Lifetime child sexual abuse assessed at age 18: A survey of college students from Kerala, India.

Estimates of child sexual abuse (CSA) during adolescence may undercount exposure and retrospective recollection in adulthood may be prone to recall bias. Thus, a more accurate and CSA may be possible if the question is examined in a large sample of 18 years old. This study examined the prevalence and psychosocial correlates of CSA among college-going 18 years olds. This is a cross-sectional survey of college students. 1424 students from 58 colleges selected by cluster random sampling completed a self-administered questionnaire incorporating standardized instruments. Prevalence rates were calculated. Chi-square test was used to compare categorical variables and binary logistic regression analysis was used to examine the correlates of CSA. CSA was reported by 13.2% and significantly more common among males compared to females (18.4% vs. 10.4%, P < 0.05). In the full model of binary logistic regression analysis, students who reported sexual abuse were significantly more likely to report psychological distress and suicidal thoughts. The finding was replicated even when male and female students were analyzed separately. To conclude, CSA is common among college students. With an increasing number of young people enrolling in colleges in India, timely interventions on campuses are important to reduce the psychological morbidity in this population. Copyright:

The impact of child sexual abuse (CSA), both immediate and long term, is well documented. A multitude of adverse outcomes has been reported in the victims, including poor academic performance, higher rates of depression, substance use, and suicidality.[12]

A systematic review and meta-analysis of the prevalence of CSA estimated it to be 8%–31% for girls and 3%–17% for boys.[3] Prevalence has also been found to be high in Indian studies, although with a wide variance, ranging from 4% to 41% in girls and 10% to 55% in boys.[14] Methodological differences including definitions, the type of sexual abuse studied, methods of assessments, and sociocultural variations have been postulated to explain this variance.[1] Another factor could be the age at assessment. Studies estimating prevalence in adolescent school-going children may undercount cases as exposure may not have occurred yet.[5] Similarly, retrospective recollection of events in older adults may be prone to recall bias.[6] Thus, a more accurate and complete estimate of CSA might be possible if the question is examined in 18 years old.

India in recent years has witnessed an increasing enrolment in colleges, and a significant proportion of 18 years old in Kerala is in the 1st year at university.[7] Hence, a college survey was deemed appropriate to ensure the availability of the targeted sample of 18 years old. Further, the research on sexual abuse and its psychological correlates among college-going young adults from India is sparse.[1] Hence, exploring the needs and addressing the psychological morbidity of this important subgroup are a public health priority.

In this background, this article reports the findings of a subgroup analysis from a large college survey that examined common psychological issues among students 18–25 years of age in the state of Kerala, India. This paper reports the lifetime prevalence of sexual abuse, and its psychosocial correlates among college-going 18 years old.

METHODS

Sampling

The study was a cross-sectional survey conducted in 58 of the total 123 colleges in the district of Ernakulam, Kerala, India. Colleges were selected randomly from institutions offering all major courses such as medicine, engineering, arts, and science. The college administration randomly allocated students enrolled in odd or even-numbered years (1st/3rd or 2nd/4th year).

Survey instrument

The instrument consisted of a self-administered questionnaire. The questionnaire in English and translated into Malayalam, with translation and back-translation following established guidelines. The participants could opt for either English or Malayalam version of the questionnaire.

Sociodemographic profile of the participant (age/sex/family structure/place of residence/academic performance) was collected using a checklist.

The other tools incorporated for assessment were as follows:

ISPCAN Child Abuse Screening Tool Children's Version:[8] four questions were selected from this instrument which has been validated in India to assess lifetime exposure to sexual abuse. The questions were as follows: (1) has someone misbehaved with you sexually against your will; (2) has someone forced you to look at pornographic materials against your will; (3) has someone forced you to fondle or fondled you against your will; (4) has someone forced you into a sexual relationship against your will. Those who reported abuse were also asked about their relationship with the perpetrator.

In addition, Kessler Psychological Distress Scale (K10),[9] a 10-item screening tool validated in Indian settings was used for assessment of psychological distress. Lifetime use of alcohol and tobacco was assessed using the Alcohol, Smoking and Substance Involvement Screening Test:[10] the survey instrument also had two screening questions aimed at assessing lifetime suicidality, namely: (1) have you ever thought of killing yourself in your lifetime and (2) have you ever tried to kill yourself in your lifetime.

Ethics

Institutional ethics approval was obtained by the review board, Government Medical College, Ernakulam. The survey was administered only to consenting students.

Statistical analysis

The data were analyzed using SPSS (Version-15).[11] The lifetime prevalence of sexual abuse and the relationship with the perpetrator, if any, was determined for those college students who were 18 years of age. The sociodemographic features of those who reported sexual abuse were compared with those with no history of abuse, using Chi-square test. A full model of binary logistic regression analysis was used to compare psychological distress, use of substances, suicidality as well as academic performance between these two groups, after controlling for significant sociodemographic variables. This was repeated for male and female students, separately. The odds ratio was calculated with statistical significance set at P < 0.05.

RESULTS

A total of 5784 students participated in the larger survey, 1424 students were 18 years of age. Among them, 79 students either did not consent or left their questionnaires unanswered. Of the 1345 responses analyzed (response rate – 94.5%), majority were female (n-885 [65.5%]). The higher proportion of females reflects the enrolment pattern of colleges in the state of Kerala.[12]

Lifetime sexual abuse was reported by 13.2%. Sexual abuse was significantly more common among males compared to females (18.4% vs. 10.4%, P < 0.05) [Table 1]. The two groups were comparable across other examined sociodemographic variables. Only 123 (69.5%) of the 177 students who reported abuse responded to the question regarding the perpetrator. A majority reported that the perpetrator was someone known to them (total – 52.8%, males – 44.2%, females – 58.7%) [Table 1].

Table 1

Prevalence and perpetrators of sexual abuse

	Male (n=463), n (%)	Female (n=882), n (%)	Total (n=1345), n (%)
Prevalence of sexual abuse (lifetime)	85 (18.4)	92 (10.4)	177 (13.2)
Perpetrators of most harm*
Strangers	22 (42.3)	29 (38.6)	51 (40.1)
Known to victim	23 (44.2)	44 (58.7)	67 (52.8)
Neighbors	7 (13.5)	13 (16)	20 (15.7)
Male relative	5 (9.6)	22 (28)	27 (21.2)
Female relative	2 (3.8)	2 (2.7)	4 (3.2)
Family friend	2 (3.8)	2 (2.7)	4 (3.2)
Classmate	7 (13.5)	5 (6.7)	12 (9.5)
Others	7 (13.5)	2 (2.7)	9 (7.1)

*Response rate for this question was 69.5% (127 responses)

In the full model of binary logistic regression analysis after controlling for significant sociodemographic variables, reporting of psychological distress and suicidal ideation was significantly higher in the group with a history of abuse. This finding was replicated, even when male and female students were analyzed separately [Table 2].

Table 2

Psychosocial and clinical correlates of sexual abuse

	Males (n=463), n (%)	Females (n=882), n (%)	Total (n=1345), n (%)	Abuse versus nonabused, adjusted OR* (95% CI)
				Males	Females	Total
Gender	85 (18.4)	92 (10.4)	177 (13.2)	-	-	1.0#
						2.50 (1.73-3.60)
Academic failures
No abuse	62 (16.4)	90 (11.4)	152 (12.8)	1.0	1.0	1.0
Sexual abuse	16 (18.8)	12 (13)	28 (15.4)	1.06 (0.55-2.04)	0.8 (0.38-1.74)	0.94 (0.59-1.52)
Alcohol use
No abuse	109 (29.1)	74 (9.4)	183 (15.3)	1.0	1.0	1.0
Sexual abuse	26 (30.6)	20 (22.5)	46 (25.8)	0.98 (0.54-1.77)	1.71 (0.91-3.20)	1.27 (0.82-1.95)
Tobacco use
No abuse	49 (13)	10 (1.3)	59 (4.9)	1.0	1.0	1.0
Sexual abuse	13 (15.3)	3 (3.3)	16 (8.8)	1.01 (0.47-2.15)	1.14 (0.28-4.74)	0.96 (0.49-1.87)
Psychological distress
No abuse	15.56±6.68	16.56±7.10	16.23±6.97	1.0	1.0	1.0
Sexual abuse	20.35±7.84	23.62±1.94	21.98±8.62	1.08 (1.04-1.12)	1.08 (1.04-1.11)	1.08 (1.05-1.10)
Suicidal thoughts
No abuse	51 (13.5)	157 (19.9)	208 (17.6)	1.0	1.0	1.0
Sexual abuse	27 (31.8)	47 (51.1)	74 (41.2)	2.29 (1.28-4.10)	2.07 (1.21-3.53)	2.20 (1.49-3.26)
Suicidal attempts
No abuse	3 (0.8)	28 (3.5)	31 (2.6)	1.0	1.0	1.0
Sexual abuse	2 (2.4)	14 (15.2)	16 (8.8)	1.51 (0.21-10.72)	1.35 (0.59-3.04)	1.41 (0.68-2.92)

#Reference category - Female. CI – Confidence interval; OR – Odds ratio

DISCUSSION

Our study found a prevalence of 13.4% for sexual abuse among college-going 18 years old from the State of Kerala in South India. By estimating CSA at the age of 18 years, we have reported the lifetime likelihood of CSA with minimal recall bias. However, college students being a high functioning group, our findings especially concerning the covariates studied, may still under-represent the overall morbidity in the community.

Previous studies examining CSA among college students in India have reported rates ranging from 4% to 55%.[141314] Our prevalence rates are comparable to the meta-analysis that reported the prevalence of CSA, across the world, to be 12.7%, when only self-report studies were examined.[15] In our study, sexual abuse was significantly higher among the male students compared to females whereas the studies from western countries have consistently reported higher rates among females.[315] This could be a true reflection of higher abuse among males in India,[16] or conversely, this could also be attributed to the strict social norms in conservative societies like ours, where any open discussion on female sexuality is discouraged, thereby leading to lower rates of disclosure.

We did not find an association of abuse with any of the other sociodemographic variables examined. Lower socioeconomic status and single-parent families have been inconsistently reported to increase the risk of CSA previously.[134] It could be possible that other variables such as social support, academic performance, and self-esteem may mediate the association of sociodemographic variables with adverse outcomes.[17] In keeping with the existing understanding, most of the perpetrators in our study were previously known to victims.[14] However, the higher proportion of males in our study reporting stranger abuse is in keeping with the findings reported that boys are more often victimized by someone outside the family.[18]

The subjects reporting sexual abuse had higher psychological distress and suicidal thoughts. There exists strong evidence of the relationship of psychological distress and suicidality to CSA.[116] Our study, however, did not replicate many other common associations of CSA. Students in our study reporting abuse had comparable academic performance with the nonabused group whereas numerous other studies have reported relatively poorer academic performance among victims.[1218] Our participants, enrolled in colleges, represent a high functioning subgroup, hence, the findings pertaining to their academic performance may not be entirely generalizable. CSA in our study was not correlated significantly with use of alcohol and tobacco although a consistent association has been reported across studies.[12] It is possible that substance use, a commonly reported behavioral outcome of the developmental transition to adulthood, is yet to manifest among our 18 years old and in their 1st year at university.

The study had its limitations. Assessment of sexual abuse and its correlates was by self-reported responses to limited queries from a standardized instrument. The impersonal nature of questionnaire-based survey methodology may not be ideal when it comes to disclosure about the experience of sexual abuse.[19] Only a limited number of psychosocial correlates were examined. The cross-sectional nature of the study precludes any conclusion on the direction of causality of the examined covariates. However, the large sample size with participating students enrolled in various courses makes our findings generalizable.

CONCLUSION

Our study replicates consistent findings that CSA is not uncommon and is associated with negative outcomes across multiple domains. In a country like India where almost 37 million young adults are enrolled in colleges, interventions for mitigation of morbidity of CSA can be effectively planned and delivered with significant reach and impact.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.