Physical Factors, Personal Characteristics, and Substance Use: Associations with Obesity

Abstract

Background

Since obesity has become a major public health problem, attention to a range of its predictors is needed. This study examined the association of physical factors, personal characteristics, and substance use with obesity in a sample (N=815) of African American and Puerto Rican young adults with a mean age of 32.

Methods

Body mass index (BMI) was calculated to assess obesity. Bivariate and multivariate logistic regression analyses were conducted.

Results

Bivariate analyses showed that protective factors such as physical activity (AOR=.82, 95% CI=.74–.91), healthy diet (AOR=.96, 95% CI=.93–.99), self control (AOR=.93, 95% CI=.87–.98), and life satisfaction (AOR=.97, 95% CI=.95–.99) were associated with a reduced probability of being obese. Marijuana use was also associated with a decreased probability of obesity (AOR=.89, 95% CI=.80–.99), but was not considered a protective factor. Risk factors such as short sleep duration (AOR=1.70, 95% CI=1.24–2.33), and depressive mood (AOR=1.05, 95% CI=1.01–1.09) were associated with an increased probability of being obese.

Conclusions

For African Americans and Puerto Ricans, programs to treat obesity should focus on increasing sleep, physical activity, and life satisfaction.

INTRODUCTION

Obesity, defined as abnormal or excessive fat accumulation that presents a health risk, has become a major public health problem (1). Obesity is associated with a number of serious illnesses such as diabetes, cardiovascular disease, and some cancers (1). Physical factors, personal factors, and substance use have been found by most, but not all, investigators to be related to obesity (2, 3–5, 6, 7–16). Alcohol use is inversely related to obesity in some research, but not all (17–20). Cannabis users are also less likely to be obese (3).

The present study is innovative in two aspects. First, this research focuses on an understudied population of urban African American and Puerto Rican young adults. According to Agyemang et al. (21), ethnic minority groups such as African Americans and Puerto Ricans, but not Asians, have higher rates of being overweight or obese than do Whites. To date, the study of obesity predictors in minority groups has been relatively uncommon. Second, the study includes numerous factors involved in obesity that might be considered in obesity treatment programs.

This study examines the association of obesity with protective factors (e.g., physical activity and life satisfaction), risk factors (e.g., short sleep duration), and substance use (e.g., marijuana use). Unlike the other studies that investigated either physical factors (22), personal factors (11, 23), or the use of substances (24), a major advantage of this study is that all of these risk and protective factors for obesity are examined at the same time in African Americans and Puerto Ricans.

Based on the literature, the hypotheses are: 1) protective factors (e.g., physical activity, healthy diet, self control, and life satisfaction) are associated with lower Body Mass Index (BMI); and 2) risk factors (e.g., short sleep duration, depressive mood, and anxiety) are associated with a higher BMI. Since the findings regarding the relationship of both alcohol use and marijuana use to obesity have been contradictory, these associations are explored, but their direction is not hypothesized.

METHOD

Participants

We examined our research questions using data from the Harlem Longitudinal Development Study (25). This study’s sample is representative of African Americans and Puerto Ricans who in 1990 attended schools serving East Harlem, NYC. Participants were 815 young adults who completed the fifth wave (T5) questionnaire between 2007 and 2010. The one pregnant participant was excluded. Of the 815 participants, 52% (n = 424) were African American and 48% (n = 391) were Puerto Rican. Forty percent (n = 324) were male. The T5 mean age was 32.6 years (SD = 1.5). The New York University School of Medicine’s Institutional Review Board approved the study. A Certificate of Confidentiality was obtained from the National Institutes of Health. We obtained informed consent from all of the participants.

Measures

All variables were measured at T5. BMI, the dependent variable, was calculated using each participant’s self reported height and weight from the formula $\text{BMI}=\frac{\mathit{weight}(lb)\times 703}{{\mathit{height}}^2(\mathit{inches})}$. The index of obesity was set at 1 when BMI≥30 and 0 otherwise (26). height ² (inches) The independent variables are listed in Table 1. Ninety-nine percent of the participants provided complete data on each of the 13 variables in the study. We used the Full Information Maximum Likelihood method to deal with the small amount of missing data (32).

Table 1.

Measures for the independent variables at T5

	T5	Alphas (no. of items)	Sample question	Answer options
Demographics	Gender	NA	Are you female or male?	1=female, 2=male
	Ethnicity	NA	Are you Hispanic?	1=African American, 2=Puerto Rican
Physical factors	Short sleep duration [27]	NA	On average, how many hours of sleep do you get per night?	1=5 hours or less, 0=otherwise
	Physical activity [28]	NA	How often do you engage in moderate physical activity (i.e. using a vacuum, bowling, washing your car, walking your dog)?	0= never, 1=seldom, 2=sometimes, 3=most days, 4=nearly every day, 5= every day
	Healthy diet [28]	.74 (4)	How often do you avoid high-fat foods (e.g., fried foods)?	Same as for physical activity
Personal factors	Depressive mood [29]	.76 (6)	You sometimes feel hopeless about the future.	1=completely false, 2=mostly false, 3= mostly true, and 4=completely true
	Anxiety [29]	.77 (3)	Over the last few years, how much were you bothered by feeling fearful?	0=not at all, 1= a little, 2= somewhat, 3=quite a bit, and 4= extremely
	Self control [30]	.65 (4)	You feel like losing your temper at people.	Same as for depressive mood
	Life satisfaction [31]	.82 (13)	During the past few years how satisfied have you been with your work?	1= completely dissatisfied, 2= somewhat dissatisfied, 3= neither satisfied nor dissatisfied, 4=somewhat satisfied, and 5=completely satisfied
Substance use	Alcohol	NA	On average, how many drinks (beer, wine, or hard liquor) did you have in the past 5 years?	0 = none, 1 = less than once a week, 2 = once a week to several times a week, 3 = 1 or 2 drinks a day, 4 = 3 or 4 drinks a day, and 5 = 5 or more drinks every day
	Marijuana	NA	On average in the past 5 years, how often have you used marijuana?	0 = never, 1 = a few times a year or less, 2 = about once a month, 3 = several times a month, and 4 = once a week or more

RESULTS

The results of the t-tests and χ² tests indicated that the obese group differed from the non-obese group on all of the factors (p<.05) except for anxiety (p>.05) and alcohol use (p>.05). The participants in the obese group had shorter sleep durations (χ²=10.22**), less physical activity (t=−3.74***), unhealthier diets (t=−2.38*), more depressive mood (t=2.23*), less satisfaction with their lives (t=−3.34***), less self control (t=−2.49*), and less frequent marijuana use (t=−2.44*) than those who were not obese. The only significant differences between African Americans and Puerto Ricans were on the healthy diet, life satisfaction, and alcohol measures. African Americans had healthier diets (t=2.6, p<.01), reported less satisfaction with their lives (t=−3.1, p<.01), and used more alcohol (t=2.2, p<.05).

Table 2 presents the results of the bivariate logistic regression analysis, controlling for gender and ethnicity, comparing the obese and non-obese participants. Individuals with short sleep duration were more likely to be obese (AOR=1.70, p<.01). More physical activity (AOR=0.82, p<.001), eating healthier food (AOR=0.96, p<.01), less depressive mood (AOR=1.05, p<.05), more self control (AOR=0.93, p<.05), greater life satisfaction (AOR=0.97, p<.001), and more frequent marijuana use (AOR=0.89, p<.05) were associated with a lower likelihood of obesity. With two exceptions, the findings of the multivariate logistic regression analysis, controlling also for the other variables in the same domain, were similar to the results of the controlled bivariate logistic regression analysis.

Table 2.

Comparisons between Obese and Non-Obese Participants Associated with Physical, Personal, and Substance Use Factors at the Mean Age of 32 (T5).

		Obesity vs. Non-obesity
		AOR (95% CI) Controlled bivariate	AOR (95% CI) Multivariate
Physical factors	Short sleep duration	1.70 ** (1.24, 2.33)	1.67 ** (1.21, 2.30)
	Physical activity	0.82 *** (0.74, 0.91)	0.83 *** (0.75, 0.93)
	Healthy diet	0.96 * (0.93, 0.99)	0.98 (0.95, 1.02)
Personal factors	Depressive mood	1.05 * (1.01, 1.09)	1.00 (0.93, 1.08)
	Anxiety	1.02 (0.96, 1.08)	0.95 (0.87, 1.05)
	Self control	0.93 * (0.87, 0.98)	0.96 (0.88, 1.06)
	Life satisfaction	0.97 *** (0.95, 0.99)	0.97 ** (0.95, 0.99)
Substance use	Alcohol	1.03 (0.88, 1.20)	1.08 (0.92, 1.27)
	Marijuana	0.89 * (0.80, 0.99)	0.88 * (0.79, 0.98)

^*p<.05,

^**p<.01,

^***p<.001

Gender and ethnicity were statistically controlled in the controlled bivariate analyses.

The variables in the same domain as well as gender and ethnicity were statistically controlled in the multivariate analyses.

AOR= Adjusted Odds Ratio, CI=Confidence Interval

DISCUSSION

The variables most significantly associated with obesity were short sleep duration, physical activity, and life satisfaction. Previous research has also found that short sleep duration is associated with obesity (4). Short sleep duration may alter thermoregulatory functions, leading to reduced energy expenditure (33). African Americans, in particular, report shorter sleep duration compared to Whites (34).

The obese participants engaged in less physical activity and had more unhealthy diets compared to the non-obese. These findings are consistent with those of several investigators (6, 35, 36). However, while an unhealthy diet had a controlled bivariate association with greater obesity, it lost significance when other physical factors were controlled. As noted by Madan et al. (35), successfully managing obesity may require physical activity.

Contrary to the results of other investigators (11,23), our findings showed that depressive mood was not related to obesity after controlling for the other personal factors, and anxiety was not related to obesity with or without these controls. This inconsistency may be due to ethnic differences between the samples (11, 23).

More life satisfaction was related to a decreased likelihood of obesity. This is consistent with findings that obese women are less likely to be satisfied with their partner relationships than non-obese women (16). Our findings extend this literature by identifying these associations in understudied groups while controlling for other personal predictors.

Other studies have found positive (20, 37) and negative (17) associations between alcohol use and obesity. Our finding of no association is partially consistent with Dallongeville’s report (18) of no association between alcohol intake and BMI in men, and an inverse correlation of alcohol consumption with BMI in women.

Although marijuana use stimulates appetite in clinical samples, usage was negatively associated with obesity, without and with alcohol use controlled. Warren et al.’s (24) results were comparable; the brain may not differentiate between food and marijuana in the activation of the dopamine reward system. From a clinical perspective, developing other reward mechanisms (i.e., physical activity) is important. The relationship between marijuana and obesity could be reciprocal. Obese persons may be less social and consequently less exposed to substance-using peers (24). Thus, greater BMI might predict less drug use.

Limitations

This study has some limitations. First is the use of self-reports for measurements of sleep duration, physical activity, diet, substance use, height, and weight. Second, our substance use measure was somewhat limited. Longitudinal studies with controls for earlier obesity and confounding variables are needed to overcome this cross-sectional study’s limits.

CONCLUSION

Despite these limitations, this study has several strengths. First, our findings are based on an understudied population (African American and Puerto Rican adults). Second, we examine the association of multiple physical factors (i.e., short sleep duration, physical activity, healthy diet), personal characteristics (i.e., depressive mood, anxiety, self control, life satisfaction), and substance use (i.e., alcohol and marijuana use) as they relate to obesity. From a clinical perspective, obesity prevention and treatment programs should focus on increasing sleep and physical activity while improving life satisfaction.