select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to glucose and aggression in married couples

A PNAS study of married couples reported that decreased blood glucose altered the way participants acted in activities such as sticking pins into a doll representing their spouse. The authors suggested low blood glucose is therefore correlated with increased aggressive behaviour between married couples.


Malte Elson, Dipl.-Psych., Research Associate, University of Münster, Germany, said

“I agree with the authors that ‘intimate partner violence affects millions of people globally’ and that indeed, poor self-control might be a potentially contributing factor. However, I do have my concerns to which extent this particular study is helpful in explaining the role of self-control (or lack thereof) in the etiology of intimate partner violence.

“The findings of this study depend on two related (and poor) measures of aggression. Neither the voodoo doll task, nor the noise blast task have been properly validated as measures of aggression, much less of violence. This lack of validation means that a larger number of needles pushed into the doll, or higher volume and duration settings may or may not actually reflect increased aggressive behavior. The fact that relationship satisfaction was not related to these measures corroborates my concerns. The study the authors cite as evidence that the noise blast task is ‘well-validated’, does not use noise blasts at all, but involves electroshocks. Whether giving or receiving electroshocks can be compared to hearing fingernails scratching on chalkboards or ambulance sirens is questionable.

“Even if it were true that pushing needles in a voodoo doll or unpleasant noise blasts would be proper measures of aggressive intents, they are certainly not measures of violence, which would involve actual harm. Should the institutional review boards that permitted this study believed harm was being inflicted to the participants, they would probably have objected to using these measures (at least one hopes).

“Therefore, I express my concerns that this study allows us to conclude that ‘low glucose levels predicted future aggressive behavior’, and strongly object to inferring this could in any way provide evidence that poor self-control is ‘one factor that contributes to intimate partner violence’. Violence was not measured in this study. This is not to say that low self-control or glucose are unrelated to intimate partner violence. However, this study does little to elucidate any potential relationship between those variables.”


Dr Andrew Przybylski, Experimental Psychologist and Research Fellow at the Oxford Internet Institute, University of Oxford, said:

“Domestic violence is a serious problem worldwide, both in the developed and developing world. This study aims to explore if there is a meaningful link between blood glucose levels and surrogate measures of aggression using a theoretical model often referred to as ‘the ego depletion model.’ This model has been applied in a range of areas (e.g. puzzle performance) but this application appears to be the most serious domain. Although the topic is important and the research goal is admirable the wider significance of the work is questionable for two reasons, one analytic, the other methodological.

“On the statistical front, this research is missing estimates of effect size. The work does a good job of reporting unstandardised slopes that relate within-person variability over time. This approach is far better than just aggregating across days but misses reporting the amount of variance accounted for by blood glucose. In other words, it is not possible to know how much of the “aggressive behaviour” blood glucose accounts for. We know some of the relations observed are statistically significant but we cannot know if it is practically significant. We would need to know this before making any kind of policy recommendation linked to blood glucose and domestic violence.

“On the methodological front, the behavioural measures used to assess “aggressive behaviour” need to be contextualised. These are proxy measures of aggression that are not validated (i.e. correlated with real-world violence) and not standardised (i.e. they can be computed a wide range of ways). Until laboratory-based and naturalistic research (like this study) take the time and effort to validate and standardise measures of “aggressive behaviours” such as these we scientists cannot make empirically-based intervention recommendations on the topic of blood glucose and domestic violence.

“It is also very surprising that there were no statistically significant trends observed linking partner gender (males more likely to be perpetrators of violence) or relationship satisfaction (lower satisfaction to higher violence) to the measures of ‘behavioural aggression’. Existing research suggests that both of these factors are robustly related to domestic violence. One would expect that these variables on the person-level should have been correlated with the outcome measure.”


Prof Jeremy Nicholson, Professor of Biological Chemistry, Imperial College London, said:

“The authors have used sticking pins in dolls and playing loud sounds to measure levels of aggression, which doesn’t of course necessarily mean they would actually be violent towards their partner.

“It is not tremendously surprising that low glucose makes people grumpy – they are hungry after all, or feel so.

“Aggression is known to be associated with poor diabetic control and alcohol consumption also causes hypoglycaemia which can contribute to aggression, although the alcohol itself is the main agent.

“I think the study is interesting as it might expose some more subtle consequences of hypoglycaemia in destabilising personal relationships – and the healthy hint is to graze eat (regular and small meals) which maintains relatively constant glucose and avoids hyperglycaemia which can link to insulin resistance.

“As for the method, it is novel and probably quite engaging for the experimental participants, but whether this complex emotional state maps on to any abstract model is debatable. The real world is always difficult to model with any certainty even when the end points are not fuzzy. If the study makes people think more rationally about the role of nutrition, lifestyle healthy living and relationships it will have a positive impact.”


‘Low glucose relates to greater aggression in married couples’ by Bushman et al. published in Proceedings of the National Academy of Sciences (PNAS) on Monday 14th April.

in this section

filter RoundUps by year

search by tag