Effects of Home Visiting on Adolescent Mothers’ Parenting Attitudes

Abstract

We examined the impact of a home visiting intervention on 227 adolescent mothers’ parenting attitudes. At enrollment, half of mothers were at risk for child maltreatment. Mothers assigned to intervention (n = 161) received home visits and case management. Intervention and comparison mothers (n = 66) participated in monthly peer group meetings. Regression analyses controlling for enrollment differences indicated that intervention group mothers had significant improvements in 3 of 5 subscales and in total Adult-Adolescent Parenting Inventory-2 scores relative to the comparison group.

It is estimated that 10 of every 1000 children are victims of maltreatment in the United States.¹ Adolescent mothers are at greater risk for perpetrating child maltreatment than are adult mothers.^2–4 Adolescents mothers tend to have multiple characteristics found to be related to an increased likelihood of maltreatment, including lower educational attainment and income,^5,6 less knowledge about child development, unrealistic expectations for children’s behavior, and a reliance on physical forms of discipline.^7,8

Home visiting programs derived from the Healthy Families America (HFA) intervention paradigm appear to be especially helpful for reducing these risks among adolescent mothers. Studies conducted in the northeastern United States demonstrated HFA program impacts on parenting beliefs related to child maltreatment, including inappropriate expectations for behavior, a lack of empathy for the child, and beliefs about the use of corporal punishment.⁹ HFA has also been shown to affect the behaviors of first-time adolescent mothers, including reducing harsh parenting behaviors and minor physical aggression against their child.¹⁰ We investigated the influence of the Thrive Program, a home visiting project derived from the HFA model, on the parenting beliefs of adolescent mothers in the southern United States, where attitudes toward physical punishments are highly favorable.¹¹

METHODS

The study design was quasiexperimental, with participants assigned on an every-other-case basis to the intervention or the comparison group. Program staff responsible for recruitment and case assignment did not adhere to the study design, resulting in unequal groups that were not demographically different (Table 1).

TABLE 1—

Sample Characteristics and Parenting Beliefs at Enrollment: Thrive Program, Southern United States, 2007–2011

Characteristic	Intervention (n = 161), % or Mean (SD)	Comparison (n = 66), % or Mean (SD)	P	Total (n = 227), % or Mean (SD)
Age	17.3 (1.4)	17.5 (1.5)	.49	17.4 (1.4)
Race/ethnicity			.24
European American	42.1	36.8		40.5
Black	47.2	60.3		51.1
Hispanic	8.8	2.9		7.0
Other	1.9	0.0		1.3
Education			.97
< high school graduate	81.0	80.6		81.0
High school graduate or equivalent	17.1	17.9		17.3
Some college or degree	1.9	1.5		1.7
Marital Status			.91
Married	11.3	5.9		9.7
Single	86.2	88.2		86.8
Divorced, separated, or widowed	0.6	4.4		1.8
Unknown	1.9	1.5		1.8
Income, $			.62
≤ 15 000	64.0	71.4		66.2
15 001–25 000	25.9	17.9		23.6
≥ 25 001	10.1	10.7		10.2
Pregnant at enrollment	22.6	18.2	.46
AAPI-2 total score	5.44 (1.56)	5.07 (1.35)	.07	5.33 (1.42)

Note. AAPI-2 = Adult-Adolescent Parenting Inventory. No differences were statistically significant at the P < .05 level for demographics or baseline AAPI-2 scores.

Paraprofessionals provided services in 4 agencies in the southern United States. Thrive adolescents received biweekly 90-minute home visits that included case management services and parenting intervention using the HFA program model and 3 curricula: Parents as Teachers’ Born to Learn, Partners for a Healthy Baby Home Visiting, and Nurturing Parenting Programs. Both groups participated in 60-minute monthly educational peer group meetings that varied in content but included parenting education. The same interventionists served adolescents in both conditions. Treatment group adolescents participated in home visiting and monthly peer group meetings an average of 25 months (SD = 14; 46 hours [SD = 46] of home visits). Comparison group adolescents participated in monthly peer group meetings an average of 24 months (SD = 12).

Participants included 227 adolescents assigned to intervention (n = 161) or comparison (n = 66) conditions. They received services in urban (n = 79; 35%) and rural (n = 148; 65%) settings.

We measured parenting beliefs using the Adult-Adolescent Parenting Inventory (AAPI-2),⁸ which is designed to assess parenting and child rearing attitudes. Measurement occurred at enrollment and every 6 months thereafter. Standard scores range from 1 to 10. Higher scores indicate more progressive beliefs, and scores of 3 or below indicate a risk for abusive parenting behaviors.

Participants assessed only at enrollment (44%) did not differ from those with follow-ups on treatment status, income, or number of children but were significantly more likely to be older, more educated, and White. We included these and other demographic characteristics in regression models calculated using full information maximum likelihood estimation^12,13 in a Structural Equations Modeling package.¹⁴ Full information maximum likelihood provides unbiased parameter estimates in the presence of missing data.¹⁵ All analyses controlled for baseline AAPI-2 scores. Regressions followed intent-to-treat conventions, including cases derived from program assignment regardless of compliance. When participants engaged in multiple follow-up assessments, the last was used (mean = 19 months postenrollment; range = 2.4–40.5).

RESULTS

At enrollment, 114 mothers were at risk for child maltreatment in at least 1 of the AAPI-2 constructs. There were no baseline differences between treatment and comparison groups in likelihood of risk or total AAPI-2 scores (Table 1).

As shown in Table 2, at follow-up we found significant associations between treatment group and inappropriate expectations of children (R² = 0.05), strong belief in the use of corporal punishment (R² = 0.05), reversing parent–child family roles (R² = 0.03), and the total AAPI-2 score (R² = 0.05).

TABLE 2—

Regressions Predicting Follow-Up Scores: Thrive Program, Southern United States, 2007–2011

Control Variables at Enrollment	Inappropriate Expectations of Children, b (SE)	Lack of Empathy for Children’s Needs, b (SE)	Strong Belief in the Use of Corporal Punishment, b (SE)	Reversing Parent–Child Roles, b (SE)	Oppressing Children’s Power and Independence, b (SE)	Total AAPI-2 Score, b (SE)
Race/ethnicity (0 = no, 1 = yes)
Black	0.68 (0.44)	–1.06 (0.57)	–0.55 (0.45)	–0.32 (0.55)	–0.01 (0.48)	–0.11 (0.31)
Hispanic	0.68 (0.61)	–1.06 (0.78)	0.25 (0.62)	–0.86 (0.75)	–0.32 (0.66)	–0.21 (0.42)
Single mother (0 = no, 1 = yes)	–0.56 (0.46)	–0.11 (0.59)	–0.05 (0.47)	–0.47 (0.57)	0.15 (0.50)	–0.26 (0.32)
Age at baseline	–0.16 (0.14)	–0.22 (0.18)	–0.33* (0.14)	–0.14 (0.17)	0.01 (0.15)	–0.16 (0.10)
Number of children	–0.41 (0.31)	–0.38 (0.40)	–0.14 (0.32)	–0.09 (0.38)	–0.18 (0.34)	–0.29 (0.21)
Education at baseline	0.50 (0.15)	0.13 (0.19)	0.18 (0.15)	0.11 (0.18)	0.05 (0.16)	0.11 (0.10)
Family income	0.01 (0.01)	0.01 (0.01)	–0.01* (0.01)	0.00 (0.01)	0.01 (0.01)	0.00 (0.00)
Urban program site	–0.07 (0.40)	0.53 (0.51)	0.01 (0.41)	1.44** (0.50)	–0.59 (0.44)	0.31 (0.28)
AAPI-2 score at baselinea	0.35** (0.08)	0.27** (0.10)	0.40** (0.08)	0.43** (0.09)	0.34** (0.10)	0.48** (0.08)
Treatment status (0 = comparison, 1 = treatment)	0.82** (0.33)	0.58 (0.42)	0.89** (0.33)	0.83* (0.41)	0.02 (0.36)	0.55** (0.23)
Constant	3.41	6.41	7.55	4.26	0.91	4.11
Final model R2	0.22	0.20	0.32	0.27	0.19	0.35

Note. AAPI-2 = Adult-Adolescent Parenting Inventory. Higher scores indicate more optimal beliefs. Standard scores range from 1 to 10. The sample size was n = 227.

^aAAPI-2 subscale score matched to outcome.

*P < .05; **P < .01.

DISCUSSION

Although preliminary, these findings indicate that home visiting activities aimed at promoting nurturing parenting can be efficacious for adolescent parents whose attitudes may be more difficult to change, such as those residing in southern United States settings where attitudes toward corporal punishment are favorable.¹¹ At follow-up, adolescent parents who received Thrive had significantly more positive parenting beliefs than did the comparison group. Findings are consistent with other adolescent home visiting studies identifying gains in parenting attitudes and behavior,^9,10,16 particularly when parenting education is paired with case management.¹⁶

Although home visiting is expensive to implement, a similar model, Nurse Family Partnership, demonstrated cost-effectiveness through reducing child maltreatment.¹⁷ We measured parenting beliefs associated with maltreatment but not substantiated maltreatment. Limitations of our study include the low follow-up rate, difficulties with the participant assignment process, and the potential of contamination of the comparison group, as the same interventionists provided services to both groups and in the same group sessions. Despite these limitations, our analyses paired with the extant literature suggest that home-based interventions may be uniquely suited to promote positive parenting and reduce risk of child maltreatment for adolescent mothers in a variety of contexts.