]. We were not able to find any studies on the associations between cell phone use and weight status. Cell phone use is, however, positively associated with some health-compromising behaviors, such as tobacco, snuff, and alcohol use [14], which in turn are related to BMI and sedentary lifestyle [15].

Given the paucity of data on the effects of information and communication technology on adolescent health, we studied the relation of the use of these technologies with weight status. Over 90% of Finnish 15–19-year-olds owned a personal cell phone at the end of 2002 [16]. Cell phone use started to spread to all age groups during the mid 90s [17]. The rapid increase of cell phone ownership has multiplied the opportunities to spend time on the phone and total phone time has steadily increased in Finland [19]; this may reduce time spent in physically active leisure activities. Therefore, we hypothesized that cell phone use would be associated with BMI and overweight.

At 1999–2000, boys aged 15–18 y spent in average 50 and girls 10 minutes using a computer during school days. During non-school days, boys spent almost 80 minutes and girls about 20 minutes using computers [18]. We hypothesized that overweight and BMI would be influenced by both computer use hours and home computer ownership, with the highest computer use hours being among adolescents having a home computer with an Internet connection. We tested these hypotheses cross-sectionally in a population sample of 17-year-old girls and boys.

During 2000–2005, the proportion of Finnish 17-year-old adolescents without a home computer steadily decreased from 18% to 8%. Weekly hours of computer use were positively associated with an increased risk of overweight. The proportion of adolescents without a cell phone decreased from 12% to just 1% during our study period. We found only a weak positive association between BMI and the monthly total cell phone bill.

How could computer and cell phone use affect weight status? An inverse association of physical activity with BMI and obesity among adolescents has been confirmed in many studies [4–8]. Television viewing time, another important behavioral correlate of obesity [3], has been hypothesized to affect weight via three main pathways: by 1) taking time away from more physically active behaviors; 2) increasing energy intake by snacking while viewing and by television commercials; and 3) decreasing the metabolic rate [29]. Conversely, video/computer game playing has been associated with increases in various metabolic and physiologic variables [30, 31]. Similarly to TV viewing, computer and cell phone use may increase the proportion of time devoted to sedentary activities. The rapid increase of cell phone ownership has multiplied the opportunities to spend time on the phone and total phone time (cell phone & landline combined) has steadily increased in Finland [19]. When a cell phone call, text message, or e-mail replaces a trip to a friend's house or eliminates the need to run an errand, daily physical activity is reduced.

Computer use has been found to be positively associated with weight status in adolescent girls, but not in boys among 4746 American teenagers in 1998–99 [9] and among 6515 Finnish teenagers in 2001 [10]. Since then, home computers have become almost ubiquitous: we thus wished to reassess their association with overweight and BMI. In this study, we observed that computer use was associated with weight status independent of gender, with significant positive associations between computer use and overweight in boys and between computer use and BMI among girls (gender-specific results not shown).

There were some incongruences in the results considering computer ownership; adolescents having a home computer with an Internet connection had the highest computer use hours, but their BMI and risk of overweight did not significantly differ from those not having a home computer. However, adolescents who had a home computer not connected to the Internet had a significantly higher BMI and a higher risk of overweight compared to adolescents not having a home computer at all. There are many possible explanations for this: either computer use might be somehow different in the presence of an Internet connection, or people with and without an Internet connection might recall their computer use hours differently, or the differences could be due to residual confounding. In any case, these questions merit further investigation.

The results were partly different when using BMI vs. overweight as an outcome, which implies that the distribution of BMI is probably different in different categories of the predictors. BMI had no overall association with physical exercise, while overweight was strongly negatively associated with exercise. Among normal-weight and underweight adolescents, differences in fat-free muscle mass can explain a large part of the BMI differences, though a high BMI appears a good indicator of excess fat mass [32].

The main limitation of this study is its cross-sectional design, which does not allow confident causal inferences. We cannot exclude the possibility that overweight or an increase in BMI increase computer or cell phone use. Prospective studies are needed to confirm our findings. It is also important to consider other potential confounders that may influence our results, such as eating habits, the availability of pocket money, and certain personality traits (impulse control, sociability). Future studies should address the association of new technology use and weight status by carefully adjusting for nutritional intake, multiple socioeconomic factors, and other potential confounders, preferably in a prospective study design.

Our study is also limited by its reliance on self-reported behaviors. Objective tracking of patterns of computer and cell phone use would yield more reliable results, but was not feasible in this large cohort of adolescents spread over a large and sparsely populated geographical area. Also, verifying billing information or actual time spent on a cell phone from cellular network providers would enhance reliable results, but was impossible because of privacy protection requirements. In Finland, the monthly cell phone bill is a relatively good proxy measure for actual time spent on the phone, because outgoing calls are typically priced on a per minute basis and text messages on a per item basis; incoming calls and text messages are usually free of charge. During the study period charging for cell phone use was uniform in Finland although the prevailing practice is about to change.

Self-reported heights and weights are also potential sources of bias. Reliance on self-report underestimates the true prevalence of overweight [33, 34]. Among 418 boys and girls whose mean age was 16.30 y, it led to under-estimation of overweight by 4.8% and obesity by 1.6% [33]. However, our main goal was not to estimate prevalence of overweight but to study the associations between information and technology use and BMI/overweight. Only if self-report bias of weight and height differs by computer and cell phone use, would the associations observed in our study be biased.

Finally, the generalizability of results from a study of twins to non-twin individuals may be questioned. But differences specific to twins in anthropometric, social, and physical activity measures during late adolescence are relatively minor [15][35–37], and they are unlikely to affect the relationship of computer and cell phone use with the risk of overweight. It is possible that hours of computer use are reduced in twin families because many families share just one home computer. However, this would closely resemble the situation of families with many non-twin children close in age.

The strengths of our study include large and population-based study sample, its high participation rates, and the availability of information from the parents. The data are very recent, reflecting developments in the 21^st century, particularly the widespread adaptation of information and communication technology. Finland is well suited for this kind of study given its early adoption and high penetration of communication technology use.

This study confirms that the global epidemic of overweight and the use of new technologies may be related phenomena: weight status is associated with computer and possibly cell phone use. However, given that computer and cell phone use constitutes only a small part of adolescents' leisure time, broader patterns of leisure time use should be addressed in a longitudinal setting to obtain a comprehensive view of their relationship with weight status. Future research should confirm and expand our findings to ensure that the use of modern communication technology does not have undesired health effects among children and adolescents.