Children's specific fears

Childhood psychopathology

The survey instrument used to produce the prevalence of clinically recognisable mental disorders among children was the Development and Well-Being Assessment (DAWBA) (Goodman et al. 2000). It was designed for use in the first national survey of child mental health in the Great Britain (Meltzer et al. 2000). When definite symptoms were identified by the structured questions, interviewers used open-ended questions and supplementary prompts to get parents and young people (aged 11 or over) to describe the problems in their own words. An abbreviated form was mailed to a teacher nominated by the family as knowing the child well. A case vignette approach was used for analysing the survey data, i.e. using clinician ratings based on a review of all the information for each child – potentially from parent, child and teacher. Diagnoses were subsequently generated based on the International Statistical Classification of Diseases and Related Health Problems – 10th Revision (World Health Organization 1993), research diagnostic criteria using the information from all available informants.

In validation studies, the DAWBA provided excellent discrimination between community and clinical samples (Goodman et al. 2000). Within the community sample, children with DAWBA diagnoses differed markedly from those without a disorder in external characteristics and prognosis, while there were high levels of agreement between the DAWBA and case notes among the clinical sample (Kendall's tau b = 0.47–0.70).

Specific childhood fears

Parents (in the majority of cases, mothers) were shown a list of the most commonly occurring fears in children and were asked whether their children had any of them.

Socio-demographic characteristics

Information on the age, sex and ethnicity of both parents and children was collected as part of the household box at the start of the interview. These questions plus those on family composition and marital status were the harmonized social survey questions recommended by the British National Statistical Office. The ethnicity question was also used in the 2001 Census in England.

Neighbourhood-level socioeconomic characteristics

The ACORN (A Classification of Regional Neighbourhoods) was used to obtain a measure of neighbourhood-level socioeconomic characteristics. It is a means of classifying areas according to various census characteristics (geographic and demographic), devised by CACI Ltd. An ACORN code is assigned to each census Enumeration District (ED) which is then copied to all postcodes within the ED. The classification consists of 56 area types. These can be collapsed into 17 higher level groups and five top level categories as used here: wealthy achievers, urban prosperity, comfortably off, moderate means and hard pressed.

Statistical analyses

To improve the representativeness of the survey, a weighting procedure was applied to the data. A weight was applied: (1) to correct for the unequal sampling probabilities of the children which arose because of the delay between selecting the area and children samples; (2) to match the age/sex/region structure of the population at the time of the survey. These data were also adjusted to take account of the missing teacher data.

Sex and age differences

Table 1 shows the extent to which girls were more likely than boys to have fears of animals [odds ratios (OR) = 2.03, 1.75–2.34]; blood, injections or injury (OR = 1.19, 1.03–1.38); elements in the natural environment (OR = 1.38, 1.10–1.72) and specific types of people, e.g. clowns, people with beards or crash helmets (OR = 1.95, 1.36–2.80).

Younger children (5–10-year-olds) were more likely to have fears of the dark (OR = 2.83, 2.31–3.47), the natural environment (OR = 1.29, 1.03–1.61), loud noises (OR = 3.31, 2.47–4.43), imaginary or supernatural beings (OR = 3.36, 2.49–4.54) and specific types of people (OR = 2.20, 1.52–3.18). However, they were less likely than older children (11–16-year-olds) to be fearful of diseases (OR = 0.76, 0.60–0.97) and enclosed places (OR = 0.47, 0.34–0.66).

Ethnicity differences

The distinction between white and non-white ethnicity does not do justice to the cultural differences between South Asian children (Indian, Pakistani and Bangladeshi) and those who are Black Caribbean and Black African. Broad categories, such as South Asian, are inadequate as there can be significant differences in religious and cultural beliefs between groups, and the impact of economic factors will be variable. Although sample numbers are quite small, comparisons with white children show quite marked differences in relation to specific fears. Three distinct patterns emerge from the data presented in Table 2.

All non-white groups were more likely than white children to have fear of animals with the highest OR for Bangladeshi children (OR = 5.32, 3.07–9.24) and in the Black Caribbean sample (OR = 3.64, 2.42–5.46). Second, Pakistani and Bangladeshi children were far more likely than white children to have fear of animals, the dark or the natural environment. Third, Bangladeshi, Pakistani and Black African children were at least three times more likely to have fears of imaginary or supernatural beings than white children did, as reported by their parents.

Family and household characteristics

The relationship between children's specific fears with family and household characteristics (family type, working status, tenure and type of neighbourhood) shows a range of significant OR (Table 3). Many of these relationships are just significant and do not show a coherent picture. The most intuitive association is type of neighbourhood being related to fears of things outside the house. For example, children living in less affluent neighbourhoods have increased odds of having fear of animals (OR = 1.22, 1.05–1.43) and fear of elements in the natural environment (OR = 1.34, 1.05–1.72).

Age differences

The increased likelihood of younger children to have fears of the dark, the natural environment, loud noises, imagined supernatural beings and specific types of people, and that of older children to be more fearful of diseases and enclosed places fit into the general proposition that age is a significant correlate of childhood fear (Sidana 1967; Gullone and King 1993; Spence & McCathie 1993; Muris et al. 1997; Muris et al. 1999; Fisher et al. 2006; Li & Morris 2007). However, there is not a consensus on the relationship between age and specific fears.

In line with our findings, Spence and McCathie (1993) found that fear of darkness showed a decrease with age among both boys and girls, and Evans and colleagues (1999) found that younger children had more fears related to strangers whereas older children have more fears related to contamination, death and fears associated with concerns of the inner city, such as burglars and assaults. Muris and colleagues (2000) also have found that fears of scary dreams pertaining imaginary beings decreases with age.

Although epidemiological research has indicated that animal fears characteristically have an early age of onset (Muris et al. 2008), we did not find any significant difference in fear of animals by age. This is consistent with the findings from Ollendick and colleagues (1985) who did not find any differences in levels and structures of fears among different age groups. Although other studies have reported differences in animal fear by age, the findings are inconsistent. Some animal fears may be developmental, and it may be that the specific animal fear changes. For example, a child may be afraid of dogs when younger but spiders in later age for the female.

Sex differences

Within our study, parents of girls were more likely to report their child's fears of animals, natural environment and strangers than parents of boys did. The greater propensity of girls to experience animal fears has been shown by Li and Morris (2007) who reported that girls reported higher levels of total fears related to small animals. Similarly, earlier studies by Sidana (1967) and King and colleagues (1994) found that girls reported significantly more fears of animals than boys did. In a study by Spence and McCathie (1993), boys reported the greatest reductions with age in fears that related to dark and unfamiliar persons.

Ethnicity

Examining the findings from other data sources can be misleading as the size and composition of minority ethnic groups in countries differs considerably.

Nevertheless, our findings are consistent with the findings from various cross-cultural studies which demonstrated that children's fears differ in nature and intensity across different ethnic groups within the same country. Culturally mediated beliefs, values and traditions play a role in the type and level of fears experienced by children (Mahat et al. 2004).

Children from some cultures are subjected to strong pressures towards compliance with strict social rules which promote development of over-controlled behaviour and discourage development of under-controlled behaviour. Mahat and colleagues (2004) suggest that cultures which sustain inhibition and obedience may contribute to increasing the level of fear among children. Elbedour and colleagues (1997) also pointed out that perhaps family closeness could either be closeness-caregiving or closeness-intrusiveness. Children, who live in a closeness-intrusive environment where independence is not fostered, may have greater fears than those who live in a more supportive environment did. Religion can also help to explain the differences in fear levels between different ethnic groups. Ingman and colleagues (1999) found that Christian children in Kenya and Nigeria reported higher levels of fear than Muslim children. More research is required to understand how cultural diversity of which ethnicity is one factor may contribute to differences in the level and type of children's fear.

Family composition

Our results regarding the relationship between family composition and child's fear are not conclusive. While children living with two parents are less likely to have fear of enclosed spaces, they are more likely to have fear of loud noises and diseases. To our knowledge no previous research has examined this issue, and with such limited information useful comment is difficult.

Socioeconomic factors

Our results suggest some differences in the presence of specific childhood fears by various measures of socioeconomic status: children characterized as having low socioeconomic status (SES) compared with a high SES have a greater propensity to fear animals, blood, the dark, doctors and dentists and using the toilet. Over half a century ago, Angelino and colleagues (1956) found that socioeconomic status was an important factor in determining the qualitative nature of fears in school children. Later studies (Sidana 1975; Crawford 1995; Erol & Sahin 1995) all found that low SES children report more fear intensity, more fear frequency and more fears than high SES children do.

Study limitations

Although the participation rate was high, about a quarter of selected households could not be contacted or refused. There is evidence from previous child psychiatric surveys that rates of psychopathology are higher among non-respondents (Fombonne 1994; Rutter et al. 1970), which leads to biased estimates of prevalence. Even though the data were weighted for non-response, it is not possible to assess the magnitude and direction of potential bias in the resulting rates.

Children under the care of local authorities were not eligible for inclusion in the household survey. Surveys carried out in 2001 and 2002 among children looked after by local authorities in the Great Britain (Meltzer et al. 2003, 2004a,b) found a rate of specific phobias of around 1.2%. However, as the proportion of children in care is about 0.5% of the total population of children, the overall rate of specific phobias and simple fears for the whole of the Great Britain would not change very much if the local authority children were included in the analysis.

The reporting of fears can also be subject to error. When parents report that their child has a fear resulting in substantial distress or avoidance, their account can generally be relied on, but there can be error in both directions. For example, a parent with a severe fear of spiders may overplay their child's lesser fear of spiders because they project their own experiences onto the child. In the opposite direction, an insensitive parent may downplay ‘silly’ fears that the parent can't understand or sympathize with. Moreover, children may sometimes succeed in hiding their fears from their parents, particularly if they know their parents are unsympathetic to these fears. It is hard to hide things like shrieking or running away on seeing a dog. But they can potentially hide substantial inner distress from insensitive or unobservant parents and maybe they can get away with excuses to hide avoidance of feared objects. In some circumstances, children may also exaggerate fears in order to get sympathy and attention – so dissimulated feelings and misleading behaviour could potentially lead to parents reporting serious fears that are not there, as well as missing serious fears that are there.

Clinical implications

Fears are widespread, and although largely mild and transient reactions, they can become distressing and entrenched in the presence of risk factors and lead to more debilitating phobic disorders. The vast majority of childhood fears do not require specialist interventions, but many of these may be helped by interventions from parents and frontline professionals working with children (teachers, nursery staff). Recognition and contextualizing the fears may be the key. Adult responses to a child's fear may do more to exacerbate rather than reduce the stress. For example, constantly checking for monsters in the cupboard may reinforce the view that monsters actually exist in the cupboard than reassure the child that they have been banished. Parents may also sometimes have specific fears that they pass on through the way they respond to situations. If the parents originate from a place where rabies is prevalent they may avoid contact with dogs and reinforce this behaviour in the child. Health staff involved in investigative and treatment procedures require ongoing training in recognizing and managing specific fears within their agency role or remit. Such training needs to take into account developmental differences and the cultural context. However, it is important to recognize heterogeneity within ethnic groups. Clinicians need to ensure that they explore the family's view rather than rely on stereotypes. For example, does the family use the threat of exposure to animals as a punishment? It is important to understand the nature of the fears and how they have been addressed in the first instance. A range of user-friendly materials (psychoeducation, Internet) based on sound evidence could help parents understand their children' specific fears. Similar materials may also be helpful for older children and adolescents, e.g. through school mental health promotion.

Finally, despite significant impairment and simple evidence-based treatments, children with specific phobias are significantly less likely to present to all public sector services, apart from teachers, suggesting that these disorders are considerably under-recognized and under-treated (Ford et al. 2008). Fears may be dismissed as ‘normal’ and the impact on the child's functioning minimized. Given the links between childhood anxiety and anxiety, depression and substance abuse in adulthood, service providers and policy makers should give some thought to the identification and management of specific phobias (Kim-Cohen et al. 2003).

Key messages