Physical and neurobiological conditions

Physical and neurobiological conditions associated with pica are listed in Box 1. Much of the literature on pica is based on case reports, case series and retrospective case-note reviews. There is a dearth of high-quality evidence in the area and it is therefore not possible to attribute causation to associated factors. The causation of pica is therefore unclear and probably multifactorial (McNaughten Reference McNaughten, Bourke and Thompson2017). There have been a number of case reports of pica occurring after gastric bypass surgery and it is suggested that it is linked to poor absorption of iron and iron deficiency anaemia (Tabaac Reference Tabaac and Tabaac2015). However, several studies indicate that the incidence of dietary deficiencies is no more common in children with pica than in those without (Barrett Reference Barrett2008). Some case reports found that pica continued after the deficiencies were corrected (Herguner Reference Herguner, Ozyldrm and Tandr2008; McNaughten Reference McNaughten, Bourke and Thompson2017), whereas others reported that pica resolved completely if, for example, an iron deficiency was corrected (Sinha Reference Sinha and Mallick2010). The role of nutritional deficiencies in pica and whether they are cause or consequence is therefore unclear.

The role of dental pain in causing pica is also unclear. There has been one case report of a child with Pallister–Killian syndrome (a rare chromosomal disorder with features including, among other problems, distinctive dysmorphic facial features and intellectual disability) in whom dental pain was felt to be one of the factors contributing to his pica (Ayanouglou Reference Ayanouglou, Pernon and Dubois2011). However, a study comparing the dental health of children with autism compared with that of controls found no association between the prevalence and severity of caries and additional diagnoses (intellectual disability, cerebral palsy, self-injurious behaviour or pica) (Loo Reference Loo, Graham and Hughes2008).

In terms of medication being linked to pica, there has been one case report of coprophagic asphyxiation in a woman with severe intellectual disability (Erickson Reference Erickson, Gambrell and Fulcher2017). She had high blood olanzapine levels at autopsy and it was postulated that her antipsychotic was a driver of pica. It has been hypothesised that pica is associated with dialysis, possibly related to anaemia, electrolyte derangements or the stress of end-stage renal disease (Katsoufis Reference Katsoufis, Kertis and McCullough2012). There are several possible explanations for the link between sickle cell disease (SCD) and pica. It has been proposed that brain injury resulting from silent infarction can occur in children with SCD, and this could be associated with behavioural difficulties as well as cognitive deficits contributing to the development of pica (El-Mekkawy Reference El-Mekkawy, Hosam and Rabah2014). Also, children with SCD seem to have a greater risk for developing depression compared with those with other chronic conditions such as diabetes, cancer and cystic fibrosis, which may then be linked to pica (El-Mekkawy Reference El-Mekkawy, Hosam and Rabah2014).

Mental disorders

Mental/developmental disorders associated with pica are listed in Box 2. In people with intellectual disability and autism, pica was previously thought to result from their inability to differentiate between food and non-food but is now thought to be likely due to learned behaviours (McNaughten Reference McNaughten, Bourke and Thompson2017). In one study, pica in adults with intellectual disability was associated with severely impaired expressive communication abilities, younger age, autism, poorer cognitive performance, greater aggression and male gender; it was found to be less common in those who had Down syndrome and who did not have impaired self-care skills (Ashworth Reference Ashworth, Hirdes and Martin2009).

Fetal alcohol spectrum disorder has been associated with pica (Wygant Reference Wygant and Cohle2019). It is thought that alcohol exposure in utero adversely affects the development of the myenteric plexus, increasing the risk of pseudo-obstruction, which can be exacerbated by pica with potentially fatal consequences such as asphyxia (Wygant Reference Wygant and Cohle2019). It has been suggested that the link between attention-deficit hyperactivity disorder (ADHD) and pica could be due to poor impulse control or dysfunctions in the dopaminergic system (Herguner Reference Herguner and Herguner2010).

As regards neuropsychiatric associations of pica, a study comparing adults with acquired brain injury and no pre-existing neuropsychiatric, developmental or degenerative diseases who demonstrated hyperorality alone with those who demonstrated pica found that those with pica had more severe semantic memory deficits and fewer frontal release signs. Neuroimaging revealed that those with pica all had a lesion in the posterior part of the middle temporal gyrus. It was therefore proposed that in this patient group semantic memory deficits (i.e. a difficulty recognising objects) were associated with pica (Funayama Reference Funayama, Muramatsu and Koreki2017). This is different from the suggested primary mechanism in people with intellectual disability and underlines that pica has many possible underlying aetiologies.

Social/environmental factors

Some see pica as a stand-alone disorder triggered and maintained by operant/environmental factors mainly to seek engagement from others, escape/avoid a disliked situation, as a display of non-social behaviour such as stealing or refusing food or, most commonlythe person enjoys eating the pica item so the behaviour is self-reinforcing (Matson Reference Matson, Hattier and Belva2011) (Box 3).

General treatment

In young children and pregnant women, pica usually remits spontaneously. Management of persistent pica should involve a multidisciplinary approach including speech and language therapists, occupational therapists, physicians, social workers, dieticians, psychologists and dentists (McNaughten Reference McNaughten, Bourke and Thompson2017). It is important to treat any associated physical conditions first, such as nutritional deficiencies, toxicity, parasitic infections, obstruction and dental problems. Thereafter, interventions should be primarily behavioural in nature (Matson Reference Matson, Hattier and Belva2011). There is only limited evidence to support pharmacological interventions in the management of pica (Matson Reference Matson, Hattier and Belva2013).

General treatment of pica includes:

• • education for parents/carers/patients about pica and good nutritional practices (McNaughten Reference McNaughten, Bourke and Thompson2017)

• • making parents/carers/patients aware of the potential severity of the condition (McNaughten Reference McNaughten, Bourke and Thompson2017)

• • supervision of the patient

• • removal of pica items from the environment wherever possible (Barrett Reference Barrett2008)

• • an assessment of the patient's capacity to decide what they eat and drink, which will inform the extent to which restrictive practices can be used to manage the pica (Hayes Reference Hayes and Boulter2019).

Behavioural treatment and physical intervention

Possible behavioural interventions for pica are listed in Box 7. Aversive methods for reducing pica have been described in the literature but in today's practice these are unlikely to be considered ethical and so are not included in this article. Most studies of behavioural treatments for pica are small (Call Reference Call, Simmons and Mevers2015), not of robust research design and many are old, so there is limited strong evidence about the best treatment. It is therefore difficult to make any strong recommendations about which of the interventions described below is best. Furthermore, as described above, the possible causes of pica are diverse, so treatment is best tailored to the features of the individual. Therefore, the information in this section is provided to inform clinicians of possible interventions rather than to recommend any particular approach.

Several studies have found benefit in positive reinforcement (i.e. if pica is not attempted or an alternative behaviour is carried out) (Matson Reference Matson, Hattier and Belva2013). Pica exchange involves teaching the person to exchange the pica item for another item. In a case study of two boys (Kern Reference Kern, Starosta and Adelman2006) with developmental disabilities, they were taught to hand over a particular pica item for a potato crisp (previously identified as a preferred food item). The exchange started on 1:1 ratio, after which the ratio was reduced to 2:1, then to 3:1. Next, a delay in response was added and the exchange was subsequently extended to other pica items. Pica levels were reduced to near zero. In another case study using pica exchange for preferred food items (such as chocolate, biscuits, coffee and diet soft drink), the distance the individual (a 72-year-old woman) had to move to do the exchange was gradually increased until the exchanger was in the next room. This reduced pica to near zero but took 2 years (Carter Reference Carter2009). There are issues about ensuring the item being exchanged for is not too unhealthy, and indeed unhealthy items were used in both these cases.

Discrimination training involves teaching the person the difference between edible and non-edible substances (McNaughten Reference McNaughten, Bourke and Thompson2017). Environmental enrichment can involve providing alternative stimulating food (e.g. highly spiced and flavoured) and also offering stimulating activities, in combination with non-contingent reinforcement (Matson Reference Matson, Hattier and Belva2013). A number of studies have found this helpful (Matson Reference Matson, Hattier and Belva2013). More broadly speaking, given the influence of social/environmental factors on pica and in behavioural difficulties in general, providing an individual with meaningful activities in a safe environment is likely to be important in most, if not all, people with pica. One study found response blocking with redirection to be successful. This involved removing pica targets from the hands or gently blocking the mouth, combined with redirection to a preferred food item or activity (Liu Reference Liu, Pesch and Lumeng2015). Another study found response blocking helpful only if it was combined with an alternative reinforcer (e.g. popcorn) and that otherwise it led to aggression (Hagopian Reference Hagopian and Adelinis2001). Some small, older studies from the 1970s, 1980s and 1990s found restraint helpful in reducing pica, but other studies found alternative methods to be as effective, such as providing staff interaction when the function of the pica was to seek engagement (Matson Reference Matson, Hattier and Belva2013). In some individuals pica may be related to sensory processing problems and input from occupational therapy may be helpful. For example, if the person has oral hyposensitivity, an occupational therapist would be able to recommend safer yet meaningful ways to fulfil this sensory need, such as ‘chewellery’ or similar-textured food items to avoid pica or substitute for the pica item.

As mentioned above, the existing studies into behavioural interventions for pica have small samples (Hagopian Reference Hagopian, Rooker and Rolider2011; Call Reference Call, Simmons and Mevers2015), but it is often thought that behavioural interventions can be very effective (McNaughten Reference McNaughten, Bourke and Thompson2017). One review paper found that 80.7% of identified studies reduced pica by more than 90% relative to baseline (Hagopian Reference Hagopian, Rooker and Rolider2011). Some argue, however, that a 90% reduction is not enough, given the potentially fatal consequences of pica (Williams Reference Williams and McAdam2012). Many studies did not provide results after long-term follow-up and there may be a problem with publication bias. Also, in many studies the interventions were carried out in a single, often clinical or institutional, setting rather than a natural setting. Although this is necessary for research, it makes it more difficult to extrapolate the findings to community settings.

Furthermore, all the interventions are intensive and require significant resources. A retrospective follow-up study over 9 years of 41 residents of an institution with intellectual disability and pica who had received both various forms of behavioural management and comprehensive management of the environment (i.e. limiting access to items small enough to be eaten) found that 85% showed a 75–100% reduction in pica. However, it was noted that, owing to the measures put in place to manage the pica, these residents led a more constricted life, as they did not go out much or have as many external activities and social opportunities as other residents (Williams Reference Williams, Kirkpatrick-Sanchez and Enzinna2009).

Given the potentially fatal consequences of pica, restrictive practices may be warranted. Judgement therefore needs to be made on an individual basis about whether the level of risk justifies the level of restriction; this is particularly relevant when considering the use of physical interventions such as those listed in Box 8. It is also essential to ensure that any use of restrictive practices is authorised by the relevant legislative framework of the jurisdiction in which the patient lives (e.g. the Adults with Incapacity (Scotland) Act or the Deprivation of Liberty Safeguards under the Mental Capacity Act 2005 for England and Wales).

Pharmacological treatment

As with most challenging behaviours, pharmacological treatment is rarely indicated (McNaughten Reference McNaughten, Bourke and Thompson2017), unless it is aimed at treating underlying mental illness. Studies and case reports of pharmacological treatments for pica in people with intellectual disability are summarised in Table 1. As with behavioural interventions for pica, the evidence base is poor and tends to be limited to a small number of uncontrolled case reports/series.

TABLE 1 Pharmacological treatments reported in the literature for pica in people with intellectual disability

Notably the studies involving individuals with comorbid ADHD have suggested a role for dopamine in pica. In the first case series mentioned in Table 1, there was a suggestion that diminished dopamine transmission had a role in the maintenance and exacerbation of pica, hence the worsening on thioridazine (Singh Reference Singh, Ellis and Crews1994). In the other case report it was postulated that stimulants improved pica by increasing dopamine levels and reducing impulsivity (Herguner Reference Herguner and Herguner2010).

There have been case reports of the successful use of other psychotropics in pica, such as asenapine in a woman with anxiety, depression and personality disorders (Tabaac Reference Tabaac and Tabaac2015), selective serotonin reuptake inhibitors (SSRIs) for pica associated with obsessive–compulsive disorder (OCD) (Gundogar Reference Gundogar, Baspinar Demir and Eren2003; Herguner Reference Herguner, Ozyldrm and Tandr2008) and venlafaxine for pica associated with depression (Mehra Reference Mehra, Sharma and Grover2018), but these studies were carried out in patients without intellectual disability so the findings may not be relevant to this patient group.