Funding from the Australian National Health and Medical Research Council (Program Grant 222708) and a grant-in-aid from Pfizer International.

We examined links between chocolate craving in people who are depressed and both personality style and atypical depressive symptoms, with a web-based questionnaire completed by nearly 3000 individuals reporting clinical depression.

Chocolate was craved by half of the respondents (more so by women), judged as beneficial for depression, anxiety and irritability, and associated specifically with personality facets encompassed by the higher-order construct of neuroticism.

The simple question of depression-associated chocolate craving appeared an efficient discriminator of DSM–IV atypical depression symptoms.

We examined links between chocolate craving in people who are depressed and both personality style and atypical depressive symptoms, with a web-based questionnaire completed by nearly 3000 individuals reporting clinical depression. Chocolate was craved by half of the respondents (more so by women), judged as beneficial for depression, anxiety and irritability, and associated specifically with personality facets encompassed by the higher-order construct of neuroticism. The simple question of depression-associated chocolate craving appeared an efficient discriminator of DSM–IV atypical depression symptoms.

Hyperphagia is commonly reported by depressed patients and is a specific DSM–IV criterion (American Psychiatric Association, 1994) for defining ‘atypical depression’, a depressive syndrome combining symptom and personality criteria. Our earlier review (Parker et al, 2002) of atypical depression weighted the primacy of a personality style of rejection sensitivity, with several symptoms (including food cravings) more reflecting self-consolatory and possibly homoeostatic strategies. In another review (Parker et al, 2006a) we overviewed studies considering mood state effects of chocolate consumption. In this study we investigate self-reported benefits of chocolate during a depressive episode and examine for any specificity of personality style to such chocolate craving.

People accessing – for whatever reason – our mood disorder consumer information website (http://www.blackdoginstitute.org.au) were invited to participate in an online survey of lifetime treatments for a depressive episode (and without any reference to the survey involving questions on chocolate consumption). Participants provided demographic data, reported symptoms and treatments of depressive episodes, and rated any increase in 35 symptoms or coping responses, degree of increased food cravings and importance of chocolate when depressed. All completed the tiered Temperament and Personality questionnaire (Parker et al, 2006b), assessing personality constructs disposing to depression from higher-order constructs (e.g. ‘neuroticism’ and ‘extraversion’) to eight lower-order facets. As participants were recruited anonymously, no data validity check was possible.

Sample
Analyses were limited to 2692 of the 3486 respondents who were 18 years or older, living in Australia, initial survey completers, and reported depressive episodes lasting 2 weeks or longer. Their mean age was 40.0 years (range 18–77) and 70.8% were female; 73.6% had previously received an antidepressant medication and 78.3% had received counselling or psychotherapy for depressive episodes.

Results
When depressed, 1465 (54.4%) reported food cravings, with 1210 (44.9%) specifically being chocolate cravers (50.7% of the women and 30.9% of the men; {chi}2=88.3, P<0.001). Only 9.5% acknowledged alternative craved foods. Of the chocolate craver group, the 736 (60.8%) who rated chocolate’s capacity to improve their depressed mood as moderately to very important were more likely to rate it as making them feel significantly (P<0.001) less anxious ({chi}2=366.7) and less irritated ({chi}2=337.1).

Temperament and Personality questionnaire scores quantified the chocolate cravers group as having significantly (P<0.001) higher mean scores on the irritability (t=6.3), rejection sensitivity (t=5.6), anxious worrying (t=5.5), self-criticism (t=5.2) and self-focused (t=4.5) scales, all derived from the higher-order neuroticism construct. Differences were not evident on scales originating from the higher-order introversion construct, i.e. personal reserve (t=1.8), social avoidance (t=1.1) and perfectionism (t=0.5). This differential finding was confirmed by this group scoring higher on the consolidated tier 2 higher-order neuroticism scale (16.8 v. 15.4, t=6.5, P<0.001) but not on the tier 2 introversion scale (12.3 v. 11.8, t=0.8, P=0.075). A logistic regression (entering all eight personality constructs as predictors of chocolate craving status) identified irritability along with rejection sensitivity as the only two significant predictors.

Examined against DSM–IV criterion B atypical depression accessory symptoms, those identified as chocolate cravers returned higher (P<0.01) scores for appetite increase (t=21.8), weight gain (t=18.8), sensitivity to rejection (t=7.3), hypersomnia (t=5.7) and limbs feeling ‘heavy like lead’ (t=5.4).

We explored the hypothesis that atypical depression symptoms have a self-comforting role. Scores on two self-comforting items (craving ‘comfort’ foods, and ‘warming up’ behaviours such as having a hot bath) increased significantly (Table 1) and linearly (F=999.2 and F=119.7 respectively; P<0.001) with increasing number of DSM–IV accessory atypical depressive symptoms, suggesting that chocolate craving might predict atypical depression status. We therefore examined the sensitivity, specificity and overall classification rate of chocolate craving predicting numbers of atypical symptoms. Sensitivity was highest (at 76.4%) for those reporting five symptoms of atypical depression, but specificity was only 57.7%. The overall classification rates were 49.2% (for a cut-off of one or more symptoms) and 57.7% (for two or more symptoms), 65.5% (for three or more symptoms), 65.0% (for four or more symptoms) and 59.2% for all five symptoms. Thus, for three or more and four or more symptoms, the probe question successfully allocated two-thirds of all participants.

GORDON PARKER, MD, PhD, DSc and JOANNA CRAWFORD, BPsych

School of Psychiatry, University of New South Wales, and Black Dog Institute, Sydney, New South Wales, Australia

Correspondence: Professor Gordon Parker, Black Dog Institute, Prince of Wales Hospital, Randwick 2031, Sydney, Australia. Email: .(JavaScript must be enabled to view this email address)

Full text