- Email: [email protected]
Child maltreatment in Nepal: prevalence and associated factors
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Child maltreatment in Nepal: prevalence and associated factors P. Kandel a, R. Kunwar b, S. Karki b, D. Kandel c, P. Lamichhane d,* a
Research and Action in Public Health, Kathmandu, Nepal School of Public Health and Community Medicine, University of New South Wales, Sydney, Australia c Kathmandu School of Law, Bhaktapur, Nepal d Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia b
article info
abstract
Article history:
Objectives: Child maltreatment is a global public health problem. There is limited infor-
Received 5 December 2016
mation about this problem in low-income countries. We aimed to document the preva-
Received in revised form
lence and factors associated with physical punishment of children less than 14 years of age
22 May 2017
in Nepal.
Accepted 24 June 2017
Study design: Population-based cross-sectional study. Methods: We conducted an in-depth analysis using data from the Nepal Multiple Indicator Cluster Survey, a nationally representative multi-stage-stratified cluster sampling survey.
Keywords:
Data were collected from 13,000 households in 520 sample enumeration areas. We
Child maltreatment
assessed prevalence of physical punishment and different child violence related acts on
Physical punishment
5081 children aged 3e14 years for whom complete information on all acts and attitude
Nepal
towards violence was available. Logistic regression was used to investigate the association
Child
between physical punishment of child and factors such as household and maternal
Violence
demographics. Results: Our results suggested violence is common across Nepal, with data showing one in every second child is physically punished. One in every third (33%) of children were spanked, hit or slapped on the bottom, 25% were hit or slapped on the face and approximately 3% were beaten up hard. Odds of facing physical punishment were higher among children aged 3e5 years (odds ratio [OR] 2.9, 95% confidence interval [CI]: 2.0e4.3), aged 6 e8 years (OR 2.8, 95% CI: 2.2e3.7), engaged in child labour activities (OR 1.4, 95% CI: 1.1 e1.7), with mother that accepted wife beating by husband is justified (OR 1.2, 95% CI: 1.1 e1.4), whose father is currently abroad (OR 1.5, 95% CI: 1.2e1.9) and whose father is away from home but in the same country (OR 1.60, 95% CI: 1.1e2.3). The risk was also higher among children living in households that believe physical punishment of children is necessary (OR 3.5, 95% CI: 2.9e4.3) and from lower caste/indigenous (dalit/janajati) ethnicity (OR 1.3, 95% CI: 1.1e1.7). Those less likely to experience physical punishment included female children (OR 0.7, 95% CI: 0.6e0.9) and children with an older mother (34 e49 years; OR 0.5, 95% CI: 0.3e0.9).
* Corresponding author. Department of Epidemiology and Preventive Medicine, Level 4, 553 St Kilda Road, Melbourne, Victoria, 3004, Australia. E-mail addresses: [email protected] (P. Kandel), [email protected] (R. Kunwar), [email protected] (S. Karki), [email protected] (D. Kandel), [email protected] (P. Lamichhane). http://dx.doi.org/10.1016/j.puhe.2017.06.020 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
107
Conclusions: Our results suggest that physical punishment of children is common across Nepal with varying severity. Prevention efforts should focus on designing and promoting interventions that support parents to adapt alternative forms of parenting practices. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction The maltreatment of children is a global public health problem. Maltreatment is a violation of child rights contributing to childhood morbidity and mortality.1 Physical maltreatment can have long-term effects on mental health such as drug and alcohol misuse, risky sexual and criminal behaviour that can persist into adulthood.1e3 Child maltreatment may occur in different forms such as physical, sexual, emotional and psychological or in various combinations.1,2 Current evidence on child maltreatment, violence and punishment comes predominantly from developed countries with little focus on low-income countries.1,2,4 Of the few studies in low-income countries, a high prevalence of child maltreatment has been shown. A multi-country study from developing and transitional economies reported that 43% and 64% of children in Africa suffered moderate and severe physical abuse respectively.4 Several factors may lead to high rates of physical maltreatment. A meta-analysis found that the most frequently studied parent-related risk factors in child physical abuse were parental stress, parent social support and single parenthood.5 This meta-analysis also showed some less frequently studied factors (e.g. parent anger/hyperreactivity, anxiety, and psychopathology) are more strongly related to physical abuse.5 However, factors seem to vary based on location and cultural factors. A study conducted in India suggested that less focus on studies, demanding behaviour, poor grades in exams, disagreements with siblings, not doing household chores, being male, younger and low educational status of father increased the likelihood of children receiving corporal punishment.6 Literature also suggests that in societies like India and Nepal, poor legal protection for corporal punishment and lack of knowledge on short- and long-term impact on children could be driving factors of corporal punishment and child violence.7,8 In Nepal, punishment of children at home and school is widely accepted.9 A study conducted in 2003 showed that corporal punishment was common for primary school children in Nepal, resulting in school drop-outs.9 Isolation, hitting, locking children in the toilet and forcing them to do household chores are some of the reported forms of punishment in Nepal.10 More recently, a study suggested that behavioural problems of children such as addictions (cigarette, alcohol), not paying attention in school studies, child's anger issues, not obeying parents and teachers and stealing may provoke physical punishment by the parents and teachers in Nepal.11 There is little documentation on the prevalence and associated risk factors in the national context. We thus aimed to assess the prevalence and factors
associated with household level physical punishment of children aged under 14 years.
Methods We used data from the Nepal Multiple Indicator Cluster Survey 2014, a nationally representative survey conducted to monitor progress towards Millennium Development Goals.12 Two stage-stratified cluster sampling design was used to interview 13,000 households from 520 enumeration areas from 15 eco-development regions of Nepal. The 15 ecodevelopment regions were obtained by cross-classifying five development (eastern, central, western, mid-western and farwestern) and three ecological (mountain, hill, terai) regions. Detail on study methodology is published elsewhere.12 We calculated descriptive statistics regarding disciplinary acts of any household members for 5081 (5069 weighted observations) children aged 3e14 years in the past month for whom complete information on all acts and attitude towards violence was available. Our analysis of correlates regarding physical punishment is based on 4103 (4125 weighted observations) children for whom complete information on child, mother and household characteristics was available. The study data were collected by Central Bureau of Statistics, Government of Nepal in collaboration with UNICEF and with the ethics approval from Nepal Health Research Council. Data were obtained from UNICEF.13
Study variables Eleven questions were used to assess disciplinary acts perpetrated by any household member on the child aged 3e14 years (Table 1). A child was defined to have been ‘physically punished in the past month’ if at least one of the following acts had been perpetrated on child by anyone in household in the past month: spanked, hit on the bottom, hit on the face, hit on hand or beat child as hard as one could. This variable was our dependent variable. We also examined attitudes towards child punishment and attitude towards intimate partner violence in the household. Positive attitudes towards child punishment was coded as 1 if a women answered yes to the question ‘Do you believe that to bring up, raise, or educate a child properly, the child needs to be physically punished?’ and 0 otherwise. Attitudes towards intimate partner violence of women were measured through a set of five questions administered to women in the household. The variable ‘acceptance of intimate partner violence’ was coded as 1 if the mother of the child justified wife beating for at least one of the reasons: if she goes without telling him, if
108
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 1 e Prevalence of disciplinary and physical punishment acts on child (n ¼ 5081 unweighted, 5069 weighted). Disciplinary and physical punishment acts Disciplinary acts Took away privileges Explained why behaviour was wrong Shook child Shouted, yelled or screamed at child Gave child something else to do Called child dumb, lazy or another name Physical punishment acts Spanked, hit or slapped child on bottom with a bare hand Hit or slapped child on the face, head or ears Hit child on the bottom or elsewhere with belt, brush, stick and forth Hit or slapped child on the hand, arm or leg Beat child up as hard as one could Child was punished physically (spanked or hit or beaten)
% (95% CI)
28.3 90.9 30.7 77.0 25.3 28.9
(26.3e30.3) (89.3e92.3) (29e32.5) (75.2e78.7) (23.5e27.1) (27.3e30.7)
32.8 (31e34.7) 13.5 (12.1e15) 13.3 (11.9e14.7) 24.9 (23.1e26.8) 2.7 (2.2e3.4) 45.8 (43.8e47.8)
Estimates are sample weighted. CI, confidence interval. Bold indicates physical punishment of child in the past month (if child was spanked or hit on the bottom/face/hand or was beaten hard).
Christian/Kirat and others. Socio-economic status was measured through the wealth quintile available in the data set. The wealth quintile in the Nepal Multiple Indicator Cluster Survey data set was constructed based on principal component analysis of household assets (such as radio, electricity, TV) and household characteristics (such as drinking water source, type of roof, wall and flooring). Households in the lowest quintile or poorest were the reference group. For household size, households with six or less members were coded 1 and 0 otherwise. We also included location in terms of rural/urban and eco-development region in the model.
Analysis We examined the prevalence of disciplinary acts perpetrated on children and its distribution by child, mother and household characteristics. We then used logistic regression to examine the association between physical punishment of the child and maternal and household characteristics. We used four different models to examine the odds of perpetration of disciplinary acts on the child, sequentially entering child characteristics, mother characteristics, attitude towards violence in household, household characteristics and location of household. Sampling weights were used throughout the analysis. All analyses were conducted using Stata 13.
Results she neglects children, if she argues with him, if she refuses to have sex with him or if she burns the food. Independent variables were chosen based on literature, Nepalese context and availability of variables in the data set.3,14e17 Besides child characteristics, we include variables such as ethnicity and geographical location which captures parenting practices, level of inequality, level of development and cultural differences across sub-region.15,18 It needs to be noted that there are 125 caste/ethnicities living across Nepal.19 Child characteristics included were age, sex and participation of child in labour activities. The variable child labour took the value of 1 if the child worked on household land, helped in the family business, produced/sold articles or engaged in any other activity for income in cash or kind. The characteristics of the mother included were education status, age and alcohol drinking habits. The education status of mother was categorized as illiterate (no education), primary school (grade 1e5), secondary school (6e10 grade) and higher secondary school (grade 10 or above). The age of the mother was categorized as less than 25 years, 25e34 years and 35 years or above. A mother's drinking habits were defined in 2 categories (no alcohol in <30 days compared with any alcoholic drink in <30 days). Household characteristics included were father's residence, ethnicity, religion, household size and socio-economic status. Father's current residence was coded 1 if father of the child was living abroad at the time of survey (temporarily or permanently), 2 if father was not at home but living within country and 0 if father was at home. Ethnicity was categorized into Brahmin/Chhetri (higher caste), Terai/Madhesis (living in southern plains), Dalit/Janajati (lower caste or indigenous), and others. Religion was categorized as Hindu, Buddhist,
The most commonly reported method of disciplining a child was an exemplary behaviour i.e. explanation of wrong behaviour (91%) followed by shouting, screaming or yelling at the child (77%) (Table 1). About one in every two children (46%) was reported to have been physically punished by a household member. One in every three (33%) child was spanked, hit or slapped in bottom with a bare hand, one in every four child was hit or slapped on the hand, arm and leg and less than 3% were beaten-up hard. We assessed the prevalence of physical punishment by eco-development region of Nepal (Table 2). The highest proportion of children receiving physical punishment was from the mid-western hill region (60%), followed by western and mid-western mountain region (57%), far-western terai (54%) and eastern mountain (53%). The lowest proportion was in central hill (34%). Our results suggest that physical violence was normally accepted in households as reflected in the attitude of women respondents (Table 3). More than one-third of the women living in the households believed that physical punishment of child was necessary for effective upbringing of children. Onethird believed wife beating was justified for neglecting children, one-quarter believed wife beating was justified if wife went out without telling her husband and less than one-tenth believed wife beating was justified if she burnt food. Approximately half of the women felt that wife beating was justified for at least one of the five reasons: goes out without telling husband, neglects children, argues, refuses sex or burns food. The prevalence was found to be high among children aged 6e8 years (55%), male children (46%) and children who were engaged in labour activities (49%; Table 4). By mother and
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
household characteristics, prevalence was high among children belonging to illiterate mothers (48%), mother younger than 24 years (63%), indigenous or lower caste mothers (48%) and children from the lowest socio-economic status (lowest wealth quintile; 51%). The proportion was high among children whose father was abroad (49%), whose father was in the same country but away from home (53%), mother who accepted wife beating is justified (50%) and household member who believed children needs to be physically punished (64%). We found statistically significant differences for all child, mother and household characteristics except household size and religion. We examine the association with multivariable logistic regression before making any conclusion on the observed statistically significant differences. Four different models were constructed to examine the association between the child characteristics, mother characteristics, attitude towards violence and household characteristics (Table 5). Age of the child, sex of the child and participation of the child in labour activities were all significantly associated with the odds of a child receiving physical punishment. Children younger than 12 years had higher odds of experiencing physical punishment compared with children aged above 12 years after adjusting for attitude, mother and household characteristics. Female children were 22% (odds ratio [OR] 0.78, 95% confidence interval [CI]: 0.6e0.9) less likely to receive physical punishment after adjusting for other confounders. Children who were engaged in some form of labour activities also had 1.4 (95% CI: 1.1e1.7) times higher odds of experiencing physical punishment (see Model IV in Table 5). The significant association between educational status of mother and physical punishment of children was overturned
Table 2 e Proportion of child physically punished by ecodevelopment region in Nepal (n ¼ 5081 unweighted, 5069 weighted). Region Eastern mountain Eastern hill Eastern terai Central mountain Central hill Central terai Western and mid-western mountain Western hill Western terai Mid-western hill Mid-western terai Far-western mountain Far-western hill Far-western terai
No [% [95% CI)]
Yes [% [95% CI)]
Observations (n)
47.2 (40.0e54.9) 52.8 (45.4e60.1)
67
53.7 (44.5e62.6) 46.4 (37.3e55.5) 53.1 (47.0e59.1) 46.0 (40.8e53.1) 55.6 (48.4e62.5) 44.4 (37.4e51.6)
264 780 96
65.8 (59.3e71.8) 34.2 (28.1e40.6) 48.5 (43.0e54.0) 51.6 (46.0e57.0) 42.8 (37.6e48.2) 57.2 (51.8e62.4)
754 825 77
58.4 (53.5e63.1) 41.6 (36.8e46.5) 57.5 (53.0e62.1) 42.5 (37.9e47.3) 39.7 (33.1e46.8) 60.3 (53.2e66.9)
622 405 370
56.8 (50.3e63.2) 43.1 (36.8e49.6)
298
58.1 (52.5e63.5) 41.9 (36.4e47.4)
97
58.4 (53.2e63.4) 41.6 (36.6e46.8)
170
45.5 (39.6e51.7) 54.5 (48.4e60.4)
244
P-value<0.001; test conducted using chi-squared tests.
109
Table 3 e Attitude towards violence against children and women (n ¼ 5081 unweighted, 5069 weighted). Attitude Child needs to be physically punished to be brought up properly Mother accepts wife beating is justified for at least one of the five reasons If she goes out without telling husband If she neglects the children Argues: if she argues with husband Refuses: if she refuses sex with husband Burns: if she burns the food
% (95% CI) 35.2 (33.2e37.3) 46.3 (44e48.6) 28.0 (25.9e30.2) 34.0 (32e36.1) 19.0 (17.1e21.1) 4.1 (3.2e5.2) 6.3 (5.3e7.6)
Estimates are sample weighted. CI, confidence interval.
after the adjustment of household characteristics, whereas mother's age remained statistically significant. Children with mothers older than 35 years were 42% (OR 0.58, 95% CI: 0.3e0.9) less likely to receive physical punishment. Alcohol drinking habits of mothers was not associated with physical punishment of children. Children whose mother thought wife beating was justified had 1.25 times higher (95% CI: 1.01e1.49) odds of receiving physical punishment. Children living in a household where the mother accepted that children need to be physically punished was also 3.6 times (OR 3.5, 95% CI: 2.9e4.3) more likely to receive physical punishment. Absence of father from home, either abroad (OR 1.5, 95% CI: 1.2e1.9) or within country (OR 1.6, 95% CI: 1.1e2.3) resulted in higher odds of children receiving physical punishment. Children belonging to a lower or indigenous caste (dalit/ janajati) had significantly 1.3 (95% CI: 1.1e1.7) times greater likelihood of receiving physical punishment as compared with children from higher caste. We observed only weak association (P values < 0.1) of physical punishment of children with socio-economic status (middle wealth quintile) and secondary level of household head education. We found that children living in central terai (OR 1.8, 95% CI: 1.1e3.1), midwestern hill (OR 1.6, 95% CI: 1.1e2.7) and far-western terai (OR 1.8, 95% CI: 1.1e3.1) had significantly higher odds of receiving physical punishment when compared with eastern mountain region.
Discussion This is the first in-depth analysis to document prevalence and factors associated with physical punishment against children using nationally representative data in Nepal. Our results suggest that physical punishment of children is common. The prevalence ranged from 34% in central hill (most developed region in terms of human development index [HDI]) to as high as 60% in mid-western hill (with low human development index) of Nepal.20 This was an almost two-fold difference between the low HDI region and high HDI region, which may be due to educational, socio-economic and childrearing differences. The observed prevalence in our study is similar to studies conducted in other low-income countries.4 Our study showed child's age, sex, mother's age, violence accepting attitudes, father's absence from home and child
110
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 4 e Physical punishment of children by child, household and mother characteristics. Variables Child age
Sex of child Child participates in labour activities Education of mothers
Age of mothers
Alcohol consumed by mother in past 30 days Household education
Ethnicity
Religion
Socio-economic status
Household size is greater than six Rural Father's current residence
Women in household accepts husband beating Child needs to be physically punished
Categories 3e5 years 6e8 years 9e11 years 12e14 years Male Female No Yes Illiterate Primary Secondary Higher secondary or above <25 years 25e34 years 35e49 years No Yes Illiterate Primary school Secondary school Higher secondary school or above Brahmin/Chhetri (higher caste) Terai (other castes from southern region) Dalit/Janajati (indigenous caste) Others/Muslim Hindu Buddhist Christian/Kirat Muslim Poorest Poor Middle Richer Richest No Yes No Yes Home Within country Abroad No Yes No Yes
Total
No (%) Yes (%) Observations (n) P-value 47.2 44.9 57.1 69.0 53.5 58.5 58.1 52.1 51.8 56.1 58.0 69.9 37.7 51.6 62.5 56.6 51.8 48.4 54.9 60.9 68.2 63.4 52.9 51.5 54.6 56.1 53.1 56.5 55.7 48.4 49.7 56.1 57.4 68.6 55.7 57.3 65.1 53.9 58.7 46.8 51.0 61.2 50.1 67.2 35.2 56.0
52.8 55.1 42.9 31.0 46.5 41.5 41.8 47.8 48.2 43.9 42.2 30.2 62.3 48.4 37.5 43.4 48.2 51.6 45.1 39.1 31.9 36.6 47.1 48.5 45.4 43.9 47.0 43.5 44.3 51.6 50.3 43.9 42.6 31.5 44.3 42.7 34.9 46.1 41.3 53.2 49.0 38.8 49.9 32.8 64.8 44.1
413 1293 1173 1263 2128 2014 2672 1470 2185 661 757 539 142 2162 1838 3614 528 1615 885 966 676 1425 588 1916 213 3603 232 132 175 828 857 825 807 825 3465 677 753 3389 3138 815 189 2182 1960 2696 1446 4142
<0.001
0.004 0.003 <0.001
<0.001
0.121 <0.001
<0.001
0.949
<0.001
0.552 <0.001
<0.001 <0.001 <0.001
Estimates are sample weighted; P values obtained using chi-squared tests.
engagement in labour activities were associated with physical punishment of children broadly consistent with the existing literature.4,5,21,22 A study from India found that male children were more likely to suffer from physical violence.21 The findings of higher likelihood of violence in children of younger age and male children could be due to 1) anti-social behaviour of the child; 2) vulnerability of age; and/or 3) higher expectations of male children. Further qualitative research particularly in Nepal's context may provide additional insights into the reasons. Our results suggested that physical violence was widely accepted in the Nepalese households. This is consistent with existing literature from both developing and developed
countries.4,23 Akmatov4 reported that parental attitudes towards corporal punishment were the strongest predictor of child abuse. This reinforces the ongoing theory that child abuse and partner abuse need to be dealt in an integrated approach.24 We found that children whose fathers were out of home at the time of survey, either within the country or overseas, had greater odds of being physically punished compared with father living at home together with the child. This could be due to increased burden of care from absence of one of the parents.25,26 Previous literature has shown that parents who are burdened by higher child care demands are more likely to physically punish children.22
111
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 5 e Odds of child receiving physical punishment among children aged 3e14 years in Nepal [OR (95% CI)]. Factors
Model I
Child age (Ref: 12e14 years) 3e5 years 3.45*** (2.45e4.84) 6e8 years 3.45*** (2.74e4.34) 9e11 years 1.85*** (1.52e2.26) Child's sex (Ref: male) Female 0.81*** (0.70e0.93) Child labour Yes 1.89*** (1.56e2.30) Mother's education (Ref: illiterate) Primary Secondary Higher sec or above Mother age (Ref: <25 years) 25e34 years 35e49 years Alcohol consumed by mother in past 30 days Yes Mother accepts beating by husband is justified Yes Child needs to be physically punished Yes Father's current residence Abroad Within country but not at home Ethnicity (Ref: Brahmin/Chhetri) Terai caste Dalit/Janajati Others Household head education (Ref: illiterate) Primary Secondary Higher sec or above Socio-economic status (Ref: poorest) Poor Middle Richer Richest Rural residence Yes Region (Ref: eastern mountain) Eastern hill Eastern terai Central mountain Central hill Central terai Western and mid-western mountain Western hill Western terai Mid-western hill Mid-western terai Far-western mountain Far-western hill Far-western terai Weighted observations 4142
Model II
Model III
Model IV
3.08*** (2.15e4.41) 3.12*** (2.46e3.95) 1.78*** (1.45e2.18)
2.99*** (2.07e4.31) 2.95*** (2.32e3.76) 1.70*** (1.37e2.11)
2.97*** (2.04e4.34) 2.88*** (2.24e3.70) 1.71*** (1.37e2.13)
0.80*** (0.70e0.93)
0.77*** (0.66e0.90)
0.78*** (0.66e0.92)
1.67*** (1.37e2.04)
1.41*** (1.15e1.73)
1.40*** (1.13e1.73)
0.72*** (0.53e0.88) 0.66*** (0.53e0.82) 0.42*** (0.31e0.57)
0.84 (0.68e1.05) 0.86* (0.64e1.04) 0.59*** (0.43e0.80)
0.94 (0.74e1.20) 0.97 (0.72e1.32) 0.82 (0.54e1.25)
0.672** (0.45e0.99) 0.48*** (0.32e0.73)
0.74 (0.49e1.13) 0.53*** (0.34e0.82)
0.80 (0.52e1.24) 0.59** (0.37e0.92)
1.19 (0.92e1.52)
1.17 (0.91e1.49)
1.09 (0.83e1.44)
1.25** (1.04e1.49)
1.23** (1.01e1.49)
3.45*** (2.85e4.16)
3.56*** (2.91e4.35) 1.55*** (1.23e1.95) 1.60** (1.07e2.38) 1.08 (0.77e1.52) 1.36*** (1.09e1.71) 0.96 (0.61e1.52) 0.92 (0.72e1.17) 0.80* (0.63e1.03) 0.81 (0.58e1.14) 0.93 (0.69e1.25) 0.67* (0.45e1.02) 0.82 (0.56e1.21) 0.80 (0.50e1.28) 0.93 (0.70e1.24)
4142
4142
1.06 (0.64e1.77) 1.28 (0.77e2.12) 0.85 (0.56e1.30) 0.92 (0.55e1.54) 1.83** (1.09e3.07) 1.61* (0.99e2.63) 1.27 (0.81e1.99) 1.24 (0.79e1.95) 1.66** (1.01e2.73) 1.09 (0.63e1.87) 0.81 (0.48e1.35) 0.76 (0.47e1.23) 1.84** (1.12e3.05) 4142
CI, confidence interval; OR, odds ratio. Estimates are sample weighted; ***P < 0.01, **P < 0.05, *P < 0.1; unweighted observations ¼ 4124. Model I e The calculated odds ratio are from logistic regression with physical punishment as dependent variable and child's characteristics (age, sex and participation in labour of child) as independent variables. In addition to variables controlled in Model I, Model II also includes mother's characteristics (education, age and alcohol consumption habit) as independent variables. Model III includes variables in Model II and variables measuring attitude towards violence as independent variables. The final and fourth (Model IV) controls for variables controlled in Model III and household characteristics (ethnicity, religion, education of head of household head, wealth status, household size and location of residence).
112
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Our results suggested that children who were engaged in some form of labour activities had significantly 1.4 times higher odds of experiencing physical punishment. Child labour is documented to be highest in Nepal of all South Asian countries with 45% of Nepali children estimated to be economically active.27 Child labour in itself is a social form of violence and can be both the cause and consequence of physical violence.27,28 Children's inability to meet financial needs or lack of active engagement (in child labour) could lead to physical violence from household members while child labour in itself could be used as a form of disciplinary act. Owing to the nature of our data, we were unable to test this hypothesis. Nevertheless, our findings warrants further investigation into links of child labour and violence in Nepal using longitudinal and qualitative studies. In 2004, the Supreme Court of Nepal nulled a previous clause of acceptance of punishment to children by guardians for the interests of the child.29 Although the court directed the government to pursue appropriate and effective measures to prevent physical punishment, there has been little progress to date.8 The Domestic Violence (Offence and Punishment) Act 2009 and its Regulation 2010 also do not explicitly prohibit corporal or any form of punishment in child-rearing.30,31 More recently, the National Child Policy adopted in 2012 states that legislation will be enacted to prohibit corporal punishment in all settings, but any initiatives in regard to this have yet to be made.31 There have been few prevention programs and campaigns against child violence in Nepal in the past.8,30 Encouragingly, in September 2014, the Ministry of Women, Children and Social Welfare launched the National Campaign against Corporal Punishment of Children in Nepal.30 Our study findings need to be considered with its limitations. The study is cross-sectional in nature meaning causality cannot be inferred. Our study also does not take into account the quality of parentechild relationships and parental perception of the child as a problem which have elsewhere been documented to be risk factors for physical abuse of children.5 It is likely that the physical punishment has been under-reported in our study, particularly given the high approval of corporal punishment observed among respondents. Studies have found that a positive attitude towards corporal punishment results in responders being less likely to report to social service agencies.23 We also do not have documentation regarding the history of experience of physical abuse which could have further triggered the perpetrating behaviour of the household member.32,33
Conclusions Physical punishment of children was found to be a common public health problem across different regions of Nepal with varying magnitude. Further study to examine differences in parenting practices across cultures and location is needed to devise prevention measures. Legislation should be enacted to provide legal protection and to discourage widespread punishment of children. Efforts should focus on promoting alternative parenting practices, discouraging use of physical violence as disciplinary methods and promoting the use of exemplary behaviour such as explanation of wrong behaviour.
Author statements Acknowledgements The authors would like to thank UNICEF for providing access to the survey data.
Ethics approval The study was originally conducted by Central Bureau of Statistics of Nepal and UNICEF Nepal which was approved by Nepal Health Research Council.
Funding None declared.
Competing interests None declared.
references
1. Reading R, Bissell S, Goldhagen J, Harwin J, Masson J, Moynihan S, et al. Promotion of children's rights and prevention of child maltreatment. Lancet 2009;373:332e43. 2. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009;373:68e81. 3. Gershoff ET. Corporal punishment by parents and associated child behaviors and experiences: a meta-analytic and theoretical review. Psychol Bull 2002;128:539. 4. Akmatov MK. Child abuse in 28 developing and transitional countriesdresults from the Multiple Indicator Cluster Surveys. Int J Epidemiol 2010:dyq168. 5. Stith SM, Liu T, Davies LC, Boykin EL, Alder MC, Harris JM, et al. Risk factors in child maltreatment: a meta-analytic review of the literature. Aggress Violent Behav 2009;14:13e29. 6. Malhi P, Ray M. Prevalence and correlates of corporal punishment among adolescents. Stud Psychol 2004;46:219e28. 7. Rimal HS, Pokharel A. Corporal punishment and its effects in children. J Kathmandu Med Coll 2014;2:156e61. 8. South Asia Initiative to End Violence against Children. Prohibition of corporal punishment of children in South Asia: a progress review. Kathmandu, Nepal. 2011. 9. UNICEF. Violence against-children in Nepal: a study of the system of School discipline. Nepal. 2004. 10. UNICEF, Des Hommes T. Adopting the rights of the child: a study on intercountry adoption and its influence on child protection in Nepal. 2008. 11. Adhikari RP, Upadhaya N, Gurung D, Luitel NP, Burkey MD, Kohrt BA, et al. Perceived behavioral problems of school aged children in rural Nepal: a qualitative study. Child Adolesc Psychiatry Ment Health 2015;9:1. 12. Central Bureau of Statistics, UNICEF. Multiple indicator cluster Survey 2014. Kathmandu. 2015. 26/02/2016. 13. UNICEF. Surveys. 2016 [cited 2016 11/01/2016]; Available from: http://mics.unicef.org/surveys. 14. Bennett L, Tamang S, Onta P, Thapa M. Unequal citizens: gender, caste and ethnic exclusion in Nepal. Kathmandu, Nepal: The World Bank and Department for International Development; 2006.
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
15. UNDP Nepal. Nepal human development report 2009. State Transformation and Human Development; 2009. 16. Black DA, Heyman RE, Slep AMS. Risk factors for child physical abuse. Aggress Violent Behav 2001;6:121e88. 17. Parenting across cultures: childrearing, motherhood and fatherhood in non-western cultures. Springer; 2013. 18. Bennett L, Dahal DR, Govindasamy P. Caste ethnic and regional identity in Nepal: further analysis of the 2006. Nepal Demographic and Health Survey; 2008. 19. Central Bureau of Statistics. National population and housing census 2011 (national report). Kathmandu, Nepal: Government of Nepal, National Planning Commission Secretariat, Central Bureau of Statistics; 2012. 20. Government of Nepal-National Planning Commission, UNDP. Nepal human development report 2014: beyond geography. Kathmandu, Nepal: Unlocking Human Potential; 2014. 21. Deb S, Modak S. Prevalence of violence against children in families in Tripura and its relationship with socio-economic factors. J Inj Violence Res 2009;2:5e18. 22. Merritt DH. Child abuse potential: correlates with child maltreatment rates and structural measures of neighborhoods. Child Youth Serv Rev 2009;31:927e34. 23. Ashton V. The relationship between attitudes toward corporal punishment and the perception and reporting of child maltreatment*, **, +. Child Abuse Negl 2001;25:389e99. 24. Slep AMS, O'leary SG. Examining partner and child abuse: are we ready for a more integrated approach to family violence? Clin Child Family Psychol Rev 2001;4:87e107.
113
25. Azzarri C, Zezza A. International migration and nutritional outcomes in Tajikistan. Food Policy 2011;36:54e70. 26. Jayatissa R, Wickramage K. What effect does international migration have on the nutritional status and child care practices of children left behind? Int J Environ Res Public Health 2016;13:218. 27. Gilligan B. An analysis of the determinants of child labour in Nepal, the policy environment and response. Kathmandu: Understanding Children’s Work; January; 2003. 28. Makhoul J, Shayboub R, Jamal J. Violence: the silent determinant of child labor. J Child Poverty 2004;10:131e47. 29. Chapagain RP. The Nepalese experience of the Judicial € f K, Enforcement of Children's rights. In: Malekian F, Nordlo editors. The sovereignty of children in law. Cambridge Scholars Publishing; 2012. p. 346e70. 30. South Asia Initiative to End Violence Against Children. National campaign against corporal punishment in the country. Kathmandu, Nepal. 2014. 31. Global initiative to end all corporal punishment of children corporal punishment of children in Nepal. 2016. 32. Gage AJ, Silvestre EA. Maternal violence, victimization, and child physical punishment in Peru. Child Abuse Negl 2010;34:523e33. 33. Frias-Armenta M. Long-term effects of child punishment on Mexican women: a structural model. Child Abuse Negl 2002;26:371e86.
Available online at www.sciencedirect.com
Public Health journal homepage: www.elsevier.com/puhe
Original Research
Child maltreatment in Nepal: prevalence and associated factors P. Kandel a, R. Kunwar b, S. Karki b, D. Kandel c, P. Lamichhane d,* a
Research and Action in Public Health, Kathmandu, Nepal School of Public Health and Community Medicine, University of New South Wales, Sydney, Australia c Kathmandu School of Law, Bhaktapur, Nepal d Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia b
article info
abstract
Article history:
Objectives: Child maltreatment is a global public health problem. There is limited infor-
Received 5 December 2016
mation about this problem in low-income countries. We aimed to document the preva-
Received in revised form
lence and factors associated with physical punishment of children less than 14 years of age
22 May 2017
in Nepal.
Accepted 24 June 2017
Study design: Population-based cross-sectional study. Methods: We conducted an in-depth analysis using data from the Nepal Multiple Indicator Cluster Survey, a nationally representative multi-stage-stratified cluster sampling survey.
Keywords:
Data were collected from 13,000 households in 520 sample enumeration areas. We
Child maltreatment
assessed prevalence of physical punishment and different child violence related acts on
Physical punishment
5081 children aged 3e14 years for whom complete information on all acts and attitude
Nepal
towards violence was available. Logistic regression was used to investigate the association
Child
between physical punishment of child and factors such as household and maternal
Violence
demographics. Results: Our results suggested violence is common across Nepal, with data showing one in every second child is physically punished. One in every third (33%) of children were spanked, hit or slapped on the bottom, 25% were hit or slapped on the face and approximately 3% were beaten up hard. Odds of facing physical punishment were higher among children aged 3e5 years (odds ratio [OR] 2.9, 95% confidence interval [CI]: 2.0e4.3), aged 6 e8 years (OR 2.8, 95% CI: 2.2e3.7), engaged in child labour activities (OR 1.4, 95% CI: 1.1 e1.7), with mother that accepted wife beating by husband is justified (OR 1.2, 95% CI: 1.1 e1.4), whose father is currently abroad (OR 1.5, 95% CI: 1.2e1.9) and whose father is away from home but in the same country (OR 1.60, 95% CI: 1.1e2.3). The risk was also higher among children living in households that believe physical punishment of children is necessary (OR 3.5, 95% CI: 2.9e4.3) and from lower caste/indigenous (dalit/janajati) ethnicity (OR 1.3, 95% CI: 1.1e1.7). Those less likely to experience physical punishment included female children (OR 0.7, 95% CI: 0.6e0.9) and children with an older mother (34 e49 years; OR 0.5, 95% CI: 0.3e0.9).
* Corresponding author. Department of Epidemiology and Preventive Medicine, Level 4, 553 St Kilda Road, Melbourne, Victoria, 3004, Australia. E-mail addresses: [email protected] (P. Kandel), [email protected] (R. Kunwar), [email protected] (S. Karki), [email protected] (D. Kandel), [email protected] (P. Lamichhane). http://dx.doi.org/10.1016/j.puhe.2017.06.020 0033-3506/© 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
107
Conclusions: Our results suggest that physical punishment of children is common across Nepal with varying severity. Prevention efforts should focus on designing and promoting interventions that support parents to adapt alternative forms of parenting practices. © 2017 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.
Introduction The maltreatment of children is a global public health problem. Maltreatment is a violation of child rights contributing to childhood morbidity and mortality.1 Physical maltreatment can have long-term effects on mental health such as drug and alcohol misuse, risky sexual and criminal behaviour that can persist into adulthood.1e3 Child maltreatment may occur in different forms such as physical, sexual, emotional and psychological or in various combinations.1,2 Current evidence on child maltreatment, violence and punishment comes predominantly from developed countries with little focus on low-income countries.1,2,4 Of the few studies in low-income countries, a high prevalence of child maltreatment has been shown. A multi-country study from developing and transitional economies reported that 43% and 64% of children in Africa suffered moderate and severe physical abuse respectively.4 Several factors may lead to high rates of physical maltreatment. A meta-analysis found that the most frequently studied parent-related risk factors in child physical abuse were parental stress, parent social support and single parenthood.5 This meta-analysis also showed some less frequently studied factors (e.g. parent anger/hyperreactivity, anxiety, and psychopathology) are more strongly related to physical abuse.5 However, factors seem to vary based on location and cultural factors. A study conducted in India suggested that less focus on studies, demanding behaviour, poor grades in exams, disagreements with siblings, not doing household chores, being male, younger and low educational status of father increased the likelihood of children receiving corporal punishment.6 Literature also suggests that in societies like India and Nepal, poor legal protection for corporal punishment and lack of knowledge on short- and long-term impact on children could be driving factors of corporal punishment and child violence.7,8 In Nepal, punishment of children at home and school is widely accepted.9 A study conducted in 2003 showed that corporal punishment was common for primary school children in Nepal, resulting in school drop-outs.9 Isolation, hitting, locking children in the toilet and forcing them to do household chores are some of the reported forms of punishment in Nepal.10 More recently, a study suggested that behavioural problems of children such as addictions (cigarette, alcohol), not paying attention in school studies, child's anger issues, not obeying parents and teachers and stealing may provoke physical punishment by the parents and teachers in Nepal.11 There is little documentation on the prevalence and associated risk factors in the national context. We thus aimed to assess the prevalence and factors
associated with household level physical punishment of children aged under 14 years.
Methods We used data from the Nepal Multiple Indicator Cluster Survey 2014, a nationally representative survey conducted to monitor progress towards Millennium Development Goals.12 Two stage-stratified cluster sampling design was used to interview 13,000 households from 520 enumeration areas from 15 eco-development regions of Nepal. The 15 ecodevelopment regions were obtained by cross-classifying five development (eastern, central, western, mid-western and farwestern) and three ecological (mountain, hill, terai) regions. Detail on study methodology is published elsewhere.12 We calculated descriptive statistics regarding disciplinary acts of any household members for 5081 (5069 weighted observations) children aged 3e14 years in the past month for whom complete information on all acts and attitude towards violence was available. Our analysis of correlates regarding physical punishment is based on 4103 (4125 weighted observations) children for whom complete information on child, mother and household characteristics was available. The study data were collected by Central Bureau of Statistics, Government of Nepal in collaboration with UNICEF and with the ethics approval from Nepal Health Research Council. Data were obtained from UNICEF.13
Study variables Eleven questions were used to assess disciplinary acts perpetrated by any household member on the child aged 3e14 years (Table 1). A child was defined to have been ‘physically punished in the past month’ if at least one of the following acts had been perpetrated on child by anyone in household in the past month: spanked, hit on the bottom, hit on the face, hit on hand or beat child as hard as one could. This variable was our dependent variable. We also examined attitudes towards child punishment and attitude towards intimate partner violence in the household. Positive attitudes towards child punishment was coded as 1 if a women answered yes to the question ‘Do you believe that to bring up, raise, or educate a child properly, the child needs to be physically punished?’ and 0 otherwise. Attitudes towards intimate partner violence of women were measured through a set of five questions administered to women in the household. The variable ‘acceptance of intimate partner violence’ was coded as 1 if the mother of the child justified wife beating for at least one of the reasons: if she goes without telling him, if
108
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 1 e Prevalence of disciplinary and physical punishment acts on child (n ¼ 5081 unweighted, 5069 weighted). Disciplinary and physical punishment acts Disciplinary acts Took away privileges Explained why behaviour was wrong Shook child Shouted, yelled or screamed at child Gave child something else to do Called child dumb, lazy or another name Physical punishment acts Spanked, hit or slapped child on bottom with a bare hand Hit or slapped child on the face, head or ears Hit child on the bottom or elsewhere with belt, brush, stick and forth Hit or slapped child on the hand, arm or leg Beat child up as hard as one could Child was punished physically (spanked or hit or beaten)
% (95% CI)
28.3 90.9 30.7 77.0 25.3 28.9
(26.3e30.3) (89.3e92.3) (29e32.5) (75.2e78.7) (23.5e27.1) (27.3e30.7)
32.8 (31e34.7) 13.5 (12.1e15) 13.3 (11.9e14.7) 24.9 (23.1e26.8) 2.7 (2.2e3.4) 45.8 (43.8e47.8)
Estimates are sample weighted. CI, confidence interval. Bold indicates physical punishment of child in the past month (if child was spanked or hit on the bottom/face/hand or was beaten hard).
Christian/Kirat and others. Socio-economic status was measured through the wealth quintile available in the data set. The wealth quintile in the Nepal Multiple Indicator Cluster Survey data set was constructed based on principal component analysis of household assets (such as radio, electricity, TV) and household characteristics (such as drinking water source, type of roof, wall and flooring). Households in the lowest quintile or poorest were the reference group. For household size, households with six or less members were coded 1 and 0 otherwise. We also included location in terms of rural/urban and eco-development region in the model.
Analysis We examined the prevalence of disciplinary acts perpetrated on children and its distribution by child, mother and household characteristics. We then used logistic regression to examine the association between physical punishment of the child and maternal and household characteristics. We used four different models to examine the odds of perpetration of disciplinary acts on the child, sequentially entering child characteristics, mother characteristics, attitude towards violence in household, household characteristics and location of household. Sampling weights were used throughout the analysis. All analyses were conducted using Stata 13.
Results she neglects children, if she argues with him, if she refuses to have sex with him or if she burns the food. Independent variables were chosen based on literature, Nepalese context and availability of variables in the data set.3,14e17 Besides child characteristics, we include variables such as ethnicity and geographical location which captures parenting practices, level of inequality, level of development and cultural differences across sub-region.15,18 It needs to be noted that there are 125 caste/ethnicities living across Nepal.19 Child characteristics included were age, sex and participation of child in labour activities. The variable child labour took the value of 1 if the child worked on household land, helped in the family business, produced/sold articles or engaged in any other activity for income in cash or kind. The characteristics of the mother included were education status, age and alcohol drinking habits. The education status of mother was categorized as illiterate (no education), primary school (grade 1e5), secondary school (6e10 grade) and higher secondary school (grade 10 or above). The age of the mother was categorized as less than 25 years, 25e34 years and 35 years or above. A mother's drinking habits were defined in 2 categories (no alcohol in <30 days compared with any alcoholic drink in <30 days). Household characteristics included were father's residence, ethnicity, religion, household size and socio-economic status. Father's current residence was coded 1 if father of the child was living abroad at the time of survey (temporarily or permanently), 2 if father was not at home but living within country and 0 if father was at home. Ethnicity was categorized into Brahmin/Chhetri (higher caste), Terai/Madhesis (living in southern plains), Dalit/Janajati (lower caste or indigenous), and others. Religion was categorized as Hindu, Buddhist,
The most commonly reported method of disciplining a child was an exemplary behaviour i.e. explanation of wrong behaviour (91%) followed by shouting, screaming or yelling at the child (77%) (Table 1). About one in every two children (46%) was reported to have been physically punished by a household member. One in every three (33%) child was spanked, hit or slapped in bottom with a bare hand, one in every four child was hit or slapped on the hand, arm and leg and less than 3% were beaten-up hard. We assessed the prevalence of physical punishment by eco-development region of Nepal (Table 2). The highest proportion of children receiving physical punishment was from the mid-western hill region (60%), followed by western and mid-western mountain region (57%), far-western terai (54%) and eastern mountain (53%). The lowest proportion was in central hill (34%). Our results suggest that physical violence was normally accepted in households as reflected in the attitude of women respondents (Table 3). More than one-third of the women living in the households believed that physical punishment of child was necessary for effective upbringing of children. Onethird believed wife beating was justified for neglecting children, one-quarter believed wife beating was justified if wife went out without telling her husband and less than one-tenth believed wife beating was justified if she burnt food. Approximately half of the women felt that wife beating was justified for at least one of the five reasons: goes out without telling husband, neglects children, argues, refuses sex or burns food. The prevalence was found to be high among children aged 6e8 years (55%), male children (46%) and children who were engaged in labour activities (49%; Table 4). By mother and
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
household characteristics, prevalence was high among children belonging to illiterate mothers (48%), mother younger than 24 years (63%), indigenous or lower caste mothers (48%) and children from the lowest socio-economic status (lowest wealth quintile; 51%). The proportion was high among children whose father was abroad (49%), whose father was in the same country but away from home (53%), mother who accepted wife beating is justified (50%) and household member who believed children needs to be physically punished (64%). We found statistically significant differences for all child, mother and household characteristics except household size and religion. We examine the association with multivariable logistic regression before making any conclusion on the observed statistically significant differences. Four different models were constructed to examine the association between the child characteristics, mother characteristics, attitude towards violence and household characteristics (Table 5). Age of the child, sex of the child and participation of the child in labour activities were all significantly associated with the odds of a child receiving physical punishment. Children younger than 12 years had higher odds of experiencing physical punishment compared with children aged above 12 years after adjusting for attitude, mother and household characteristics. Female children were 22% (odds ratio [OR] 0.78, 95% confidence interval [CI]: 0.6e0.9) less likely to receive physical punishment after adjusting for other confounders. Children who were engaged in some form of labour activities also had 1.4 (95% CI: 1.1e1.7) times higher odds of experiencing physical punishment (see Model IV in Table 5). The significant association between educational status of mother and physical punishment of children was overturned
Table 2 e Proportion of child physically punished by ecodevelopment region in Nepal (n ¼ 5081 unweighted, 5069 weighted). Region Eastern mountain Eastern hill Eastern terai Central mountain Central hill Central terai Western and mid-western mountain Western hill Western terai Mid-western hill Mid-western terai Far-western mountain Far-western hill Far-western terai
No [% [95% CI)]
Yes [% [95% CI)]
Observations (n)
47.2 (40.0e54.9) 52.8 (45.4e60.1)
67
53.7 (44.5e62.6) 46.4 (37.3e55.5) 53.1 (47.0e59.1) 46.0 (40.8e53.1) 55.6 (48.4e62.5) 44.4 (37.4e51.6)
264 780 96
65.8 (59.3e71.8) 34.2 (28.1e40.6) 48.5 (43.0e54.0) 51.6 (46.0e57.0) 42.8 (37.6e48.2) 57.2 (51.8e62.4)
754 825 77
58.4 (53.5e63.1) 41.6 (36.8e46.5) 57.5 (53.0e62.1) 42.5 (37.9e47.3) 39.7 (33.1e46.8) 60.3 (53.2e66.9)
622 405 370
56.8 (50.3e63.2) 43.1 (36.8e49.6)
298
58.1 (52.5e63.5) 41.9 (36.4e47.4)
97
58.4 (53.2e63.4) 41.6 (36.6e46.8)
170
45.5 (39.6e51.7) 54.5 (48.4e60.4)
244
P-value<0.001; test conducted using chi-squared tests.
109
Table 3 e Attitude towards violence against children and women (n ¼ 5081 unweighted, 5069 weighted). Attitude Child needs to be physically punished to be brought up properly Mother accepts wife beating is justified for at least one of the five reasons If she goes out without telling husband If she neglects the children Argues: if she argues with husband Refuses: if she refuses sex with husband Burns: if she burns the food
% (95% CI) 35.2 (33.2e37.3) 46.3 (44e48.6) 28.0 (25.9e30.2) 34.0 (32e36.1) 19.0 (17.1e21.1) 4.1 (3.2e5.2) 6.3 (5.3e7.6)
Estimates are sample weighted. CI, confidence interval.
after the adjustment of household characteristics, whereas mother's age remained statistically significant. Children with mothers older than 35 years were 42% (OR 0.58, 95% CI: 0.3e0.9) less likely to receive physical punishment. Alcohol drinking habits of mothers was not associated with physical punishment of children. Children whose mother thought wife beating was justified had 1.25 times higher (95% CI: 1.01e1.49) odds of receiving physical punishment. Children living in a household where the mother accepted that children need to be physically punished was also 3.6 times (OR 3.5, 95% CI: 2.9e4.3) more likely to receive physical punishment. Absence of father from home, either abroad (OR 1.5, 95% CI: 1.2e1.9) or within country (OR 1.6, 95% CI: 1.1e2.3) resulted in higher odds of children receiving physical punishment. Children belonging to a lower or indigenous caste (dalit/ janajati) had significantly 1.3 (95% CI: 1.1e1.7) times greater likelihood of receiving physical punishment as compared with children from higher caste. We observed only weak association (P values < 0.1) of physical punishment of children with socio-economic status (middle wealth quintile) and secondary level of household head education. We found that children living in central terai (OR 1.8, 95% CI: 1.1e3.1), midwestern hill (OR 1.6, 95% CI: 1.1e2.7) and far-western terai (OR 1.8, 95% CI: 1.1e3.1) had significantly higher odds of receiving physical punishment when compared with eastern mountain region.
Discussion This is the first in-depth analysis to document prevalence and factors associated with physical punishment against children using nationally representative data in Nepal. Our results suggest that physical punishment of children is common. The prevalence ranged from 34% in central hill (most developed region in terms of human development index [HDI]) to as high as 60% in mid-western hill (with low human development index) of Nepal.20 This was an almost two-fold difference between the low HDI region and high HDI region, which may be due to educational, socio-economic and childrearing differences. The observed prevalence in our study is similar to studies conducted in other low-income countries.4 Our study showed child's age, sex, mother's age, violence accepting attitudes, father's absence from home and child
110
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 4 e Physical punishment of children by child, household and mother characteristics. Variables Child age
Sex of child Child participates in labour activities Education of mothers
Age of mothers
Alcohol consumed by mother in past 30 days Household education
Ethnicity
Religion
Socio-economic status
Household size is greater than six Rural Father's current residence
Women in household accepts husband beating Child needs to be physically punished
Categories 3e5 years 6e8 years 9e11 years 12e14 years Male Female No Yes Illiterate Primary Secondary Higher secondary or above <25 years 25e34 years 35e49 years No Yes Illiterate Primary school Secondary school Higher secondary school or above Brahmin/Chhetri (higher caste) Terai (other castes from southern region) Dalit/Janajati (indigenous caste) Others/Muslim Hindu Buddhist Christian/Kirat Muslim Poorest Poor Middle Richer Richest No Yes No Yes Home Within country Abroad No Yes No Yes
Total
No (%) Yes (%) Observations (n) P-value 47.2 44.9 57.1 69.0 53.5 58.5 58.1 52.1 51.8 56.1 58.0 69.9 37.7 51.6 62.5 56.6 51.8 48.4 54.9 60.9 68.2 63.4 52.9 51.5 54.6 56.1 53.1 56.5 55.7 48.4 49.7 56.1 57.4 68.6 55.7 57.3 65.1 53.9 58.7 46.8 51.0 61.2 50.1 67.2 35.2 56.0
52.8 55.1 42.9 31.0 46.5 41.5 41.8 47.8 48.2 43.9 42.2 30.2 62.3 48.4 37.5 43.4 48.2 51.6 45.1 39.1 31.9 36.6 47.1 48.5 45.4 43.9 47.0 43.5 44.3 51.6 50.3 43.9 42.6 31.5 44.3 42.7 34.9 46.1 41.3 53.2 49.0 38.8 49.9 32.8 64.8 44.1
413 1293 1173 1263 2128 2014 2672 1470 2185 661 757 539 142 2162 1838 3614 528 1615 885 966 676 1425 588 1916 213 3603 232 132 175 828 857 825 807 825 3465 677 753 3389 3138 815 189 2182 1960 2696 1446 4142
<0.001
0.004 0.003 <0.001
<0.001
0.121 <0.001
<0.001
0.949
<0.001
0.552 <0.001
<0.001 <0.001 <0.001
Estimates are sample weighted; P values obtained using chi-squared tests.
engagement in labour activities were associated with physical punishment of children broadly consistent with the existing literature.4,5,21,22 A study from India found that male children were more likely to suffer from physical violence.21 The findings of higher likelihood of violence in children of younger age and male children could be due to 1) anti-social behaviour of the child; 2) vulnerability of age; and/or 3) higher expectations of male children. Further qualitative research particularly in Nepal's context may provide additional insights into the reasons. Our results suggested that physical violence was widely accepted in the Nepalese households. This is consistent with existing literature from both developing and developed
countries.4,23 Akmatov4 reported that parental attitudes towards corporal punishment were the strongest predictor of child abuse. This reinforces the ongoing theory that child abuse and partner abuse need to be dealt in an integrated approach.24 We found that children whose fathers were out of home at the time of survey, either within the country or overseas, had greater odds of being physically punished compared with father living at home together with the child. This could be due to increased burden of care from absence of one of the parents.25,26 Previous literature has shown that parents who are burdened by higher child care demands are more likely to physically punish children.22
111
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Table 5 e Odds of child receiving physical punishment among children aged 3e14 years in Nepal [OR (95% CI)]. Factors
Model I
Child age (Ref: 12e14 years) 3e5 years 3.45*** (2.45e4.84) 6e8 years 3.45*** (2.74e4.34) 9e11 years 1.85*** (1.52e2.26) Child's sex (Ref: male) Female 0.81*** (0.70e0.93) Child labour Yes 1.89*** (1.56e2.30) Mother's education (Ref: illiterate) Primary Secondary Higher sec or above Mother age (Ref: <25 years) 25e34 years 35e49 years Alcohol consumed by mother in past 30 days Yes Mother accepts beating by husband is justified Yes Child needs to be physically punished Yes Father's current residence Abroad Within country but not at home Ethnicity (Ref: Brahmin/Chhetri) Terai caste Dalit/Janajati Others Household head education (Ref: illiterate) Primary Secondary Higher sec or above Socio-economic status (Ref: poorest) Poor Middle Richer Richest Rural residence Yes Region (Ref: eastern mountain) Eastern hill Eastern terai Central mountain Central hill Central terai Western and mid-western mountain Western hill Western terai Mid-western hill Mid-western terai Far-western mountain Far-western hill Far-western terai Weighted observations 4142
Model II
Model III
Model IV
3.08*** (2.15e4.41) 3.12*** (2.46e3.95) 1.78*** (1.45e2.18)
2.99*** (2.07e4.31) 2.95*** (2.32e3.76) 1.70*** (1.37e2.11)
2.97*** (2.04e4.34) 2.88*** (2.24e3.70) 1.71*** (1.37e2.13)
0.80*** (0.70e0.93)
0.77*** (0.66e0.90)
0.78*** (0.66e0.92)
1.67*** (1.37e2.04)
1.41*** (1.15e1.73)
1.40*** (1.13e1.73)
0.72*** (0.53e0.88) 0.66*** (0.53e0.82) 0.42*** (0.31e0.57)
0.84 (0.68e1.05) 0.86* (0.64e1.04) 0.59*** (0.43e0.80)
0.94 (0.74e1.20) 0.97 (0.72e1.32) 0.82 (0.54e1.25)
0.672** (0.45e0.99) 0.48*** (0.32e0.73)
0.74 (0.49e1.13) 0.53*** (0.34e0.82)
0.80 (0.52e1.24) 0.59** (0.37e0.92)
1.19 (0.92e1.52)
1.17 (0.91e1.49)
1.09 (0.83e1.44)
1.25** (1.04e1.49)
1.23** (1.01e1.49)
3.45*** (2.85e4.16)
3.56*** (2.91e4.35) 1.55*** (1.23e1.95) 1.60** (1.07e2.38) 1.08 (0.77e1.52) 1.36*** (1.09e1.71) 0.96 (0.61e1.52) 0.92 (0.72e1.17) 0.80* (0.63e1.03) 0.81 (0.58e1.14) 0.93 (0.69e1.25) 0.67* (0.45e1.02) 0.82 (0.56e1.21) 0.80 (0.50e1.28) 0.93 (0.70e1.24)
4142
4142
1.06 (0.64e1.77) 1.28 (0.77e2.12) 0.85 (0.56e1.30) 0.92 (0.55e1.54) 1.83** (1.09e3.07) 1.61* (0.99e2.63) 1.27 (0.81e1.99) 1.24 (0.79e1.95) 1.66** (1.01e2.73) 1.09 (0.63e1.87) 0.81 (0.48e1.35) 0.76 (0.47e1.23) 1.84** (1.12e3.05) 4142
CI, confidence interval; OR, odds ratio. Estimates are sample weighted; ***P < 0.01, **P < 0.05, *P < 0.1; unweighted observations ¼ 4124. Model I e The calculated odds ratio are from logistic regression with physical punishment as dependent variable and child's characteristics (age, sex and participation in labour of child) as independent variables. In addition to variables controlled in Model I, Model II also includes mother's characteristics (education, age and alcohol consumption habit) as independent variables. Model III includes variables in Model II and variables measuring attitude towards violence as independent variables. The final and fourth (Model IV) controls for variables controlled in Model III and household characteristics (ethnicity, religion, education of head of household head, wealth status, household size and location of residence).
112
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
Our results suggested that children who were engaged in some form of labour activities had significantly 1.4 times higher odds of experiencing physical punishment. Child labour is documented to be highest in Nepal of all South Asian countries with 45% of Nepali children estimated to be economically active.27 Child labour in itself is a social form of violence and can be both the cause and consequence of physical violence.27,28 Children's inability to meet financial needs or lack of active engagement (in child labour) could lead to physical violence from household members while child labour in itself could be used as a form of disciplinary act. Owing to the nature of our data, we were unable to test this hypothesis. Nevertheless, our findings warrants further investigation into links of child labour and violence in Nepal using longitudinal and qualitative studies. In 2004, the Supreme Court of Nepal nulled a previous clause of acceptance of punishment to children by guardians for the interests of the child.29 Although the court directed the government to pursue appropriate and effective measures to prevent physical punishment, there has been little progress to date.8 The Domestic Violence (Offence and Punishment) Act 2009 and its Regulation 2010 also do not explicitly prohibit corporal or any form of punishment in child-rearing.30,31 More recently, the National Child Policy adopted in 2012 states that legislation will be enacted to prohibit corporal punishment in all settings, but any initiatives in regard to this have yet to be made.31 There have been few prevention programs and campaigns against child violence in Nepal in the past.8,30 Encouragingly, in September 2014, the Ministry of Women, Children and Social Welfare launched the National Campaign against Corporal Punishment of Children in Nepal.30 Our study findings need to be considered with its limitations. The study is cross-sectional in nature meaning causality cannot be inferred. Our study also does not take into account the quality of parentechild relationships and parental perception of the child as a problem which have elsewhere been documented to be risk factors for physical abuse of children.5 It is likely that the physical punishment has been under-reported in our study, particularly given the high approval of corporal punishment observed among respondents. Studies have found that a positive attitude towards corporal punishment results in responders being less likely to report to social service agencies.23 We also do not have documentation regarding the history of experience of physical abuse which could have further triggered the perpetrating behaviour of the household member.32,33
Conclusions Physical punishment of children was found to be a common public health problem across different regions of Nepal with varying magnitude. Further study to examine differences in parenting practices across cultures and location is needed to devise prevention measures. Legislation should be enacted to provide legal protection and to discourage widespread punishment of children. Efforts should focus on promoting alternative parenting practices, discouraging use of physical violence as disciplinary methods and promoting the use of exemplary behaviour such as explanation of wrong behaviour.
Author statements Acknowledgements The authors would like to thank UNICEF for providing access to the survey data.
Ethics approval The study was originally conducted by Central Bureau of Statistics of Nepal and UNICEF Nepal which was approved by Nepal Health Research Council.
Funding None declared.
Competing interests None declared.
references
1. Reading R, Bissell S, Goldhagen J, Harwin J, Masson J, Moynihan S, et al. Promotion of children's rights and prevention of child maltreatment. Lancet 2009;373:332e43. 2. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009;373:68e81. 3. Gershoff ET. Corporal punishment by parents and associated child behaviors and experiences: a meta-analytic and theoretical review. Psychol Bull 2002;128:539. 4. Akmatov MK. Child abuse in 28 developing and transitional countriesdresults from the Multiple Indicator Cluster Surveys. Int J Epidemiol 2010:dyq168. 5. Stith SM, Liu T, Davies LC, Boykin EL, Alder MC, Harris JM, et al. Risk factors in child maltreatment: a meta-analytic review of the literature. Aggress Violent Behav 2009;14:13e29. 6. Malhi P, Ray M. Prevalence and correlates of corporal punishment among adolescents. Stud Psychol 2004;46:219e28. 7. Rimal HS, Pokharel A. Corporal punishment and its effects in children. J Kathmandu Med Coll 2014;2:156e61. 8. South Asia Initiative to End Violence against Children. Prohibition of corporal punishment of children in South Asia: a progress review. Kathmandu, Nepal. 2011. 9. UNICEF. Violence against-children in Nepal: a study of the system of School discipline. Nepal. 2004. 10. UNICEF, Des Hommes T. Adopting the rights of the child: a study on intercountry adoption and its influence on child protection in Nepal. 2008. 11. Adhikari RP, Upadhaya N, Gurung D, Luitel NP, Burkey MD, Kohrt BA, et al. Perceived behavioral problems of school aged children in rural Nepal: a qualitative study. Child Adolesc Psychiatry Ment Health 2015;9:1. 12. Central Bureau of Statistics, UNICEF. Multiple indicator cluster Survey 2014. Kathmandu. 2015. 26/02/2016. 13. UNICEF. Surveys. 2016 [cited 2016 11/01/2016]; Available from: http://mics.unicef.org/surveys. 14. Bennett L, Tamang S, Onta P, Thapa M. Unequal citizens: gender, caste and ethnic exclusion in Nepal. Kathmandu, Nepal: The World Bank and Department for International Development; 2006.
p u b l i c h e a l t h 1 5 1 ( 2 0 1 7 ) 1 0 6 e1 1 3
15. UNDP Nepal. Nepal human development report 2009. State Transformation and Human Development; 2009. 16. Black DA, Heyman RE, Slep AMS. Risk factors for child physical abuse. Aggress Violent Behav 2001;6:121e88. 17. Parenting across cultures: childrearing, motherhood and fatherhood in non-western cultures. Springer; 2013. 18. Bennett L, Dahal DR, Govindasamy P. Caste ethnic and regional identity in Nepal: further analysis of the 2006. Nepal Demographic and Health Survey; 2008. 19. Central Bureau of Statistics. National population and housing census 2011 (national report). Kathmandu, Nepal: Government of Nepal, National Planning Commission Secretariat, Central Bureau of Statistics; 2012. 20. Government of Nepal-National Planning Commission, UNDP. Nepal human development report 2014: beyond geography. Kathmandu, Nepal: Unlocking Human Potential; 2014. 21. Deb S, Modak S. Prevalence of violence against children in families in Tripura and its relationship with socio-economic factors. J Inj Violence Res 2009;2:5e18. 22. Merritt DH. Child abuse potential: correlates with child maltreatment rates and structural measures of neighborhoods. Child Youth Serv Rev 2009;31:927e34. 23. Ashton V. The relationship between attitudes toward corporal punishment and the perception and reporting of child maltreatment*, **, +. Child Abuse Negl 2001;25:389e99. 24. Slep AMS, O'leary SG. Examining partner and child abuse: are we ready for a more integrated approach to family violence? Clin Child Family Psychol Rev 2001;4:87e107.
113
25. Azzarri C, Zezza A. International migration and nutritional outcomes in Tajikistan. Food Policy 2011;36:54e70. 26. Jayatissa R, Wickramage K. What effect does international migration have on the nutritional status and child care practices of children left behind? Int J Environ Res Public Health 2016;13:218. 27. Gilligan B. An analysis of the determinants of child labour in Nepal, the policy environment and response. Kathmandu: Understanding Children’s Work; January; 2003. 28. Makhoul J, Shayboub R, Jamal J. Violence: the silent determinant of child labor. J Child Poverty 2004;10:131e47. 29. Chapagain RP. The Nepalese experience of the Judicial € f K, Enforcement of Children's rights. In: Malekian F, Nordlo editors. The sovereignty of children in law. Cambridge Scholars Publishing; 2012. p. 346e70. 30. South Asia Initiative to End Violence Against Children. National campaign against corporal punishment in the country. Kathmandu, Nepal. 2014. 31. Global initiative to end all corporal punishment of children corporal punishment of children in Nepal. 2016. 32. Gage AJ, Silvestre EA. Maternal violence, victimization, and child physical punishment in Peru. Child Abuse Negl 2010;34:523e33. 33. Frias-Armenta M. Long-term effects of child punishment on Mexican women: a structural model. Child Abuse Negl 2002;26:371e86.