Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Feb 9.
Published in final edited form as: Prev Sci. 2012 Feb;13(1):64–74. doi: 10.1007/s11121-011-0247-0

Risk for Maternal Harsh Parenting in High-Risk Families From Birth to Age Three: Does Ethnicity Matter?

Christina Gamache Martin 1,✉, Philip A Fisher 2, Hyoun K Kim 3
PMCID: PMC5300780  NIHMSID: NIHMS848820  PMID: 21935658

Abstract

Child maltreatment prevention programs typically identify at-risk families by screening for risk with limited consideration of how risk might vary by ethnicity. In this study, longitudinal data from mothers who participated in a randomized clinical trial of a home-visitation, child maltreatment prevention program (N=262) were examined to determine whether risk for harsh parenting differed among mothers who identified themselves as Spanish-speaking Latinas (n=64), English-speaking Latinas (n=102), or non-Latina Caucasians (n=96). The majority of the participants were first-time mothers (58.4%), and the average age of all participants was 23.55 years (SD=6.04). At the time of their infants’ births, the Spanish-speaking Latina mothers demonstrated higher SES risk, whereas the English-speaking Latina and non-Latina Caucasian mothers demonstrated higher psychosocial risk. Three years later, the English-speaking Latina and non-Latina Caucasian mothers reported harsher parenting behaviors than the Spanish-speaking Latina mothers. The need for prevention programs to consider how risk and protective factors differ by ethnic group membership when identifying at-risk mothers is discussed.

Keywords: Harsh parenting, Child maltreatment risk, Ethnicity, Acculturation, Latinos

The documented prevalence and negative outcomes associated with child maltreatment has led to a multitude of prevention programs primarily aimed at parents. Parents are responsible for approximately 80% of the substantiated cases of child abuse and neglect in the United States, and children aged 3 years and younger are maltreated at higher rates than older children (U.S. Department of Health and Human Services 2009). Early identification of at-risk families is thus imperative for preventing maltreatment and for developing cost-effective programs.

Although maltreatment occurs across racial, ethnic, and cultural groups, most studies of maltreatment have focused predominantly on Caucasian samples and have largely ignored ethnicity (Behl et al. 2001; Miller and Cross 2006). In their content analysis of the maltreatment literature published between 1999 and 2002, Miller and Cross reported that, although approximately 76% of the researchers collected ethnicity data, only about 33% included ethnicity in their analyses, and only about 13% focused on ethnicity as a variable of interest. Of particular relevance, about 52% of the studies that focused on ethnicity reported significant effects for ethnicity. However, because the direction of these findings has been inconsistent (Elliott and Urquiza 2006), a closer examination of ethnicity could better clarify differences in risk for and prevalence rates of maltreatment to better tailor preventative efforts.

One reason for the omission of ethnicity from prior maltreatment research might be the inherent challenges in measuring maltreatment across cultural groups, which often have divergent parenting styles and discipline tactics with varying child outcomes (Deater-Deckard et al. 1998). Thus, differences in child outcomes under similar types of parenting and disciplinary techniques might result from differences in culturally based motives or meanings underlying the use of these tactics.

Latino Families and Maltreatment

The maltreatment literature involving Latino samples is particularly inconsistent. Although national data indicate that the percentage of Latino children with confirmed cases of abuse and neglect increased by 35% from 1999 to 2007 (U.S. Department of Health and Human Services 2001, 2009), some researchers have found that Latinos are less likely to be investigated by child protective services and are generally underrepresented in the child welfare system (Dworsky et al. 2007; Lu et al. 2004). In contrast, other researchers have reported higher substantiated maltreatment rates and out-of-home placements for Latinos compared to non-Latino Caucasians even when the number of maltreatment reports between groups has been largely equivalent (Alzate and Rosenthal 2009; Church et al. 2005).

Such high rates of maltreatment among Latinos might be related to the economic disparities that many Latino families face (Church et al. 2005). Poverty and low socioeconomic status (SES) are well-established risk factors for maltreatment (Berger 2004, 2005), and Latinos in the United States experience poverty at a disproportionately higher rate than non-Latino Caucasians. In 2007, more than twice as many Latinos (21.5%) were living in poverty compared to non-Latino Caucasians (8.2%), and the median income for non-Latino Caucasians was 1.5 times that of Latinos (DeNavas-Walt et al. 2008). These elevated poverty rates and increased SES risk factors (Ayón and Lee 2005) might impose additional stress on Latino parents, perhaps making them more vulnerable to harsh parenting tactics and child maltreatment or more likely to be identified by child protective services as having less social and economic insulation.

In addition to SES risk factors, psychosocial risk factors such as a history of maltreatment, depression, substance abuse, and a lack of social support have been consistently linked to maltreatment risk in the general population (Berger 2005; Coohey 1996; Ferrari 2002; Kotch et al. 1999; Scannapieco and Connell-Carrick 2005). However, discrepancies regarding the prevalence of such psychosocial risk factors in Latino families further complicate the identification of concrete risk factors for this population. Although Latinos are frequently thought to experience less depression and substance abuse based on nationally representative samples (Alegría et al. 2008), the prevalence rates of depression (Mendelson et al. 2008) and the severity of depression (Huang et al. 2006) in Latinos might not differ from non-Latino Caucasians. Similarly, findings are inconsistent regarding Latino parenting styles and beliefs. Some researchers have reported that Latina mothers maintain more negative parenting beliefs and are less nurturing toward their children than non-Latina Caucasian mothers (Acevedo 2000; Cardona et al. 2000). In contrast, Domenech Rodríguez et al. (2009) suggested that Latino parents display greater warmth and protectiveness toward their children.

The inconsistencies in the literature might have resulted from a failure to recognize heterogeneity among Latinos. Ethnic lumping (i.e., the grouping of ethnic populations based on broad cultural connections), although more parsimonious for analysis, fails to differentiate variability within the Latino population (Fontes 1993). Although the term was originally used to describe circumstances under which diverse ethnic groups were collapsed into one broad group (e.g., Mexicans and Puerto Ricans both referred to as Latinos), much variability, especially in terms of acculturation, has been found within particular ethnic groups (e.g., Mexican-Americans; Lara et al. 2005). Acculturation is “the process by which individuals adopt the attitudes, values, customs, beliefs, and behaviors of another culture” (Abraído-Lanza et al. 2006, p. 1342). Acculturation is a complex construct to measure, and, among various methodologies, language preference is one of the most frequently used proxies of acculturation (Folsom et al. 2007; Lara et al. 2005). Language has been shown to predict unique variance in various health outcomes when considered with other acculturation measures (Montez and Eschbach 2008; Padilla et al. 2011; Salinas and Sheffield 2011) and to account for most of the variation in acculturation (Marín 1992). Language preference among Latinos has consistently demonstrated differences in a pattern termed the Latino paradox.

The Latino paradox is the finding that less acculturated Latinos (i.e., Spanish-speaking Latinos), who tend to be economically disadvantaged in comparison to English-speaking Latinos and Caucasians, tend to partake in fewer health-risking behaviors and maintain lower levels of physical and mental health problems compared to more acculturated Latinos (i.e., English-speaking Latinos; Alegría et al. 2007; Folsom et al. 2007; Lara et al. 2005; Padilla et al. 2011). This phenomenon is not yet fully understood, but it has been suggested that traditional Latino cultural values protect against stress and unhealthy behaviors (Alegría et al. 2008). Such protective factors might be related to less acculturated Latinos maintaining strong social networks, living in safe neighborhoods, and experiencing less discrimination or social exclusion than more acculturated Latinos (Dettlaff et al. 2009; Viruell-Fuentes 2007).

Although evidence supporting the Latino paradox has frequently been reported within the physical and mental health domains, less is known regarding its association to risk factors for child maltreatment (e.g., harsh parenting). Dettlaff et al. (2009) found that, within the child welfare system, more acculturated, U.S.-born Latino parents are significantly more likely to have drug abuse problems, intellectual or cognitive impairments, poor parenting skills, recent histories of arrest, and higher levels of family stress than less acculturated, foreign-born Latino parents. Similarly, Drake et al. (2011) found evidence supporting the Latino paradox, with Latino children coming from families with disproportionately higher rates of poverty but lower rates of maltreatment than non-Latino Caucasian children. However, level of acculturation was not considered as all Latinos were grouped together. Given the limited information available in terms of how acculturation might be associated with risk factors for and rates of maltreatment in at-risk families, additional research in this area is necessary to better inform preventive intervention efforts. Moreover, provided that inclusion criteria for identifying at-risk families for child maltreatment prevention programs tend to be made across ethnicity and across levels of acculturation in particular, examining potential differences in this domain is warranted.

In the present study, we examined whether there were different risk factors for harsh parenting between Latina and non-Latina Caucasian mothers. Because preferred language is frequently used and has been found to be an effective proxy for acculturation (Folsom et al. 2007; Lara et al. 2005; Marín 1992; Padilla et al. 2011; Salinas and Sheffield 2011), we also examined differences between the Latina mothers whose preferred language was Spanish versus English. The mothers were identified as being at high risk for child maltreatment at childbirth and were followed until child age 3 years. We hypothesized that both groups of Latina mothers (i.e., Spanish speaking and English speaking) would have more SES risk factors but fewer psychosocial risk factors than the non-Latina Caucasian mothers. Consistent with the Latino paradox, we further hypothesized that the Spanish-speaking Latina mothers would have more SES risk but fewer psychosocial risk factors than the English-speaking Latina mothers. Similarly, we hypothesized that the Latina mothers would engage in less harsh parenting than the non-Latina Caucasians and that the Spanish-speaking Latinas would engage in less harsh parenting than the English-speaking Latinas.

Method

Participants

The data for the present study were taken from the Healthy Families America San Diego clinical trial (Landsverk et al. 2002). The Healthy Families America intervention is a widely implemented home visitation program for high-risk families with newborns to improve parenting, promote child health and development, and prevent child abuse and neglect (see Daro and Harding 1999). The San Diego clinical trial involved an examination of the intervention’s efficacy through age 3 years.

All participants gave birth at Mary Birch Hospital in San Diego between February 1996 and March 1997 and were screened for child maltreatment risk according to the two-step process of the Healthy Families America intervention. First, the mothers’ medical charts were reviewed via the hospital’s computer system using the 15-item Hawaii Risk Indicators checklist (Hawaii Family Stress Center 1994). Second, the mothers who were identified in Step 1 to have certain risk factors (i.e., not married, received inadequate prenatal care, previously had or attempted to have an abortion, or showed multiple risk factors) were assessed in Step 2 using Kempe’s Family Stress Checklist (Kempe and Kempe 1976), a 10-item measure used to predict future child maltreatment (see description below). Scores of 25 or greater on this measure and not having an open case with child protective services were required for participation.

A total of 488 mothers participated in the study. The current analyses include 262 of these mothers from both treatment conditions (intervention and control) who self-identified as Spanish-speaking Latina (n=64; 24.4%), English-speaking Latina (n=102; 38.9%), or non-Latina Caucasian (n=96; 36.6%). The participants included in the current analyses tended to be first-time mothers (58.4%), and the average age of all participants was 23.55 years (SD=6.04). The mothers who self-identified as African American, Asian, or “other” were not included in the current analyses (n=145) because of the study’s focus. In addition, 81 mothers (30 Spanish-speaking Latinas, 29 English-speaking Latinas, and 22 non-Latina Caucasians) were excluded from the present study because they had not completed the harsh parenting measure when their children were 3 years old. The mothers excluded because of missing data did not differ in age, ethnicity, number of children, or average number of psychosocial risk factors, but they had significantly more SES risk factors (M=3.74, SD=1.30) compared to mothers who were included (M=3.22, SD=1.44), t(341)=2.92, p<.01. Further examination of the individual SES risk factors revealed that the excluded mothers had less education, χ2(1, N=343)=5.11, p<.05, and were less likely to have been employed during the previous year, χ2(1, N=343)=6.05, p<.05, than the participating mothers. There were no group differences on the remaining five SES risk factors. The potential implications of excluding these mothers are discussed as a limitation of the study.

Procedure

All recruitment and intervention procedures were approved by the San Diego State University IRB. Informed consent was obtained from the mothers, and the mothers were compensated $25 for each interview attended. Within 1–2 weeks of childbirth, each mother completed a baseline home interview with a trained family assessor to collect data on demographics, SES, social support, past and present substance use, psychological health, prenatal care, and other harsh parenting risk factors. The baseline interview was followed by annual interviews up to child age 3 years. The baseline measures and a measure of harsh parenting were assessed each year. For the present study, we used the measures obtained at baseline and at the age-3 assessment. All measures were parent self-report instruments.

Baseline Risk for Harsh Parenting

SES Risk

We assessed baseline SES risk as a general indicator of poverty level. A cumulative index, or summary score, was created based on prior research (Wekerle et al. 2007). Scores ranged 0–7 with an average of 3.22 (SD=1.44; see Table 1).

Table 1.

Percentages of mothers’ SES and psychosocial risk factors by ethnic/language group

Risk factor Full sample
(N=262)		 Spanish-speaking
Latinas (N=64)		 English-speaking
Latinas (N=102)		 Non-Latina
Caucasians
(N=96)		 Χ 2
SES Risk
 Marital status 8.96*
  Married 17 29 11 15
  Single 83 71 89 85
 Number of children 10.35**
  0-2 92 86 90 99
  3 or more 8 14 10 1
 People living per bedroom 12.48**
  Less than 3.5 90 81 89 98
  3.5 or more 10 19 11 2
 Education 41.24***
  Less than high school 55 86 55 34
  High school degree/GED 45 14 45 66
 Employment 19.10***
  Employed over past year 53 34 49 68
  Unemployed in past year 47 66 51 32
 Public assistance 0.51
  No 49 53 48 48
  Yes 51 47 52 52
 Income 8.18*
  Above poverty level 24 15 21 34
  Under poverty level 76 85 79 66
Psychosocial Risk
 Maltreated as a child 3.36
  No 26 30 30 20
  Yes 74 70 70 80
 History of SA, MH problems, or criminal behavior 87.83***
  No 39 84 37 10
  Yes 61 16 63 90
 Prior CPS involvement 6.65*
  No 90 94 93 83
  Yes 10 6 7 17
 Depression/low self-esteem 13.67**
  No 16 3 16 25
  Yes 84 97 84 75
 Multiple stressors 2.02
  No 16 12 14 20
  Yes 84 88 86 80
 Potential for violence 0.79
  No 79 83 78 77
  Yes 21 17 22 23
 Negative/unrealistic expectations for child 11.84**
  No 50 38 45 64
  Yes 50 62 55 36
 Risk for harsh discipline 6.39*
  No 61 73 54 60
  Yes 39 27 46 40
 Negative perceptions of child 3.32
  No 92 91 96 90
  Yes 8 9 4 10
 Risk for poor attachment 0.37
  No 11 11 10 12
  Yes 89 89 90 88
Open in a new tab

SA substance abuse, MH mental health, CPS child protective services

*

p<.05,

**

p<.01,

***

p<.001

Psychosocial Risk

We used Kempe’s Family Stress Checklist to assess maternal psychosocial vulnerability at baseline. This semistructured assessment interview measures 10 psychosocial stressors associated with risk for child maltreatment: 0 (not problematic) to 10 (severely problematic). The mothers’ psychosocial risk scores ranged 15–75 with an average of 35.97 (SD=11.02; see Table 1).

Harsh Parenting at Age 3

We used the Parent–Child Conflict Tactics Scales (Psychological Aggression and Minor Physical Aggression scales; Straus et al. 1998)—a parent self-report questionnaire that has demonstrated adequate test-retest reliability, internal consistency, and construct validity—to assess intent to cause pain or injury to the child through psychological and physical aggression. The measure assesses how often the caregiver engaged in harmful parenting tactics in the prior year: 0 (never) to 20 (more than 20 times). Sample items from the Psychological Aggression scale include the following: “Shouted, yelled, or screamed at your child,” “Swore or cursed at your child,” and “Threatened to spank or hit your child but did not actually do it.” Sample items from the Minor Physical Aggression subscale include the following: “Hit your child on the bottom with something like a belt, hairbrush, a stick, or some other hard object,” “Pinched your child,” and “Slapped your child on the face or head or ears.” The items from these scales were summed to form an 11-item index of harsh parenting. Two participants missed a single item on the questionnaire. As suggested by the original authors of the scale (Straus 2001), the missing data point for these two participants was replaced with the average of the other values within the subscale.

Statistical Analyses

Risk for Harsh Parenting

Omnibus chi-square tests were used to examine whether individual SES and psychosocial risk items varied by ethnic/language group. Subsequently, one-way ANOVAs were used to examine differences in cumulative risk indices for SES and psychosocial risk across ethnic/language groups. Reverse Helmert contrasts were used to test the hypotheses that the Latina mothers would have greater SES risk and lower psychosocial risk compared to the non-Latina Caucasian mothers and that a similar pattern would be found when comparing the Spanish-speaking Latina mothers to the English-speaking Latina mothers.

Harsh Parenting

The raw scores from the Parent–Child Conflict Tactics Scales were transformed using a logarithmic transformation to correct for positive skew prior to analysis as has been performed in similar samples (Lee et al. 2008). ANCOVA analyses were then tested to examine whether harsh parenting differed by ethnic/language group, while controlling for SES and psychosocial risk. In addition, as previously noted, the mothers in this study were part of an intervention. Although no significant effects on reducing child maltreatment or related outcomes between the intervention and control groups have been found to date (Landsverk et al. 2002), treatment group status (intervention vs. control) was included as a covariate to control for any potential intervention effects. Reverse Helmert contrasts were used to test the hypotheses that the Latina mothers would report less harsh parenting compared to the non-Latina Caucasian mothers and that the Spanish-speaking Latina mothers would report less harsh parenting than the English-speaking Latina mothers.

Results

Correlations

Bivariate correlation analyses indicate that SES and psychosocial risk scores were significantly related: Mothers high in SES risk were high in psychosocial risk, r(260)=.15, p<.05. SES and psychosocial risk were also significantly associated with harsh parenting: Rates of harsh parenting increased as SES risk decreased, r(260)=−.16, p<.01, and as psychosocial risk increased, r(260)=.15, p<.05.

SES and Psychosocial Risk for Harsh Parenting by Ethnicity

Chi-square tests were used to test whether individual SES and psychosocial risk factors varied by ethnic/language group (see Table 1). Significant ethnic/language group differences were found for all SES risk indicators except public assistance. The Spanish-speaking Latina mothers maintained higher rates on the majority of these indicators: more children, more people living per bedroom, less education, more unemployment, and a greater percentage living below the poverty level. The Spanish-speaking Latinas reported lower rates of risk for being single and for utilizing public assistance compared to the other groups. The non-Latina Caucasian mothers maintained lower rates for the majority of these risk factors, and the English-speaking Latina mothers typically fell between the other two groups.

Significant ethnic/language group differences were found for 4 of the 10 psychosocial risk items from Kempe’s Family Stress Checklist (see Table 1). Specifically, the Spanish-speaking Latina mothers reported higher rates of depression and acknowledged more unrealistic expectations for child behavior than the other groups. The English-speaking Latina mothers exhibited the highest rates of harsh discipline. The non-Latina Caucasian mothers demonstrated higher rates of substance abuse, mental health concerns, criminal activity, and prior involvement with child protective services than the other groups. Significant ethnic differences were not found for history of maltreatment, amount of stress faced by mothers, potential for violence, negative perceptions of their infants, or risk for developing poor attachment styles.

The differential SES and psychosocial risk between groups was further supported when individual risk indicators were examined as cumulative risk indices (see Table 2 for descriptive statistics). One-way ANOVAs revealed significant ethnic/language differences for SES risk, F(2, 259)=14.89, p<.001, and psychosocial risk, F(2, 259)=5.57, p<.01. Reverse Helmert contrasts revealed that the Latina mothers had significantly higher levels of SES risk than the non-Latina Caucasian mothers, whereas the difference between the Spanish-speaking Latina mothers and the English-speaking Latina mothers approached significance (see Table 2). In terms of psychosocial risk, the Latina mothers had significantly less psychosocial risk than the non-Latina Caucasian mothers, and the Spanish-speaking Latina mothers had significantly less psychosocial risk than the English-speaking Latina mothers.

Table 2.

Mothers’ mean scores on harsh parenting risk indices by ethnic/language group

Reverse Helmert contrasts
Spanish-speaking
Latinas (N=64)		 English-speaking
Latinas (N=102)		 Non-Latina
Caucasians (N=96)		 Latinas vs. non-Latina
Caucasians		 Spanish- vs. English-
speaking Latinas
M SD M SD M SD t p t p
SES risk 3.80 1.37 3.39 1.44 2.65 1.30 −5.34a <.001 −1.85c .07
Psychosocial risk 32.50 8.54 35.93 10.36 38.33 12.54 2.79b <.01 −2.32d <.05
Open in a new tab

SES risk score range is 0–7. Psychosocial risk score range is 0–100

a

Latinas>non-Latina Caucasians

b

Non-Latina Caucasians>Latinas

c

Spanish-speaking Latinas>English-speaking Latinas

d

English-speaking Latinas>Spanish-speaking Latinas

Effects of Harsh Parenting by Ethnicity

Given the significant relationship between the ethnic/language groups and the SES and psychosocial risk factors, ANCOVAs were utilized to control for these factors and for treatment group status. Harsh parenting differed by ethnicity at child age 3 even when risk factors and treatment group status were taken into account, F(2, 254)=5.73, p<.01, η2 =0.04. Reverse Helmert contrasts revealed that the Latina mothers reported significantly fewer harsh parenting practices than the non-Latina Caucasian mothers and that the Spanish-speaking Latina mothers reported significantly fewer harsh parenting tactics than the English-speaking Latina mothers (see Table 3). In terms of the covariates, SES was significantly associated with harsh parenting, F(1, 254)=5.32, p<.05, η2=0.02. Neither treatment group status, F(1, 254)=2.86, p>.05, η2=0.01, nor psychosocial risk, F(1, 254)=2.91, p>.05, η2=0.01, was significantly associated with harsh parenting.

Table 3.

Mothers’ mean scores on harsh parenting rates by ethnic/language group

Reverse Helmert contrasts
Spanish-speaking
Latinas (N=64)		 English-speaking
Latinas (N=102)		 Non-Latina
Caucasians (N=96)		 Latinas vs. non-
Latina Caucasians		 Spanish- vs. English-
speaking Latinas
M Madj SE M Madj SE M Madj SE t p t p
Harsh parenting 11.27 0.81 0.07 20.53 1.02 0.05 24.71 1.13 0.06 2.83a <.01 2.36b <.05
Psychological aggression 6.73 0.63 0.07 12.74 0.84 0.05 16.15 0.96 0.06 3.04a <.01 2.48b <.05
Physical aggression 4.53 0.51 0.07 7.79 0.64 0.05 8.56 0.67 0.05 1.39a .16 1.62b .11
Open in a new tab

M means for raw scores (reported for ease of interpretation), Madj adjusted means for the log transformed scores from ANCOVAs. Harsh parenting score range is 0–275. Psychological aggression score range is 0–125. Physical aggression score range is 0–150

a

Non-Latina Caucasians>Latinas

b

English-speaking Latinas>Spanish-speaking

To better understand how harsh parenting differed by ethnicity, harsh parenting was disaggregated into psychological and physical aggression. A subsequent ANCOVA revealed that, after controlling for SES, psychosocial risk factors, and treatment group status, psychological aggression differed significantly by ethnicity, F(2, 254)=6.48, p<.001, η2 =0.05. The Latina mothers reported fewer instances of psychological aggression than the non-Latina Caucasian mothers (see Table 3). Further, the Spanish-speaking Latina mothers reported fewer instances of psychological aggression than the English-speaking Latina mothers. Only SES risk was a significant covariate of psychological aggression, F(1, 254)=6.80, p<.05, η2=0.03. Neither treatment group status, F(1, 254)=3.73, p>.05, η2=0.01, nor psychosocial risk, F(1, 254)=2.87, p>.05, η2=0.01, was significantly associated with psychological aggression. In contrast, physical aggression did not differ by ethnicity, F(2, 254)=1.93, p>.05, η2=0.02, or by SES risk, F(1, 254)=2.78, p>.05, η2=0.01. Treatment group status was not associated with physical aggression, F(1, 254)=2.22, p>.05, η2=0.01. Psychosocial risk, however, was significantly associated with physical aggression, F(1, 254)=5.67, p<.05, η2=0.02.

We conducted post-hoc analyses to examine the interactive effects of ethnicity and SES risk on psychological aggression. We categorized mothers who were one standard deviation above the mean for SES risk into a high-risk group (n=51) and mothers who were one standard deviation below the mean for SES into a low-risk group (n=39). Bonferroni comparisons from a one-way ANOVA revealed that, for the high-risk group, the Spanish-speaking Latina mothers engaged in less psychological aggression (M=4.43, SD=7.53) compared to the English-speaking Latina mothers (M=13.39, SD=17.69), t(44)=−2.42, p=.05. No other significant differences were found. The mothers in the low-risk group did not differ from one another in terms of psychological aggression. Thus, increased SES risk was related to increased psychological aggression for the English-speaking Latina mothers but not for the Spanish-speaking Latina or the non-Latina Caucasian mothers.

Discussion

Prevention programs are central elements in public policy efforts to reduce child maltreatment, and identifying potential risk factors for child maltreatment is paramount. We examined whether risk factors for and rates of harsh parenting differed between Latina and non-Latina Caucasian mothers using a longitudinal sample of 262 at-risk mothers. Our hypothesis that the Latina mothers would maintain higher levels of SES risk and lower levels of psychosocial risk than the non-Latina Caucasian mothers was supported. Similarly, our hypothesis that the Spanish-speaking Latina mothers would be at lower psychosocial risk than the English-speaking Latina mothers was supported. In contrast to our expectations, the Spanish- and English-speaking Latinas did not differ significantly in terms of SES risk, although their risk scores were in the predicted direction and approached significance. These findings suggest that the English-speaking Latina mothers faced more varied risks than the other two groups, sharing high SES risk with the Spanish-speaking Latina mothers and high psychosocial risk with the non-Latina Caucasian mothers. Finally, our hypothesis that the Latina mothers in comparison to the non-Latina Caucasian mothers and the Spanish-speaking Latina mothers in comparison to the English-speaking Latina mothers would maintain lower rates of harsh parenting (controlling for SES and psychosocial risk factors and treatment group status) was also supported: The Spanish-speaking Latina mothers, although at high SES risk, exhibited less psychosocial risk and reported less harsh parenting than the other mothers.

Although SES indicators are well-established risk factors for child maltreatment (Berger 2005), our findings suggest that SES cannot be the sole risk indicator. In this sample, the mothers with greater SES risk, largely comprised of Spanish-speaking Latina mothers, were less likely to engage in harsh parenting practices. Further, when solely examining mothers with high SES risk, the English-speaking Latina mothers reported more psychological aggression than the other groups. Thus, living in poor SES conditions might be a more potent risk factor for harsh parenting in English-speaking Latinas compared to Spanish-speaking Latinas or non-Latina Caucasians. Further, our results suggest that differences in harsh parenting result more from the interplay of ethnic and acculturative differences with SES risk than from SES risk alone.

Psychosocial vulnerability appears to be a particularly potent risk for physical aggression, regardless of ethnicity. Based on the results from other prospective studies, such as the Nurse Family Partnership program (Olds et al. 1999) and the Healthy Families New York study (DuMont et al. 2008), a mother who is more psychologically vulnerable—having lower intelligence, more mental health problems, and less belief in her ability to control her life—at childbirth is at greater risk for maltreating her child. In addition, within the New York study, the intervention mothers at high psychological risk were less likely to maltreat their children compared to the control group mothers with similar levels of psychological risk, indicating that the mothers with higher psychological risk benefit from home-based interventions.

The differences found between the Spanish-and English-speaking Latina mothers in our study are congruent with the Latino paradox. Even though the Spanish-speaking Latina mothers lived in worse SES conditions than the English-speaking Latina mothers, they exhibited less psychosocial risk and less harsh parenting. Similar distinctive patterns of child maltreatment risk among immigrant and U.S.-born Latinos have been found, with immigrant Latinos exhibiting less risk and violent behavior (Dettlaff et al. 2009). Although level of acculturation was not directly assessed at baseline, preferred language is frequently used as an effective proxy for acculturation (Folsom et al. 2007; Lara et al. 2005; Marín 1992; Padilla et al. 2011; Salinas and Sheffield 2011). These differences illuminate the heterogeneity among Latinos and stress the importance of examining Latino subgroups.

One possible explanation for these results is that less acculturated Latinos are better able to maintain cultural beliefs and attitudes that mitigate their psychosocial risk and deter harsh parenting. One such culture-specific protective value that has been extensively cited in previous work with Latino populations is familism (i.e., the desire to live in close proximity to family members and the placing of higher value on supporting family-based vs. individual-based interests; Sarkisian et al. 2007). Levels of familism are often high in less acculturated Latino families, declining as acculturation progresses (Miranda et al. 2000). Familism has been shown to mediate negative parenting outcomes due to stress and living in poor economic conditions (Behnke et al. 2008) and to be a protective factor against child maltreatment in Latino and non-Latino families (Coohey 2001). Thus, familism, which is particularly salient in traditional Latino families, might mitigate child maltreatment risk.

Prevention Implications

Healthy Families America is one of the most widely utilized child maltreatment prevention programs in the nation, with more than 450 cities implementing the program throughout the United States and Canada (Prevent Child Abuse America 2003). The intervention has been shown to improve parent–child interactions, but limited progress has been found in preventing child maltreatment (Daro and Harding 1999; Duggan et al. 1999, 2004, 2007; Harding et al. 2007; Landsverk et al. 2002). Although it was not specifically examined in the current study, one possible explanation might be the programs’ primary reliance on risk for inclusion without incorporating strength-based or protective factors into the risk index, which might have incorporated some families less at risk than initially believed. Such protective factors might help explain why the Spanish-speaking Latina mothers were less likely to use harsh parenting tactics with their children even though they had the greatest SES risk. Future prevention research should continue to identify protective factors for at-risk families. In addition, incorporating strategies to help less acculturated Latina mothers maintain these protective factors and more acculturated Latina mothers and other at-risk mothers to adopt these values and behaviors could be beneficial.

Limitations

Several limitations of the present study should be acknowledged. First, although the data were collected from a prospective study of ethnically diverse, at-risk mothers, we were limited to the measures and variables included in the original study. Second, our analyses focused on ethnic differences based on self-identified Latina or non-Latina Caucasian ethnicity and language preference. Further, we examined the implications of acculturation for the Latina mothers through preferred language but did not directly assess it in this study. Future research using more stringent criteria for Latina subgroups (i.e., acculturation level, country of origin or descent, racial identity, and cultural values) would strengthen these results. Third, because parents might underreport child maltreatment (Brown et al. 1998), it is likely that lower, less accurate rates of harsh parenting were self-reported in this study. Further, substantiated cases of child maltreatment were not directly measured. However, harsh parenting is a significant indicator of risk that has been considered a proxy for child maltreatment (Lee et al. 2008). Straus (2000) contended that a parent’s propensity to use corporal punishment increases the probability that harsh discipline will advance to physical abuse, and Scannapieco and Connell-Carrick (2005) found harsh parenting to be a significant predictor of substantiated cases of child maltreatment. Future studies should include self-report measures of maltreatment and official reports of maltreatment. Fourth, although significant differences were found between our ethnic/language group comparisons, these groups were limited in size. Future research should incorporate larger samples to better discriminate any group differences. In addition, given that those mothers who were excluded from the study tended to have less education and were less likely to have been employed during the previous year, our findings should be interpreted with caution. Although all of the mothers included in the study exhibited SES risk factors, these results do not necessarily generalize to all mothers.

Despite these limitations, our results indicate that risk factors and rates of harsh parenting differ by ethnicity and level of acculturation. At childbirth, the Spanish- and English-speaking Latina mothers were more likely to report SES risk factors than the non-Latina Caucasian mothers, whereas the English-speaking Latina and non-Latina Caucasian mothers were more likely to report psychosocial risk factors than the Spanish-speaking Latina mothers. When their infants were 3 years old, the English-speaking Latina and non-Latina Caucasian mothers reported engaging in significantly harsher parenting, especially more psychological aggression, compared to the Spanish-speaking Latina mothers, even when treatment group status, SES risk factors, and psychosocial risk factors were controlled. Thus, perhaps the cultural values shared by Spanish-speaking Latina mothers serve as protective factors against psychosocial risk and harsh parenting.

Acknowledgments

Author Note Support for this research was provided by the following grants: P30 DA023920, NIDA, U.S. PHS; R01 HD045894, NICHD, U.S. PHS; and R01 MH059780, NIMH, U.S. PHS. The authors thank the families who participated in the study, their colleagues from the Child and Adolescent Services Research Center in San Diego, and Matthew Rabel for editorial assistance.

Contributor Information

Christina Gamache Martin, University of Oregon and Oregon Social Learning Center , 1227 University of Oregon, Eugene, OR 97403-1227, USA.

Philip A. Fisher, University of Oregon, Oregon Social Learning Center, and Center for Research to Practice, Oregon Social Learning Center, 10 Shelton McMurphey Boulevard, Eugene, OR 97401, USA

Hyoun K. Kim, Oregon Social Learning Center, 10 Shelton McMurphey Blvd., Eugene, OR 97401, USA

References

  1. Abraído-Lanza AF, Armbrister AN, Flórez KR, Aguirre AN. Toward a theory-driven model of acculturation in public health research. American Journal of Public Health. 2006;96:1342–1346. doi: 10.2105/AJPH.2005.064980. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Acevedo MC. The role of acculturation in explaining ethnic differences in the prenatal health-risk behaviors, mental health, and parenting beliefs of Mexican American and European American at-risk women. Child Abuse and Neglect. 2000;24:111–127. doi: 10.1016/s0145-2134(99)00121-0. [DOI] [PubMed] [Google Scholar]
  3. Alegría M, Canino G, Shrout PE, Woo M, Duan N, Vila D, et al. Prevalence of mental illness in immigrant and non-immigrant U.S. Latino groups. American Journal of Psychiatry. 2008;165:359–369. doi: 10.1176/appi.ajp.2007.07040704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Alegría M, Mulvaney-Day N, Torres M, Polo A, Cao Z, Canino G. Prevalence of psychiatric disorders across Latino subgroups in the United States. American Journal of Public Health. 2007;97:68–75. doi: 10.2105/AJPH.2006.087205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Alzate MA, Rosenthal JA. Gender and ethnic differences for Hispanic children referred to child protective services. Children and Youth Services. 2009;31:1–7. [Google Scholar]
  6. Ayón C, Lee CD. A comparative analysis of child welfare services through the eyes of African American, Caucasian, and Latino parents. Research on Social Work Practice. 2005;15:257–266. [Google Scholar]
  7. Behl LE, Crouch JL, May PF, Valente AL, Conyngham HA. Ethnicity in child maltreatment research: A content analysis. Child Maltreatment. 2001;6:143–147. doi: 10.1177/1077559501006002006. [DOI] [PubMed] [Google Scholar]
  8. Behnke AO, MacDermid SM, Coltrane SL, Parke RD, Duffy S, Widaman KF. Family cohesion in the lives of Mexican American and European American parents. Journal of Marriage and Family. 2008;70:1045–1059. [Google Scholar]
  9. Berger LM. Income, family structure, and child maltreatment risk. Children and Youth Services Review. 2004;26:725–748. [Google Scholar]
  10. Berger LM. Income, family characteristics, and physical violence toward children. Child Abuse and Neglect. 2005;29:107–133. doi: 10.1016/j.chiabu.2004.02.006. [DOI] [PubMed] [Google Scholar]
  11. Brown J, Cohen P, Johnson JG, Salzinger S. A longitudinal analysis of risk factors for child maltreatment: Findings of a 17-year prospective study of officially recorded and self-reported child abuse and neglect. Child Abuse and Neglect. 1998;22:1065–1078. doi: 10.1016/s0145-2134(98)00087-8. [DOI] [PubMed] [Google Scholar]
  12. Cardona PG, Nicholson BC, Fox RA. Parenting among Hispanic and Anglo-American mothers with young children. Journal of Social Psychology. 2000;140:357–365. doi: 10.1080/00224540009600476. [DOI] [PubMed] [Google Scholar]
  13. Church WT, II, Gross ER, Baldwin J. Maybe ignorance is not always bliss: The disparate treatment of Hispanics within the child welfare system. Children and Youth Services Review. 2005;27:1279–1292. [Google Scholar]
  14. Coohey C. Child maltreatment: Testing the social isolation hypothesis. Child Abuse and Neglect. 1996;20:241–254. doi: 10.1016/s0145-2134(95)00143-3. [DOI] [PubMed] [Google Scholar]
  15. Coohey C. The relationship between familism and child maltreatment in Latino and Anglo families. Child Maltreatment. 2001;6:130–142. doi: 10.1177/1077559501006002005. [DOI] [PubMed] [Google Scholar]
  16. Daro DA, Harding KA. Healthy Families America: Using research to enhance practice. The Future of Children. 1999;9:152–176. [PubMed] [Google Scholar]
  17. Deater-Deckard K, Dodge KA, Bates JE, Pettit GS. Multiple risk factors in the development of externalizing behavior problems: Group and individual differences. Development and Psychopathology. 1998;10:469–493. doi: 10.1017/s0954579498001709. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. DeNavas-Walt C, Proctor BD, Smith JC. Income, poverty, and health insurance coverage in the United States: 2007. U.S. Government Printing Office; Washington, DC: 2008. [Google Scholar]
  19. Dettlaff AJ, Earner I, Phillips SD. Latino children of immigrants in the child welfare system: Prevalence, characteristics, and risk. Children and Youth Services Review. 2009;31:775–783. [Google Scholar]
  20. Domenech Rodríguez MM, Donovick MR, Crowley SL. Parenting styles in a cultural context: Observations of “protective parenting” in first generation Latinos. Family Process. 2009;48:195–210. doi: 10.1111/j.1545-5300.2009.01277.x. [DOI] [PubMed] [Google Scholar]
  21. Drake B, Jolley JM, Lanier P, Fluke J, Barth RP, Jonson-Reid M. Racial bias in child protection? A comparison of competing explanations using national data. Pediatrics. 2011;127:471–478. doi: 10.1542/peds.2010-1710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Duggan A, Caldera D, Rodriguez K, Burrell L, Rohde C, Crowne SS. Impact of a statewide home visiting program to prevent child abuse. Child Abuse and Neglect. 2007;31:801–827. doi: 10.1016/j.chiabu.2006.06.011. [DOI] [PubMed] [Google Scholar]
  23. Duggan A, McFarlane E, Fuddy L, Burrell L, Higman SM, Windham A, et al. Randomized trial of a statewide home visiting program: Impact in preventing child abuse and neglect. Child Abuse and Neglect. 2004;28:597–622. doi: 10.1016/j.chiabu.2003.08.007. [DOI] [PubMed] [Google Scholar]
  24. Duggan AK, McFarlane EC, Windham AM, Rohde CA, Salkever DS, Fuddy L, et al. Evaluation of Hawaii’s healthy start program. The Future of Children. 1999;9:66–90. [PubMed] [Google Scholar]
  25. DuMont K, Mitchell-Herzfeld S, Greene R, Lee E, Lowenfels A, Rodriguez M, et al. Healthy Families New York (HFNY) randomized trial: Effects on early child abuse and neglect. Child Abuse and Neglect. 2008;32:295–315. doi: 10.1016/j.chiabu.2007.07.007. [DOI] [PubMed] [Google Scholar]
  26. Dworsky A, Courtney ME, Zinn A. Child, parent, and family predictors of child welfare services involvement among TANF applicant families. Children and Youth Services Review. 2007;29:802–820. [Google Scholar]
  27. Elliott K, Urquiza A. Ethnicity, culture, and child maltreatment. Journal of Social Issues. 2006;62:787–809. [Google Scholar]
  28. Ferrari AM. The impact of culture upon child rearing practices and definitions of maltreatment. Child Abuse and Neglect. 2002;26:793–813. doi: 10.1016/s0145-2134(02)00345-9. [DOI] [PubMed] [Google Scholar]
  29. Folsom DP, Gilmer T, Barrio C, Moore DJ, Bucardo J, Lindamer LA, et al. A longitudinal study of the use of mental health services by persons with serious mental illness: Do Spanish-speaking Latinos differ from English-speaking Latinos and Caucasians? American Journal of Psychiatry. 2007;164:1173–1180. doi: 10.1176/appi.ajp.2007.06071239. [DOI] [PubMed] [Google Scholar]
  30. Fontes L. Considering culture and oppression: Steps toward an ecology of sexual child abuse. Journal of Feminist Family Therapy. 1993;5:25–54. [Google Scholar]
  31. Harding K, Galano J, Martin J, Huntington L, Schellenbach CJ. Healthy Families America effectiveness: A comprehensive review of outcomes. Journal of Prevention and Intervention in the Community. 2007;34:149–179. doi: 10.1300/J005v34n01_08. [DOI] [PubMed] [Google Scholar]
  32. Hawaii Family Stress Center . Early identification training materials: Healthy Families America. Author; Honolulu, HI: 1994. [Google Scholar]
  33. Huang FY, Chung H, Kroenke K, Spitzer RL. Racial and ethnic differences in the relationship between depression severity and functional status. Psychiatric Services. 2006;57:498–503. doi: 10.1176/ps.2006.57.4.498. [DOI] [PubMed] [Google Scholar]
  34. Kempe R, Kempe CH. Assessing family pathology. In: Helfer RE, Kempe CH, editors. Child abuse and neglect. Ballinger; Cambridge, MA: 1976. pp. 115–126. [Google Scholar]
  35. Kotch JB, Browne DC, Dufort V, Winsor J, Catellier D. Predicting child maltreatment in the first 4 years of life from characteristics assessed in the neonatal period. Child Abuse and Neglect. 1999;23:305–319. doi: 10.1016/s0145-2134(99)00003-4. [DOI] [PubMed] [Google Scholar]
  36. Landsverk J, Carrilio T, Connelly CD, Ganger WC, Slymen DJ. Healthy Families San Diego clinical trial: Technical report. San Diego Children’s Hospital and Health Center, Child and Adolescent Services Research Center; San Diego, CA: 2002. [Google Scholar]
  37. Lara M, Gamboa C, Kahramanian MI, Morales LS, Bautista DEH. Acculturation and Latino health in the United States: A review of the literature and its sociopolitical context. Annual Review of Public Health. 2005;26:367–397. doi: 10.1146/annurev.publhealth.26.021304.144615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lee SJ, Guterman NB, Lee Y. Risk factors for paternal physical child abuse. Child Abuse and Neglect. 2008;32:846–858. doi: 10.1016/j.chiabu.2007.11.006. [DOI] [PubMed] [Google Scholar]
  39. Lu YE, Landsverk J, Ellis-Macleod E, Newton R, Ganger W, Johnson I. Race, ethnicity, and case outcomes in child protective services. Children and Youth Services Review. 2004;26:447–461. [Google Scholar]
  40. Marín G. Issues in the measurement of acculturation among Hispanics. In: Geisinger KF, editor. Psychological testing of Hispanics. American Psychological Association; Washington, DC: 1992. pp. 23–51. [Google Scholar]
  41. Mendelson T, Rehkopf DH, Kubzansky LD. Depression among Latinos in the United States: A meta-analytic review. Journal of Consulting and Clinical Psychology. 2008;76:355–366. doi: 10.1037/0022-006X.76.3.355. [DOI] [PubMed] [Google Scholar]
  42. Miller AB, Cross T. Ethnicity in child maltreatment research: A replication of Behl et al.’s content analysis. Child Maltreatment. 2006;11:16–26. doi: 10.1177/1077559505278272. [DOI] [PubMed] [Google Scholar]
  43. Miranda AO, Estrada D, Firpo-Jimenez M. Differences in family cohesion, adaptability, and environment among Latino families in dissimilar stages of acculturation. The Family Journal: Counseling and Therapy for Couples and Families. 2000;8:341–350. [Google Scholar]
  44. Montez JK, Eschbach K. Country of birth and language are uniquely associated with intakes of fat, fiber, and fruits and vegetables among Mexican-American women in the United States. Journal of the American Dietetic Association. 2008;108:473–480. doi: 10.1016/j.jada.2007.12.008. [DOI] [PubMed] [Google Scholar]
  45. Olds DL, Henderson CR, Jr., Kitzman HJ, Eckenrode JJ, Cole RE, Tatelbaum RC. Prenatal and infancy home visitation by nurses: Recent findings. The Future of Children. 1999;9:44–65. [PubMed] [Google Scholar]
  46. Padilla R, Steiner JF, Havranek EP, Beaty B, Davidson AJ, Bull S. A comparison of different measures of acculturation with cardiovascular risk factors in Latinos with hypertension. Journal of Immigrant Minority Health. 2011;13:284–292. doi: 10.1007/s10903-010-9434-5. [DOI] [PubMed] [Google Scholar]
  47. Prevent Child Abuse America . Prevent Child Abuse America (Newsletter 13) Author; Chicago, IL: 2003. [Google Scholar]
  48. Salinas JJ, Sheffield KM. English language use, health and mortality in older Mexican Americans. Journal of Minority Health. 2011;13:232–238. doi: 10.1007/s10903-009-9273-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Sarkisian N, Gerena M, Gerstel N. Extended family integration among Euro and Mexican Americans: Ethnicity, gender, and class. Journal of Marriage and Family. 2007;69:40–54. [Google Scholar]
  50. Scannapieco M, Connell-Carrick K. Focus on the first years: Correlates of substantiation of child maltreatment for families with children 0 to 4. Children and Youth Services Review. 2005;27:1307–1323. [Google Scholar]
  51. Straus MA. Corporal punishment and primary prevention of physical abuse. Child Abuse and Neglect. 2000;24:1109–1114. doi: 10.1016/s0145-2134(00)00180-0. [DOI] [PubMed] [Google Scholar]
  52. Straus MA. Scoring the CTS2 and CTSPC. University of New Hampshire, Family Research Laboratory; Durham, NH: 2001. [Google Scholar]
  53. Straus MA, Hamby SL, Finkelhor D, Moore DW, Runyan D. Identification of child maltreatment with the Parent–Child Conflict Tactics Scales: Development and psychometric data for a national sample of American parents. Child Abuse and Neglect. 1998;22:249–270. doi: 10.1016/s0145-2134(97)00174-9. [DOI] [PubMed] [Google Scholar]
  54. U.S. Department of Health and Human Services . Child Maltreatment 1999. U.S. Government Printing Office; Washington, DC: 2001. [Google Scholar]
  55. U.S. Department of Health and Human Services . Child Maltreatment 2007. U.S. Government Printing Office; Washington, DC: 2009. [Google Scholar]
  56. Viruell-Fuentes EA. Beyond acculturation: Immigration, discrimination, and health research among Mexicans in the United States. Social Science and Medicine. 2007;65:1524–1535. doi: 10.1016/j.socscimed.2007.05.010. [DOI] [PubMed] [Google Scholar]
  57. Wekerle C, Wall AM, Leung E, Trocmé N. Cumulative stress and substantiated maltreatment: The importance of caregiver vulnerability and adult partner violence. Child Abuse and Neglect. 2007;31:427–443. doi: 10.1016/j.chiabu.2007.03.001. [DOI] [PubMed] [Google Scholar]

RESOURCES