Obesity and the relationship with positive and negative affect

Abstract

Objective:

To examine the cross-sectional association between overweight and obesity and positive and negative affect.

Method:

Participants included 273 women, aged 29–84 years, who were enrolled in the Geelong Osteoporosis Study (GOS). Weight and height were measured and overweight and obesity determined from body mass index (BMI; kg/m²) according to WHO criteria. Medical history and lifestyle exposures were assessed by questionnaire. Positive and negative affect scores were derived using the validated 20-item Positive and Negative Affect Schedule (PANAS) and categorised into tertiles.

Results:

A pattern of greater negative affect scores was observed for increasing levels of BMI. Setting normal weight as the referent category, the odds for having a negative affect score in the highest tertile were sequentially increased for women who were overweight (OR = 1.31, 95% CI: 0.72–2.40) and obese (OR = 1.95, 95% CI: 1.02–3.73). The association between obesity and increased negative affect was diminished by adjusting for physical illness (adjusted OR = 1.76, 95% CI: 0.91–3.42). These associations were not substantially influenced by positive affect score or other exposures. No association was detected between BMI categories and positive affect scores.

Conclusions:

We report data suggesting that obesity is associated with greater negative affect scores, reflecting emotions such as distress, anger, disgust, fear and shame, and that this association is attenuated by physical illness. Further investigations are now warranted to explore possible mechanistic interplay between pathological, neurobiological and psychosocial factors.

Keywords

Affect body mass index depression mood obesity

Introduction

In 2008, the World Health Organization (WHO) estimated that 1.4 billion adults were overweight, including 200 million men and 300 million women who were obese (WHO, 2013). In Australia, approximately one-third of the population is obese, and 60% exceed the healthy weight range (Pasco et al., 2012a). In parallel to the obesity epidemic, data from the United Kingdom (Collishaw et al., 2004), United States of America (Twenge et al., 2010) and Australia (Goldney et al., 2010) suggest that the prevalence of psychological distress, including depression, may also be increasing. Because both obesity and depression confer an escalated risk for chronic illness, such as cardiovascular disease and type 2 diabetes, the relationship between the conditions warrants close investigation.

The mood literature examining psychological consequences of obesity is equivocal (McElroy et al., 2004). Meta-analyses provide evidence of a link between obesity and depression (Luppino et al., 2010), with some evidence of a bi-directional relationship (Pan et al., 2012), while others have reported a null effect between increasing body mass index (BMI) and depressive symptoms (Istvan et al., 1992) and subjective wellbeing (Dierk et al., 2006). Positive affect includes pleasurable feelings such as happiness, joy, interest, excitement, contentment, enthusiasm and alertness; on the other hand, negative affect encompasses feelings that reflect unpleasant engagement, and includes distress, anger, disgust, fear and shame (Crawford and Henry, 2004; Pressman and Cohen, 2005; Watson et al., 1988). It is unclear whether positive and negative affect are opposing extremes of a continuum or whether they represent separate orthogonal factors (Pressman and Cohen, 2005). If they are independent, positive affect would not just be the absence of negative affect; similarly, negative affect would not just be the absence of positive affect. As the mechanistic link between obesity and mood remains unclear, we aimed to examine the association between overweight and obesity defined by BMI, and positive and negative affect. We further aimed to examine the role of medical comorbidity in any observed relationships between body habitus and affect.

Methods

Participants

This cross-sectional analysis utilises data generated by the Geelong Osteoporosis Study (GOS). This is an age-stratified cohort of 1494 women selected at random from electoral rolls for the Barwon Statistical Division in south-eastern Australia, and recruited for the GOS between 1994 and 1997, with a response of 77.1% (Pasco et al., 2012b). This analysis focuses on the 304 women who completed questionnaires as part of the 8-year follow-up assessment, at which the measure of positive and negative affect was administered. Twenty-eight women were excluded because their self-report documenting affect was incomplete; a further three were excluded because they were classified as underweight; therefore, data for 273 women (aged 29–84 years) were included. Dietary data were missing or incomplete for five participants. All participants provided informed, written consent. Participant anonymity has been preserved in this report. The study was approved by the Barwon Health Human Research Ethics Committee.

Data

Self-rated affect was derived from two 10-item scales that comprise the Positive and Negative Affect Schedule (PANAS) that recorded feelings and emotions over the few days prior to assessment (Watson et al., 1988). The exposure of interest was BMI. Weight and height were measured to the nearest 0.1 kg and 0.001 m; BMI was calculated as weight/height² (kg/m²) and categorised as normal (if BMI < 25.0 kg/m²), overweight (if BMI was in the range 25.0–29.9 kg/m²), or obese (if BMI ≥ 30.0 kg/m²) (World Health Organization, 1998). Cigarette smoking was recognised if practised currently and regularly at the time of assessment. Dietary intake was estimated using a validated food frequency questionnaire developed by the Cancer Council of Australia (Victoria) (Giles and Ireland, 1996); alcohol consumption was expressed as grams of ethanol per day, and daily total energy intake (EI) determined from food and alcohol combined (expressed in MJ). Basal metabolic rate (BMR) was estimated from age and weight (Schofield, 1985) and the ratio EI/BMR calculated. Habitual physical activity level was categorised as very active if the subject reported that they ‘move, walk and work energetically and participate in vigorous exercise’, otherwise they were described as sedentary. Physical illnesses, including cardiovascular disease, diabetes and cancer, and antidepressant use, were self-reported. Osteoporosis, at either the spine or femoral neck, was defined according to World Health Organization (WHO) criteria (World Health Organization Study Group, 1994) utilising thresholds from the Australian reference ranges for areal bone mineral density (BMD) (Henry et al., 2004); BMD was measured at the spine (L2–4) and femoral neck (Lunar DPX-L, software version 1.31) and the short-term precision in vivo was 0.6% and 1.6%, respectively. Area-based socioeconomic status (SES) was determined using the Socio-Economic Index for Areas (SEIFA) values based on 2001 census data from the Australian Bureau of Statistics. SEIFA values were used to assign an Index of Occupation and Education, which was categorised into quintiles according to cut-off points for the study region.

Statistics

Characteristics for participants categorised by level of BMI were compared using analysis of variance (ANOVA) for parametric continuous variables, Kruskal–Wallis for non-parametric continuous variables and the χ² test for discrete variables. Odds ratios (ORs; with 95% confidence intervals; CIs) were determined using binary logistic regression models to explore associations between BMI (the exposure of interest designated as obese, overweight or normal weight) and the likelihood of having a high positive or negative affect score (highest tertile compared to mid-and-low tertiles pooled). Age, smoking, alcohol consumption, physical activity, SES, physical illness and exposure to antidepressants were investigated as confounders and effect modifiers. Statistical analyses were performed using Minitab (version 15; Minitab Inc., State College, PA, USA) and Stata (release 9.0; StataCorp LP, College Station, TX, USA).

Results

Of the 273 women, 103 (37.7%) were classified in the normal weight category, 102 (37.4%) as overweight and 68 (24.9%) as obese (Table 1). Overweight and obese participants were sequentially older and less likely to be physically active. There was a decreased tendency for obese individuals to consume alcohol and to smoke. Individuals with normal weight were less likely to use antidepressants. There was a trend of decreasing EI/BMR with increasing BMI (Table 1) and of increasing negative affect scores with increasing BMI (Figure 1).

Table 1.

Characteristics of subjects categorised into different levels of body mass index (BMI).

	Normal BMI n = 103	Overweight n = 102	Obese n = 68	p-value
Age, years	51.6 (43.1–62.7)	55.3 (48.5–66.0)	57.0 (46.9–67.5)	0.03
Smokers	18 (17.5%)	10 (9.8%)	6 (8.8%)	0.1
Alcohol > 20 g/day	14 (13.6%)	14 (13.7%)	3 (4.4%)	0.1
Physically active	30 (29.1%)	14 (13.7%)	6 (8.8%)	0.001
Antidepressant use	4 (3.9%)	13 (12.8%)	7 (10.3%)	0.07
Physical illness	47 (45.6%)	64 (62.8%)	48 (70.6%)	0.003
Energy intake (MJ/day)	6.41 (5.31–7.93)	6.65 (5.48–7.91)	6.92 (5.47–8.91)	0.4
EI/BMR^a	1.21 (0.96–1.46)	1.15 (0.94–1.43)	1.04 (0.87–1.34)	0.1
Socioeconomic status				0.07
quintile 1 (most disadvantaged)	18 (17.5%)	22 (21.6%)	9 (13.2%)
quintile 2	23 (22.3%)	25 (24.5%)	10 (14.7%)
quintile 3	18 (17.5%)	26 (25.5%)	25 (36.8%)
quintile 4	17 (16.5%)	14 (13.7%)	13 (19.1%)
quintile 5	27 (26.2%)	15 (14.7%)	11 (16.2%)

Values are presented as median (interquartile range) or n (%).

EI/BMR: energy intake/basal metabolic rate.

Figure 1.

Boxplot showing scores for positive affect and negative affect according to body mass index (BMI) categories: normal BMI (BMI 18.5–24.9 kg/m²), overweight (BMI 25.0–29.9 kg/m²) and obese (BMI ≥ 30.0 kg/m²) (Kruskal–Wallis test, p = 0.5 and 0.01, for positive and negative affect scores, respectively). The median positive and negative affect scores and interquartile ranges are represented in each box, together with an indication of the range of scores for each category of BMI.

A total of 159 (58.2%) women were identified with physical illness (cardiovascular disease: n = 106, asthma: n = 54, osteoporosis: n = 22, diabetes: n = 16, cancer: n = 11, osteoarthritis: n = 5, chronic pain: n = 5, haemochromatosis: n = 2, migraine: n = 1, ulcerative colitis: n = 1). The presence of physical illness increased with BMI (Table 1). Compared with women with no physical illness, those with physical illness were more likely to have high negative affect scores (age-adjusted OR = 1.81, 95% CI: 1.05–3.13); no association was observed between physical illness and positive affect scores (age-adjusted OR = 0.98, 95% CI: 0.57–1.69).

Setting the normal-weight group as the reference, the odds for having a high negative affect score were greater for those who were obese (age-adjusted OR = 1.95, 95% CI: 1.02–3.73), while the point estimate for the overweight category was intermediate between that of normal weight and obese categories (age-adjusted OR = 1.31, 95% CI: 0.72–2.40). This pattern persisted after adjusting for the positive affect scores and lifestyle exposures listed in Table 1.

However, this association was attenuated by physical illness (overweight: adjusted OR = 1.23, 95% CI: 0.67–2.27; obese: OR = 1.76, 95% CI: 0.91–3.42). In this model, further adjustment for other exposures had only minor impact on the association between BMI categories and high negative affect score.

There was no association detected between BMI categories and positive affect scores. With normal weight as the reference, the age-adjusted odds for having a high positive affect score did not differ for those who were overweight (OR = 0.86, 95% CI: 0.47–1.59) or obese (OR = 1.24, 95% CI: 0.64–2.39).

Discussion

We present data that suggest that obese women are more likely to have negative affect as indicated by negative affect scores. Curiously, obesity was not associated with altered levels of positive affect. We found that negative affect scores were greater for obese women compared to those with normal weight and that physical illness explained this association. It is reasonable to assume that obesity is a risk factor for physical illness, although the direction of the relationship between obesity and physical illness in this group of women is unclear. Our findings support the notion that women who are obese may not experience less positive feelings, but may have more negative feelings such as distress, anger, disgust, fear and shame, as they are indicative of high negative affect (Crawford and Henry, 2004; Watson et al., 1988) and that this may be due, at least in part, to physical illness.

The bulk of extant literature suggests that obesity is negatively associated with psychological wellbeing (Godin et al., 2012; Heo et al., 2006; Simon et al., 2006; Williams et al., 2009). However, there are also studies that report only a weak or no association (Dierk et al., 2006; Istvan et al., 1992), or a positive association that emerges after adjusting for factors such as physical ill health (Carr et al., 2007; Jorm et al., 2003). Some of these differences may plausibly arise from methodological issues or consideration of psychosocial and physical mechanisms linking these conditions.

There are several psychosocial disadvantages to being obese. Social stigma may be a consequence of obesity (Lippa and Sanderson, 2012). Social disadvantage is more likely among obese individuals who may have poor self-esteem (Lippa and Sanderson, 2012) and perceive lower levels of social integration and support, particularly among women (Crosnoe, 2012). Social disadvantage is a risk factor shared with depression (Williams et al., 2011) and poor self-image is frequently associated with social isolation, which has clear detrimental effects on mood (Teo, 2012).

Biologically, obesity and depression share an inflammatory diathesis (Maes et al., 2011; Pasco et al., 2010a, 2010b). Physical comorbidities similarly are more common among those with obesity, as they are in those with depression (Bell et al., 2011) and, thus, inflammation associated with physical illness may underpin the link between obesity and negative affect.

Physical inactivity (Jacka et al., 2011b; Pasco et al., 2011a, 2011b) and poor quality diet (Jacka et al., 2010) are key underlying lifestyle behaviours that predispose both obesity and mood. Whilst it is suggested that food intake generally enhances mood (Schulz and Laessle, 2010), awareness of excessive food intake and emotional eating may contribute to the links between obesity and psychological distress. In our study, we did not detect differences in dietary energy intake between those who were obese, overweight and of normal weight. In this context, it is interesting to note a sequential reduction in the EI/BMR ratio observed with increasing BMI, suggesting that under-reporting of food intake, or caloric restriction, may be more common with increasing BMI. Although this pattern may be a consequence of ill-health, no difference was detected between physical illness as defined in this study and the ratio EI/BMR (data not shown). It cannot be assumed that obese individuals have nutritionally adequate diets (Pasco et al., 2009). Some data suggest that poor quality diets characterised by higher intakes of unhealthy ‘Western-type’ foods are more common among individuals with mood disorders, while better quality diets, rich in vegetables, lean meats, fish and wholegrains, are less common (Jacka et al., 2010, 2011a).

The strengths of our study were that participants were randomly selected from the population and that weight and height were measured. It has previously been demonstrated that self-reported weight is increasingly underestimated with increasing BMI (Nyholm et al., 2007), with the likely impact of misclassification in lower categories of BMI. We also acknowledge some limitations in our study: we relied on self-reports to identify lifestyle factors and exposure to medications and diseases, and there is the possibility of differential recall influenced by emotions and wellbeing. Adjustment for comorbid physical illness attenuated the association between obesity and high negative affect; however, it is possible that with a larger sample the attenuated association may have retained significance; therefore, a type II error cannot be excluded. Although not measured, we recognise that physical illness and its treatment may have led to physical inactivity and weight gain, and that the negative affect could be secondary. It is also possible that sequelae of obesity that were not measured in our study, such as social disengagement, may have contributed to the apparent detrimental effects of high BMI. Measures of depression and anxiety were not considered; such measures would be useful in elucidating links between BMI and affect. As this is a cross-sectional study, directional causality cannot be addressed. Further, the limitations of BMI as an indicator of adiposity have been demonstrated, especially for muscular body builds and the elderly (Pasco et al., 2012a), and this was not considered in this study.

In conclusion, this study supports the contention that the impact of obesity on feelings and emotions is associated with changes in negative affect rather than positive affect, and that this is mediated by comorbid physical illness. Although causality cannot be claimed, these results warrant further investigations into the mechanistic interplay between neurobiological, pathological and psychosocial factors that might underlie this association.

Footnotes

Funding

The research was funded by the National Health and Medical Research Council (NHMRC).

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

Bell

Curran

Byrne

. (2011) High incidence of obesity co-morbidities in young children: A cross-sectional study. Journal of Paediatrics and Child Health 47: 911–917.

Carr

Friedman

Jaffe

(2007) Understanding the relationship between obesity and positive and negative affect: The role of psychosocial mechanisms. Body Image 4: 165–177.

Collishaw

Maughan

Goodman

. (2004) Time trends in adolescent mental health. Journal of Child Psychology and Psychiatry 45: 1350–1362.

Crawford

Henry

(2004) The positive and negative affect schedule (PANAS): Construct validity, measurement properties and normative data in a large non-clinical sample. British Journal of Clinical Psychology 43(Pt 3): 245–265.

Crosnoe

(2012) Obesity, family instability, and socioemotional health in adolescence. Economics and Human Biology 10: 375–384.

Dierk

Conradt

Rauh

. (2006) What determines well-being in obesity? Associations with BMI, social skills, and social support. Journal of Psychosomatic Research 60: 219–227.

Giles

Ireland

(1996) Dietary Questionnaire for Epidemiological Studies (Version 2). Melbourne: Cancer Council Victoria.

Godin

Elbejjani

Kaufman

(2012) Body mass index, blood pressure, and risk of depression in the elderly: A marginal structural model. American Journal of Epidemiology 176: 204–213.

Goldney

Eckert

Hawthorne

. (2010) Changes in the prevalence of major depression in an Australian community sample between 1998 and 2008. Australian and New Zealand Journal of Psychiatry 44: 901–910.

10.

Henry

Pasco

Pocock

. (2004) Reference ranges for bone densitometers adopted Australia-wide: Geelong Osteoporosis Study. Australasian Radiology 48: 473–475.

11.

Heo

Pietrobelli

Fontaine

. (2006) Depressive mood and obesity in US adults: Comparison and moderation by sex, age, and race. International Journal of Obesity (2005) 30: 513–519.

12.

Istvan

Zavela

Weidner

(1992) Body weight and psychological distress in NHANES I. International Journal of Obesity and Related Metabolic Disorders 16: 999–1003.

13.

Jacka

Pasco

Mykletun

. (2010) Association of Western and traditional diets with depression and anxiety in women. American Journal of Psychiatry 167: 305–311.

14.

Jacka

Pasco

Mykletun

. (2011a) Diet quality in bipolar disorder in a population-based sample of women. Journal of Affective Disorders 129: 332–337.

15.

Jacka

Pasco

Williams

. (2011b) Lower levels of physical activity in childhood associated with adult depression. Journal of Science and Medicine in Sport 14: 222–226.

16.

Jorm

Korten

Christensen

. (2003) Association of obesity with anxiety, depression and emotional well-being: A community survey. Australian and New Zealand Journal of Public Health 27: 434–440.

17.

Lippa

Sanderson

(2012) Impact of information about obesity genomics on the stigmatization of overweight individuals: An experimental study. Obesity (Silver Spring) 20: 2367–2376.

18.

Luppino

de Wit

Bouvy

. (2010) Overweight, obesity, and depression: A systematic review and meta-analysis of longitudinal studies. Archives of General Psychiatry 67: 220–229.

19.

McElroy

Kotwal

Malhotra

. (2004) Are mood disorders and obesity related? A review for the mental health professional. Journal of Clinical Psychiatry 65: 634–651, quiz 730.

20.

Maes

Galecki

Chang

. (2011) A review on the oxidative and nitrosative stress (O&NS) pathways in major depression and their possible contribution to the (neuro)degenerative processes in that illness. Progress in Neuro-Psychopharmacology and Biological Psychiatry 35: 676–692.

21.

Nyholm

Gullberg

Merlo

. (2007) The validity of obesity based on self-reported weight and height: Implications for population studies. Obesity (Silver Spring) 15: 197–208.

22.

Pan

Sun

Czernichow

. (2012) Bidirectional association between depression and obesity in middle-aged and older women. International Journal of Obesity (2005) 36: 595–602.

23.

Pasco

Henry

Kotowicz

. (2009) Paradoxical nutritional deficiency in overweight and obesity: The importance of nutrient density. Medical Journal of Australia 191: 44–45.

24.

Pasco

Jacka

Williams

. (2010a) Clinical implications of the cytokine hypothesis of depression: The association between use of statins and aspirin and the risk of major depression. Psychotherapy and Psychosomatics 79: 323–325.

25.

Pasco

Nicholson

Williams

. (2010b) Association of high-sensitivity C-reactive protein with de novo major depression. British Journal of Psychiatry 197: 372–377.

26.

Pasco

Jacka

Williams

. (2011a) Don’t worry, be active: Positive affect and habitual physical activity. Australian and New Zealand Journal of Psychiatry 45: 1047–1052.

27.

Pasco

Williams

Jacka

. (2011b) Habitual physical activity and the risk for depressive and anxiety disorders among older men and women. International Psychogeriatrics 23: 292–298.

28.

Pasco

Nicholson

Brennan

. (2012a) Prevalence of obesity and the relationship between the body mass index and body fat: Cross-sectional, population-based data. PLoS One 7: e29580.

29.

Pasco

Nicholson

Kotowicz

(2012b) Cohort profile: Geelong Osteoporosis Study. International Journal of Epidemiology 41: 1565–1575.

30.

Pressman

Cohen

(2005) Does positive affect influence health? Psychological Bulletin 131: 925–971.

31.

Schofield

(1985) Predicting basal metabolic rate, new standards and review of previous work. Human Nutrition. Clinical Nutrition 39(Suppl 1): 5–41.

32.

Schulz

Laessle

(2010) Associations of negative affect and eating behaviour in obese women with and without binge eating disorder. Eating and Weight Disorders 15: e287–293.

33.

Simon

Von Korff

Saunders

. (2006) Association between obesity and psychiatric disorders in the US adult population. Archives of General Psychiatry 63: 824–830.

34.

Teo

(2012) Social isolation associated with depression: A case report of hikikomori. International Journal of Social Psychiatry Epub ahead of print 8 March 2012. DOI: 10.1177/0020764012437128.

35.

Twenge

Gentile

DeWall

. (2010) Birth cohort increases in psychopathology among young Americans, 1938–2007: A cross-temporal meta-analysis of the MMPI. Clinical Psychology Review 30: 145–154.

36.

Watson

Clark

Tellegen

(1988) Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology 54: 1063–1070.

37.

Williams

Brennan

Henry

. (2011) Area-based socioeconomic status and mood disorders: Cross-sectional evidence from a cohort of randomly selected adult women. Maturitas 69: 173–178.

38.

Williams

Pasco

Henry

. (2009) Lifetime psychiatric disorders and body composition: A population-based study. Journal of Affective Disorders 118: 173–179.

39.

World Health Organization Study Group (1994) Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Technical Report Series 843. Geneva: WHO.

40.

World Health Organization (1998) Obesity: Preventing and managing the global epidemic. Report of a WHO Consultation on Obesity; Geneva, 3–5 June 1997. Geneva: WHO.

41.

World Health Organization (2013) Obesity and overweight. Fact sheet no. 311. Available at: www.who.int/mediacentre/factsheets/fs311/en/index.html (accessed 13 March 2013).