top of page
Folie Conditions

Now I turn to look at this delusional type, as it involves a small social grouping between people who are often related to each other. It seems to involve processes where somehow a social group has become isolated, inward looking and the beliefs the individuals have, have a fixity to them that is resistant to treatment and often can only be broken upon separating the group members. Maintaining our social group has been a vital source of survival in the past, to the extent that strong efforts are made to keep the group together or to find another group. Would an investigation of the folie delusions provide some insight here? Are folie delusions another evolutionary way to maintain social groupings in the face of adversity?

These types of delusions in essence involve group members sharing delusional beliefs. The number of people involved in the group is often two, but there are cases where three, four, or even five people can share the same delusional beliefs. These individuals are said to have exactly the same delusional content, where people are closely related, etc. From a clinical viewpoint, there seems to be five types (Kusum 2009, Soni 1974).

  • Folie impose – the delusional person ‘transmits’ the delusions to another.

  • Folie simultanée – simultaneous psychosis in two vulnerable individuals, who have been associated with each other for a while. Separation is often ineffective here.

  • Folie communiqué – delusions transferred after initial resistance and are maintained even on separation and both are often psychotic.

  • Folie induite – delusions transferred to another psychotic individual.

  • Folie a beaucoup was described by Brusell and involves a large number of people becoming delusional (Dewhurst 1957).

Some authors feel that the secondary patient, the one who becomes delusional, must have some pre-existing condition that makes them more vulnerable, e.g. intellectual disabilities, dependent personality disorder. Others say that there needs to be a third person who does the ‘transmitting’ to the secondary patient. Sometimes it comes with other delusions and has been found in patients with schizophrenia, delusional disorder, psychotic depression and dementia. Still others have attracted a diagnosis of OCD, somatoform or dissociation (Kusum 2009). One case of an elderly lady with schizophrenia, who was taken in by a younger male of a kind nature. After a few months he began to follow the lady on her travels, e.g. collecting rubbish, holding water bottles for her. His self-care reduced and he became shabby. He developed paranoid delusions and gradually she became his only social contact and he even took her to work and sought her advice. Initially he shared her delusions but later he developed his own (Kusum 2009). A case of folie a trois, involving the grandmother, mother and the daughter, all of whom had developed schizophrenia. The second case described how when the delusional pair were rehoused, the delusions resurfaced against new neighbours. A secondary in a fourth case, drunk alcohol with the initial aim of allowing himself to believe his wife’s delusions. In the fifth case, the secondary was a 14 year old boy who was emotionally close to the primary, his mother. For the primary in the sixth case, the secondary was intellectually inferior to her. In the seventh case, the primary became psychotic, possibly in the context of underlying disease and she and the primary had two episodes of folie a deux over the years (Soni 1974). There was a case of folie a cinq, in five with evidence of borderline traits (Mirabzadeh 2007). It has been reported in recurrent depressive disorder, in this report the patient felt vermin was moving under her skin, which resolved with antidepressants and ECT (Mahler 2008). One case of shared paranoid delusions in the context of eating disorder found that the patient’s mother was the primary (Baweja 2013).

Two reviews of cases of folie a deux have been carried out, from 1942 – 1993 and 1993 – 2005.  The mean ages of the primaries were 48.1 and 52.7 respectively and the secondary’s was 42.9 and 45.9 years respectively. The majority were female and over 90% were from the same family, with half being married, a quarter were sisters and about 70% involved the parent and child. The majority of cases displayed social isolation and were together for a mean of 72.98 months. Interestingly the primaries often had delusional disorder, schizophrenia and affective disorders, the secondary’s often were diagnosed with folie a deux (Arnone 2006). A review of patients from the Japanese literature found that there were 97 dominant patients, 151 submissive people, with about 75% of cases involving 2 people and over 90% involved family members. The relationships most often affected were those involving a mother and her child, married people, followed by siblings. The most common delusion was paranoia, followed by religious delusions (Kashiwase 1997). Another review did not find any age differences between the inducer and induced patients. Nor did they find a gender bias. About 30% of cases involved married people, siblings and parent-child relationships. Interestingly about 20% involved identical twins. Paranoid and then grandiose delusions were the commonest delusions found (Silveria 1995).

In terms of physical disorders, it has been associated with lymphoedema, in one report the lady was living in an area rich in bancrofitian filariasis (Dreyer 2008). Many patients have been found to have physical conditions, but it was not possible to attribute some causal connection (Balducci 2017). There are reports in blind patients (Bidaki 2017), in SLE (Caribe 2013), dementia (Draper 1990), with stimulant use (Hill 2001), with multiple sclerosis (Malik 2000). One case involved deafness in the secondary patient (Soni 1974). It has been found with downs syndrome (Meakin 1987). There have been reports of folie a quatre (Kathiresan 2018), in a family with a history of Huntington’s chorea (Sims 1977).

Delusional infestation has often been the subject of folie conditions. A 44 year old man had six years of itchiness without attracting a dermatological diagnosis. His wife, 16 and 18 year old daughters had similar problems, after initially resisting. It started after the wife developed delusions that the neighbours had inflicted a worm infestation on her and her family. She was diagnosed with paranoid disorder and her family with shared psychotic disorder (Daniel 2004). Another case involved a man, his wife and three children. The man had a diagnosis of paranoid schizophrenia, antisocial personality disorder and alcoholism. He had prevented his family from leaving the house and they too started to believe in his paranoid delusions. While the father was in hospital, the family recovered somewhat but the whole family were eventually lost to follow up (Srivastava 2010). Still another involved the delusional belief that outsiders had deliberately infested them in some way, with the mother dominating the husband and children (Daniel 2004; Sugahama 2000). Another family of 4 developed paranoid delusions after the primary, the daughter, developed delusions of infestation while in university (Katsigiannopoylos 2006). The coexistence of folie a deux and delusional parasitosis is about 5 to 15%. A 55 year old woman felt she was infested by bugs, and claimed to be seeing them all around her. Her son who helped her to look after an uncle and lived with her, felt that the bugs had transferred to him. They both had itchy sensations even though they stayed in hotels at times (Kim 2003). Two weeks after the primary developed it, her husband did also. She cleaned everywhere and used numerous chemicals to clean her house (Matthews 2005). case of shared paranoid delusions in the context of eating disorder found that the patient’s mother was the primary (Baweja 2013).

The previous section mentioned regaining of social resources as a prelude to what we call delusional behaviour. There is evidence that the sharing of such beliefs is very common, with the groups closing ranks, providing protection, etc. A pivotal figure arises, proclaiming some future happening which results in people joining the cause (not all of whom are mentally ill), they come to believe in the same beliefs, which may then elicit benefits to the primary. Sometimes the primary is placed into a leadership position and some have speculated that this may help in dividing ancient social groups into smaller units to help protect society, other feel that it is the gaining of social benefit which is of importance (Hagen 1995, Stevens 2000).

Typically the finding of folie a deux occurs between people who are socially close, such as husbands and wives (Ahmad 2009). When it comes to families and the folie conditions, the families may be isolated, with dependent and ambivalent relationships between them. They may experience crises more often, sometimes with violence. Within such a family network, there tends to be more stability, with one member being dominant to the other less. (Glassman 1987). The isolation though does not necessarily mean that people are physically isolated from people, it could simply be that they are isolated due to language or religious reasons (Majumder 2014).

Occasionally there have been unfortunate cases whereby the presence of folie a deux leads to the death of a child (Kraya 1994). A mother with schizoaffective disorder with previous history of religious delusions and eventually her husband came to believe the same religious beliefs. They believed in demon possession, the sovereignty of god, that prayer will raise the dead, conspiracies theories and alternative medication, all of which were at odds with the church they attended (Rahman 2013). There has also been some reports where we see folie a deux by proxy, i.e. the inducer complains that the secondary has issues (Hussain 2018).

Treatment of the folie Conditions involves separation of the pair often resolves the delusion in the secondary individual, though in folie communiqué this may not happen (Kusum 2009). ECT has been used successfully (Adler 1946).

Univariate Analysis of Folie Dataset

So I had collected case reports of patients with folie delusions and recorded information on various variables, e.g. age, gender, etc. Cases were found via Google Scholar, Pubmed and other resources and information was entered into a STATA datasheet. I did not have any specific hypothesis that I wanted to test, rather I just wanted to explore the data.

Age – the mean age was 38.7 years, SD was 17.8. The youngest person in the group was aged 3 and the oldest was aged 89. The histogram of ages is shown below.

Picture 1.png

So it seems that most people were between the ages of 20 to 60, when they were written up in cases. The normal probability plot shows a reasonable fit to a normal distribution.

Picture 2.png

Data on the duration of the Folie conditions were present in 87 people only.

Picture 3.png

Histogram showing duration of folie delusions in years.

It appears to follow a geometric distribution, with most people having less than 5 years of being delusional. The median duration was 1.5 years, the mean duration was 4.4 years with SD being 6.6 years.

Below is the basic descriptive analysis of the folie dataset.

Glancing at the above, we see that 360 cases were either a sister / daughter or a mother to the primary. With other male family roles appearing less frequently. In terms of the binary variables, it is reasonable to assume that a null distribution of 0.5 would be appropriate, i.e. that there would be an equal chance of being an inducer or a secondary, or that there is an equal chance of the patient being a male or female, etc. An analysis found that all binary variables were significantly different from the null. This may simply reflect the absence of relevant material in the case reports. 

Summary of univariate data suggest that most are middle aged, and tend to suffer from folie delusions for less than 5 years – of course, they may have continued to suffer afterwards, but they were not written up again. Most groups had only 2 people in them with females making up two thirds of the total number of people. In terms of relations to primary, sisters / daughters / mother / wives were more often seen then those who were husbands / brothers / sons or fathers to the primary. Most of the folie related delusions were paranoid in nature and there were often more than 1 type of folie delusions present. Interestingly only about 200 isolated, whilst the other did not – this seems to go against the prevailing view that such patients are isolated. Slightly less than half received treatment and about 2/3 of these responded to treatment, whether it was medications, separation or other forms of treatment. 

It seems that the vast majority of folie delusions are related to paranoia and somatic delusions. Could these have survival advantages? If the inducer is paranoid about his or her safety, than in our past when we have depended on the group for survival, than the group taking on the paranoia might have a survival advantage. The chance that harm will come from being wrong may be outweighed by the harm that could come from not anticipating an attack, etc. An interesting fact is that of the 193 people who isolated themselves, 112 were suffering from paranoia. Isolating yourself would be a good defence against being attacked. 122 suffering from somatic concerns, which related to thoughts of infestation, hypochondriasis, etc. Now the costs of missing a disease may be greater than assuming that you have the disease that the inducer has. We all know how seeing one person itch can make us itchy, so again assuming a disease is present which leads to disease related behaviour again might have survival benefits. 

79 had religious delusions and 77 had grandiose delusions. What would be the advantage of believing the inducers religious or grandiose delusions? May it simply relate to staying with the hunter, the care giver, or does it have other survival needs?? One theory of schizophrenia says that it may develop when there are too many people in the area, so that the local resources are insufficient to feed everyone and the group needs to split. So the arrival of a ‘special person’, someone who is communicating with gods, or has special powers, tends to result in that person developing followers. When the social order is strained, they and their group get kicked. This may have benefits to both those who stay and those that go, by reducing competition for resources. To answer these questions we would need to know a lot more of the patient’s employment history, the threats that are perceived to be in the community by the group [as opposed to well others], religious affiliations, etc. before we can provide answers here. 

Also it is interesting that there seems to low levels of jealousy (n=4) and erotomania (n=8) being represented as folie delusions. Now if the group are genetically related it does not make sense to have sexual relations with them. But if the group contains spouses or other unrelated individuals, then this might not hold. Of course if the inducer feels jealous or erotomania, it does not make sense for the submissive to also feel jealousy or erotomania, especially if they are the target of the jealousy or erotomania. As these emotions / delusions are related to various mate retention behaviours, why would the submissive admit that they have jealousy over their [not others] behaviours or say that they love the inducer when they don’t. So sexual selection might not relate to folie conditions. 

From an empirical point of view, it would make sense that most of the groups only have 2 members in them. After all, it becomes harder and harder to convince more and more people that what you are saying is correct. Though we do see that there are some large groups, but the probability of seeing them reduces as the group size increases. Now it would also make sense that larger groups are more advantageous from a survival point of view. There would be more people to protect from outsiders, to gather food, to provide care for children, etc. But on the other hand there are more people in need of protection, say from disease, from attack, etc. Would a larger folie group reflect this in some way? E.g. are there difference in the types of delusions across the different sizes. Do they reflect other processes that are not involved in smaller groups? 

According to most of the psychiatric literature about folie conditions, those members of folie groups tend to be together for some time. The so called submissive individual is said to one that tends to be younger and is often female. Relative isolation of the group from others may allow the dominant person to lead as it were, further enhanced by mistrusting those who are in the immediate surroundings. This isolation is enhanced if there are language barriers, etc as well as being enhanced by physical isolation from others. The dominant person is said to be intelligent, is the source of pleasure and because of the isolation, the dominated cannot use social influences that run counter to the dominant individual. Interestingly some have remarked that this situation may allow the submissive person to actually control the dominant person, a passive aggressive style as it were (Lozzi 1992). 

Some have remarked on other similar situations, such as brainwashing, hypnosis or therapy. There may be an initial reduction in resistance due to isolation, followed by identifying with the dominant / aggressive person which results in safety from the aggression. Then a final imposing of a contextual component to explain what is happening. The dominated person may also be more suggestible than others (Lozzi 1992). 

Formation of group norms eventually result in what is considered acceptable in that group and taking them on may reflect a desire to be accepted. Some have highlighted the similarity with groupthink, where people in a group make a choice because of a need for universal agreement. This involves censoring yourself, believing in the group, criticising those who expound alternative ideas, etc. (Salganik 2006).

Bivariate Analysis of Folie Dataset

Table showing pairwise correlation coefficients between the variables that remain significant after Bonferroni correction.

Additionally is seems that the correlation between duration of folie delusions was positively related to the size of the folie group. This means that the delusions tended to be present for a longer period of time if the size of the folie group was larger. This could mean that the longer the group exists, the greater the chance of recruiting others. Or it could mean that larger delusional groups are more resistance to outside ‘corrective’ measures.

Age – I ran 20 tests looking at whether age was different across the levels of the categorical variables. As a result I adjusted my p-value to indicate significance, i.e.

  • Inducers were significantly older than secondaries (35 v 44 yr) with t=-5.6, df=442, p<0.0001. Does this simply mean that they lived long enough to become delusional and went searching for others who believed them? Does it reflect that in our past, we may have tended to be alone and would this then be a way to recreate a social group?

  • Those in relationships were also older (38 v 45), t=-3.7, df=328, p=0.0002. Again this may reflect the fact that the older you are, the more likely you are to be in a relationship.

  • Those with hallucinations were also slightly older (t=-2.8, p=0.006). Could this reflect the fact they they’ve lived long enough to develop brain pathology? Those who received treatment tended to be older (30 v 39.5, t=-3.2, p=0.0016). This may reflect that fact that services may have had more time to get them into treatment.


When looking at ages across the different relationships that the secondaries had to there inducers. Brothers were the baseline group and compared to them, all others tended to be older, except for daughters and sons. The only group that was not significant was the ‘other’ group, but this contained many people who were friends, distance relations, etc, as opposed to a single familial relationship type.

Mean age of patients, broken down by familial relation to inducer.

Gender - When it came to gender if appeared that females were more likely to submissive compared with males (chi2 =4.59, p=0.032, OR 0.67 05% CI 0.45 – 0.98. Females also appeared more often when we looked at each folie group size (chi2 =6.03, df=10, p=0.813). They were more likely not to work (chi2 =29.5, df=1, p<0.001, OR=5.6 95% CI 2.8-11.3). Interestingly females were also less likely to have a family psychiatric history (chi2 =10.2, p=0.001, OR=0.3, 95% CI 0.1-0.7). Also they were more likely to respond to get treatment, though p=0.059.

Interestingly submissives were more likely to get treatment (chi2=7.7, p-0.005) and to respond to treatment (chi2 =4.3, p=0.037). When it came those with children, they were more likely to be in a relationship.

When I looked at each group member relationship to the primary case, I found:

Size of 

folie                                         place    in family

group   brother daughter          father   husband           mother other    sister    son       wife      Total


2          19        38                    3          51                    47        22        80        20        51        331

3          11        17                    6          10                    21        11        25        14        9          124

4          4          18                    8          4                      11        2          6          9          6          68

5          2          18                    5          2                      7          5          1          3          2          45

6          8          5                      2          8                      4          8          38        9          9          91


Total    44        96        24        75        90        48        150      55        77        659

When combining siblings into a group, mother and father into parents, daughters and sons in children, the following table was obtained:

Size of

folie                 place    in family

group   kids      other    parent  partner sibs      Total


2          58        22        50        102      99        331

3          31        11        27        19        36        124

4          27        2          19        10        10        68

5          21        5          12        4          3          45

6          14        8          6          17        46        91


Total    151      48        114      152      194      659

In general a sizeable proportion were in groups of size 2, makes sense as groups of size 2 were the commonest. Siblings appeared to be the commonest group, followed by children and partners, then parents and finally others. Again this could simply reflect the fact that we are physically closer to such individuals without needing to evoke an evolutionary role. Of course in our evolutionary past we would have tended to around family members and staying with them was essential for survival. And as we are more likely to be helped by family as opposed to strangers, we could argue that a family sharing the same ideas may serve as a survival tacit. Now would this extend to sharing the same delusional beliefs??

Multivariable Analysis of Folie Group Size

None of the multi-regression analysis run after multiple imputation showed any significant variables with different dependant variables.


Maintaining a social grouping (as folie delusions could be argued to do), may be important in terms of pathogen avoidance, i.e. if we stay together and isolated, we are less likely to suffer from illness. This might explain that high incidence of paranoid delusions and delusions of infestations. Paranoia keeps us away from ‘danger’ and this could be reinforced if we say we are infested. In that we restrict ourselves as we blame others for our predicament. If this keeps the ancient social group alive, it may mean that being in a group is more important than the advantages that can be found out of the group. This probability would change with time. Could we see a greater incident of folie delusions (and possibly those with paranoia / infestation) when there is famine / war / restriction of resources / surrounding infections? Would this mean that in the modern world, we would see this people in deprived areas?

There was a lot of missing information when cases were presented in the literature. Also variables that might potentially be ok importance, were not measured or mentioned. I would recommend that further studies are run.

bottom of page