Television Watching and Other Sedentary Behaviors in Relation to Risk of Obesity and Type 2 Diabetes Mellitus in Women FREE

Current public health campaigns to reduce obesity and type 2 diabetes have largely focused on increasing exercise levels, but have paid little attention to the reduction of sedentary behaviors. Television (TV) watching is a major sedentary behavior in the United States. In a survey conducted in 1997, an adult male spent approximately 29 hours per week watching TV, and an adult female spent 34 hours per week.¹ In recent decades, in parallel with increasing obesity, there has been a steady increase in the number of homes with multiple TV sets, videocassette recorders (VCRs), cable TV, and remote controls, as well as the number of hours spent watching TV.¹

Compared with other sedentary activities such as sewing, playing board games, reading, writing, and driving a car, TV watching results in a lower metabolic rate.² Constant exposure to food advertising leads to increased food and calorie intake and unhealthy eating patterns.^3- 5 It is well established that prolonged TV watching is associated with obesity in children.^6- 8 However, the role of TV watching compared with other sedentary behaviors, such as sitting at work or reading, in the development of obesity and type 2 diabetes among adults has not been well studied, especially among women. We therefore examined prospectively the relationship between several common sedentary behaviors and incidence of obesity and type 2 diabetes in a large cohort of women.

The Nurses' Health Study cohort was established in 1976 when 121 700 female registered nurses aged 30 to 55 years and residing in 1 of 11 US states responded to mailed questionnaires regarding their medical history and health practices; details have been published elsewhere.⁹ For this study, we excluded women with diagnosed cardiovascular disease (n = 3102), cancer (n = 9917), or diabetes (n = 4373) in 1992. The final sample for the diabetes analysis consisted of 68 497 women. For the obesity analysis, we also excluded women who were already obese (body mass index [BMI] ≥30) in 1992 or earlier cycles, leaving 50 277 women in the final analysis.

In 1992, participants were asked to report their average weekly time spent sitting at home while watching TV or VCR, sitting at work or away from home or while driving, and other sitting at home (eg, reading, meal times, at desk). They were also asked to report time spent standing or walking around at home or at work. The responses included 9 categories (ranging from 0 h/wk to >90 h/wk). In the current analyses, 5 categories were coded consistently across all the 5 items (0-1, 2-5, 6-20, 21-40, and >40 h/wk).

In 1992, 1994, and 1996, participants were asked the amount of time they spent on average per week on each of the following physical activities: walking, jogging, running, bicycling, calisthenics/aerobics/aerobic dance/rowing machine, lap swimming, squash/racquetball, and tennis. They were also asked about their usual walking pace, specified as easy/casual (<2 miles per hour [mph]), normal (2-2.9 mph), brisk (3-3.9 mph), or very brisk/striding (≥4 mph). From this information, weekly energy expenditure in metabolic equivalent hours (MET-hours) was calculated.² For example, brisk walking requires an energy expenditure of about 4 METs and is considered to be a moderate-intensity activity. In this cohort, walking is the most common type of activity (60% of all women reported that they walked ≥1 h/wk). Because the results using baseline and updated physical activity measures were similar, we reported only baseline analyses.

The reproducibility and validity of the physical activity questionnaire has been described elsewhere.¹⁰ In a representative sample (n = 147) of participants in the Nurses' Health Study II cohort, the 2-year test-retest correlation for activity was 0.59. The correlation between physical activity reported on 1-week recalls and that reported on the questionnaire was 0.79. The correlation between activity reported in diaries and that reported on the questionnaire was 0.62. In a separate study on a population aged 20 to 59 years recruited from a university community (n = 103), the correlation between physical activity score on a very similar questionnaire and maximum oxygen consumption was 0.54.¹¹ In a parallel cohort of men, average hours of TV watching were significantly associated with higher levels of leptin and low-density lipoprotein cholesterol and lower levels of high-density lipoprotein cholesterol and apolipoprotein A1.¹²

Body weight was self-reported in the biennial questionnaires. Self-reported weights were highly correlated with measured weights (r = 0.96; mean difference, 1.5 kg).¹³ In 1976, the nurses were asked to report their height to the closest inch. Body mass index (BMI) is calculated as weight in kilograms divided by the square of height in meters. Incidence of obesity is defined as the transition from nonobese (BMI <30) in 1992 to a BMI of 30 or more at the end of follow-up in 1998. Thus, only individuals with a BMI of less than 30 across all time points (between 1976 and 1992) were included in the obesity analysis (n = 50 277).

A supplementary questionnaire regarding symptoms, diagnostic tests, and hypoglycemic therapy was mailed to women who indicated on any biennial questionnaire that they had been diagnosed with diabetes. A case of diabetes was considered confirmed if at least 1 of the following was reported on the supplementary questionnaire: (1) classic symptoms plus elevated glucose levels (a fasting plasma glucose concentration ≥140 mg/dL [7.7 mmol/L] or a randomly measured concentration of at least 200 mg/dL [11.1 mmol/L]); (2) at least 2 elevated plasma glucose concentrations on different occasions in the absence of symptoms (levels as above or ≥200 mg/dL [11.1 mmol/L] after ≥2 hours of oral glucose tolerance testing); and (3) treatment with oral hypoglycemic agents or insulin. Our criteria for diabetes classification are consistent with those proposed by the National Diabetes Data Group.¹⁴

The validity of this diagnostic procedure has been verified in a subsample of this study population.¹⁵ In addition, another substudy assessing the prevalence of undiagnosed diabetes suggested a very low rate of false-negative results.¹⁶

Person-time for each participant was calculated from the date of return of the 1992 questionnaires to the date of confirmed type 2 diabetes (for diabetes analysis only), the year of first reported obesity (for obesity analysis only), death from any cause, or June 1, 1998, whichever came first. Incidence rates of obesity or type 2 diabetes were obtained by dividing the number of cases by person-years in each category of average time spent on each sedentary behavior (eg, TV watching). Relative risks (RRs) were computed as the incidence rate in a specific category of TV watching divided by that in the reference category, with adjustment for 5-year age categories. Tests for linear trend across increasing categories of average time spent watching TV were conducted by treating the categories as a continuous variable and assigning the middle score for the category as its value.

We used Cox regression to adjust estimated incidence rate ratios simultaneously for potential confounding variables. In the multivariate analyses of onset of obesity, we adjusted for age (<50, 50-54, 55-59, 60-64, ≥65 years), smoking (never, past, current 1-14, 15-24, and ≥25 cigarettes/d), alcohol consumption (0, 0.1-4.9, 5-14.9, ≥15 g/d), and physical activity (METs in quintiles). We further adjusted for dietary variables including total energy intake, total fat, glycemic load, and cereal fiber (all in quintiles). For multivariate analyses of diabetes, our covariates included age, smoking, alcohol consumption, and family history of diabetes. Further analyses adjusted for dietary intakes of polyunsaturated fat, glycemic load, cereal fiber, and trans fat (all in quintiles).¹⁷

We calculated the population-attributable risk to estimate the percentages of obesity and type 2 diabetes cases in this population that are attributable to the joint effects of 2 risk factors (either >10 h/wk of TV watching or <30 min/d walking or equivalent energy expenditure).¹⁸P = .05 was considered significant. Statistical analysis was performed using SAS statistical software, version 8.2 (SAS Institute Inc, Cary, NC).

Women who spent more time watching TV were more likely to smoke and drink alcohol and less likely to exercise (Table 1). However, the correlation between TV watching and physical activity levels was minimal (r = −0.03). These women also had higher intake of total energy, total and saturated fats, red meat, processed meat, refined grain products, snacks, sweets/desserts, and lower intakes of fish, vegetables, fruits, and whole grains.

During 6 years of follow-up (282 141 person years), 3757 women who were not obese at baseline (7.5%) became obese in 1998. Time spent watching TV was positively associated with risk of obesity (Table 2). The age-adjusted RRs across categories of TV watching (0-1, 2-5, 6-20, 21-40, >40 h/wk) were 1.0, 1.23, 1.42, 1.68, and 2.00, respectively (P for trend, < .001). Further adjustment for exercise levels and other covariates did not appreciably alter the RRs. Sitting at work or away from home or driving was also significantly associated with elevated risk of obesity. In contrast, time spent standing or walking around at home was associated with a lower risk of obesity (multivariate RR comparing extreme categories, 0.77; 95% confidence interval [CI], 0.61-0.97; P for trend, <.001). Further adjustment for baseline BMI substantially attenuated the RRs. The multivariate RR comparing the extreme categories of TV watching was 1.29 (95% CI, 1.01-1.66). These results suggest that women who watched more TV were already on a trajectory to become obese at baseline.

During 6 years of follow-up (396 900 person-years), we documented 1515 newly diagnosed cases of type 2 diabetes. After adjustment for age, average time spent watching TV was significantly associated with increased risk of type 2 diabetes (Table 3). The RRs across categories of average hours spent watching TV per week were 1.0, 1.10, 1.30, 1.53, and 1.98 (P for trend, <.001). Further adjustment for dietary factors slightly attenuated the RRs. As expected, these RRs were substantially attenuated after further adjustment for BMI (RR comparing extreme categories, 1.17; 95% CI, 0.82-1.67), suggesting that the increased risk of diabetes associated with TV watching was largely mediated through obesity.

Sitting at work and other sitting at home (last category only) were both associated with significantly increased risk of diabetes in multivariate analyses adjusting for dietary and nondietary covariates (Table 3). Standing or walking around at home was associated with a modestly lower risk of diabetes.

To compare the predictive role of various sedentary behaviors and physical activity, we conducted a multivariate analysis including simultaneously these behaviors (in continuous form), physical activity (4 METs/d, equivalent to 1 h/d of brisk walking), and nondietary and dietary covariates. Among various sedentary behaviors, time spent watching TV was most strongly associated with obesity risk (Figure 1). For each 2-h/d increase in time spent watching TV, there was a 23% (95% CI, 17%-30%) increase in obesity risk. In contrast, other sitting at home or standing or walking around home was associated with a significantly lower risk of obesity. Each 1-h/d increase in brisk walking was associated with a 24% (95% CI, 19%-29%) reduction in obesity.

Watching TV was also most strongly associated with elevated risk of type 2 diabetes (Figure 1). Each 2-h/d increment in time spent watching TV was associated with a 14% (95% CI, 5%-23%) increase in diabetes risk. Standing or walking around at home was associated with a lower risk of diabetes (12% [95% CI, 7%-16%] risk reduction for each 2-h/d increment). Brisk walking was strongly associated with a decreased risk of type 2 diabetes (34% [95% CI, 27%-41%] risk reduction for each 1-h/d increment).

To address the possibility that surveillance for diabetes may have varied according to physical activity levels, we did an analysis restricted to cases reporting at least 1 symptom of diabetes at diagnosis. Results from this subgroup were not appreciably different from those for the entire cohort. The multivariate RRs across categories of average hours spent watching TV per week were 1.0, 1.09, 1.30, 1.44, and 1.70.

In multivariate analyses, we observed independent effects of TV watching and exercise levels on obesity and type 2 diabetes (Figure 2). Compared with women who were in the most active (the highest tertile of METs/wk) and the lowest TV watching category (<6 h/wk watching TV), those who were in the lowest of METs per week and most sedentary category (≥20 h/wk watching TV) had a significantly increased risk of obesity (RR, 1.90; 95% CI, 1.61-2.24) and type 2 diabetes (RR, 2.89; 95% CI, 2.21-3.79).

We estimated that 30% (95% CI, 24%-36%) of new obesity cases and 43% (95% CI, 32%-52%) of type 2 diabetes cases could be attributable to the joint effects of 2 risk factors (either >10 h/wk of TV watching or <30 min/d walking or equivalent energy expenditure). Only 15% of participants in our cohort belonged to the joint low-risk group (defined as <10 h/wk of TV watching and ≥30 min/d of walking or equivalent energy expenditure).

In this large prospective cohort of women, sedentary behaviors, especially TV watching, were significantly associated with risk of obesity and type 2 diabetes. Sedentary occupation reflected by long hours of sitting or standing at work was significantly associated with risk of obesity. In contrast, even light activities such as standing or walking around at home (which probably reflects household work) and brisk walking were associated with a significantly lower risk of obesity and diabetes. Our study suggests that 30% of obesity cases and 43% of type 2 diabetes cases can be potentially prevented by following a relatively active lifestyle (<10 h/wk TV watching and ≥30 min/d of brisk walking).

Our results are consistent with previous studies of TV watching in relation to obesity and weight gain in children^4,6- 7 and adults.^19- 22 They are also consistent with our previous analysis showing a positive association between prolonged TV watching and type 2 diabetes among men in the Health Professionals Follow-up Study.⁵ Our findings extend the literature showing that moderate-intensity activity such as brisk walking is strongly protective against type 2 diabetes^15,23- 29 and indicate a continuum in the relationship between physical activity levels and obesity and diabetes risk, and TV watching has the highest risk for obesity and diabetes among several sedentary behaviors.

There are at least 3 potential mechanisms for the observed positive association between TV watching and obesity and diabetes risk.³⁰ First, TV watching typically displaces physical activity and thus reduces energy expenditure.^20- 21 In our study, women who spent more time watching TV tended to exercise less, but the effects of TV watching and exercise on the development of obesity and diabetes were largely independent. Second, TV watching results in increased food and total energy intake because individuals tend to eat while watching TV despite their low physical activity levels.^3- 5 Also, participants who spent more time watching TV tended to follow an unhealthy eating pattern.⁵ Such an eating pattern is directly related to commercial advertisements and food cues appearing on TV³¹ and has been associated with risk of obesity and diabetes.³² Another potential mechanism is that TV viewing results in lower energy expenditure compared with other sedentary activities such as sewing, reading, writing, and driving a car.² The combinations of these factors may explain our findings that TV watching is more strongly associated with obesity than other sedentary behaviors.

Our findings could have important public health implications. The prevalence of obesity and type 2 diabetes has increased dramatically in the past several decades in the United States.³³ Although leisure-time physical activity levels are generally low in the United States, there is no evidence that recreational physical activity has declined in recent decades. However, sedentary lifestyle has become more prevalent and pervasive as reflected by the large numbers of TV sets, VCRs, and remote controls per household and increasing time spent watching TV in the past several decades.¹ We speculated that increasing sedentary behaviors, especially TV watching, may have contributed to the obesity epidemic in the United States. Given the strong relationship observed between sedentary lifestyle and obesity and diabetes risk, public health campaigns to reduce obesity and diabetes should not only promote increasing exercise levels, but also decreasing sedentary behaviors, especially prolonged TV watching. Substantial health benefits can be gained by even light to moderate activity such as doing household chores and by engaging in simple and convenient activities such as walking.

The major strengths of this study include its large sample size, prospective design, and detailed measures of physical activity, sedentary behaviors, and a multitude of dietary and nondietary covariates. However, the observational nature of this study cannot prove a causal relationship between TV watching and obesity because of the potential reciprocal relationship between TV watching behavior and obesity. This issue could be addressed in randomized clinical trials among adults. However, school-based intervention studies have already demonstrated that reductions in TV watching were effective in preventing obesity among children.⁸

In conclusion, our data provide strong evidence that sedentary behaviors, especially prolonged TV watching, are directly related to obesity and diabetes risk. In contrast, even light- to moderate-intensity activity substantially lowers the risk. While these findings lend further support to current guidelines that promote physical activity, they also suggest the importance of reducing sedentary behaviors in the prevention of obesity and type 2 diabetes.