Age-17 Follow-up of Home Visiting Intervention

This study is a longitudinal follow-up of 670 primarily African-American women and their
18-year-old firstborn children enrolled since 1990 in a randomized controlled trial (RCT) of
prenatal and infancy home visiting by nurses. Nurses in this program are charged with
improving pregnancy outcomes, child health and development, and maternal economic
self-sufficiency.1 This follow-up examines whether earlier program effects on maternal and
child functioning 2-7 lead to less violent antisocial behavior, psychopathology, substance
use and use-disorders, and risk for HIV; whether these effects are greater for those who
carry genetic susceptibility to the environment and are at environmental risk; and whether
program effects replicate those found with whites in an earlier trial.8-10 Results from
earlier phases of follow-up from this trial found that the Memphis program affected women's
prenatal health, fertility, partner relations, and use of welfare; children's injuries,
cognition, language, achievement, conduct, depression/anxiety, and use of substances through
child age 12.2-7 Program effects on maternal life-course were concentrated among mothers with
higher psychological resources (better intellectual functioning, mental health, and sense of
mastery), probably because higher-resource mothers could envision their success in the world
of work, leading to better pregnancy planning and employment. Program effects on children
were greater for those born to mothers with low psychological resources, because without
help, low-resource mothers are especially challenged in the care of their children and their
children function less well. Given the damaging effects of early stressors on developing
neural circuitry, and given that many early neural developmental insults do not become fully
evident until synaptic pruning is complete in late adolescence and early adulthood, there was
reason to expect this early intervention would have enduring effects at youth age 18.

Hypotheses for Primary Grant

We specified hypotheses based upon the pattern of results found through child age 12, and
separated them into primary and secondary hypotheses. Following the original formulation of
hypotheses, we edited them to take into account results from the earlier Elmira trial10 that
were analyzed following the submission of the proposal for the current phase of follow-up in
Memphis. We had originally hypothesized that program effects would be more pronounced for
mothers and children living in the most disadvantaged neighborhoods in Memphis, but realized
as these data were being gathered that virtually all of the participants in the Memphis trial
lived in neighborhoods that were so disadvantaged that there was little meaningful variation
among neighborhoods, and therefore removed this aspect of our hypotheses. We also found that
it was impossible to consistently gather information from children's school records on
outcomes like conduct grades from hundreds of schools, so substituted high school graduation
as a secondary outcome. These refined hypotheses were specified prior to the completion of
data gathering and any analysis of treatment-control differences. We specify the original
hypotheses and then indicate the revised hypotheses for maternal and child outcomes. Compared
to control-group counterparts:

Original Maternal Outcomes Hypotheses

1. (Primary) The program will continue to improve maternal life-course (fewer short
inter-birth intervals, less use of welfare, more stable partner relations), especially
for mothers with higher psychological resources.

2. (Secondary) The program will reduce maternal substance use disorders (SUDs) and
depression, effects that will be more pronounced for a) mothers with low psychological
resources, and b) those living in the most disadvantaged neighborhoods at registration.

Revised Maternal Outcome Hypotheses

1. (Primary) The program will continue to improve maternal life-course (reflected in total
costs of welfare - SNAP, TANF, Medicaid), especially for mothers with higher
psychological resources.

2. (Secondary) The program will reduce maternal substance use disorders (SUDs) and
depression.

Original Child Outcomes Hypotheses

3. (Primary) The program will improve the health and development of firstborn children who
will exhibit: a) superior cognitive, language, and academic functioning, and executive
cognitive functioning (ECF); b) less depression and anxiety; c) fewer failed conduct
grades and school disciplinary actions, d) less violent behavior and gang membership,
and fewer arrests, juvenile detentions, and convictions - especially for crimes
involving interpersonal violence.

4. (Primary) The program will reduce youth risk for HIV infection, including a) use of
substances and SUDs; b) risky sexual behaviors; c) sexually transmitted infections
(STIs) and d) pregnancies.

5. (Primary) Program effects on youth will be more pronounced for a) males, b) those born
to low-resource mothers, and c) those living in the most disadvantaged neighborhoods at
registration.

Revised Child Outcome Hypotheses 3. (Primary) The program will improve the health and
development of firstborn children who will exhibit: a) superior cognitive, language, and
academic functioning; b) less depression and anxiety; d) less gang membership, and fewer
arrests, convictions, and self-reported antisocial behavior - especially for crimes involving
interpersonal violence.

4. (Primary) The program will reduce youth risk for HIV infection, pregnancies, births, use
of substances, and SUDs.

5. (Secondary) The program will improve firstborn children's executive cognitive functioning
(ECF); and rates of high school graduation.

6. (Primary) Program effects on cognitive, language, and academic functioning, and executive
cognitive functioning will be more pronounced among those born to low-resource mothers and on
arrests and convictions among females.

Maternal and Child Outcomes (Not Revised) 7. (Secondary) Program effects on mothers and
youth, in preliminary analyses, will be more pronounced for those with genetic
vulnerabilities:

1. Effects on youth depression and anxiety will be greater for those with low-activity
genotypes (S/S, LG/LG, S/LG) of the serotonin transporter gene (SLC6A4) promoter
polymorphism, 5-HTTLPR, compared to those with high-activity genotypes (LA/LA); effects
on these outcomes will be of intermediate magnitude for those with intermediate
activity-level genotypes (S/LA, LA/LG).

2. Effects on youth violent antisocial behavior, SUDs, and risky sexual behavior will be
more pronounced among males with the MAOA-LPR low activity alleles compared to males
with MAOA-LPR high activity alleles, and among both males and females with 2 copies of
the high-activity Val allele of the COMT Val158Met polymorphism compared to those with 2
copies of the low-activity met allele or heterozygotes.

3. Effects on maternal SUDs will be concentrated among mothers with 2 copies of the Val158
alleles.

4. Effects on child outcomes will be more pronounced among youth born to mothers with
either 1) the S/S, S/LG and "LG/LG" (low-activity) genotypes of 5-HTTLPR (conferring
susceptibility for depression under adversity) or 2) 2 copies of the high activity COMT
Val158 allele (conferring susceptibility to compromised ECF and SUDs under conditions of
adversity).

8. (Secondary) Program effects on adolescent functioning will be explained by its
improvement in prenatal health, early care of the child, maternal life-course, and
earlier child academic and behavioral functioning.

Examination of Intervention Effects on Subsequent Children With an administrative
supplement, we addressed the following questions focused on subsequent children born
within 5 years of the first child. Note that these questions were framed with no
specific hypotheses about the degree to which particular subgroups would benefit from
the intervention, given that intervention impact on pregnancy planning had been most
pronounced on women with higher psychological resources.

1. To what degree does this program improve the health and development of subsequent
children in terms of their a) language, academic, and executive cognitive
functioning (ECF); b) depression and anxiety; c) failed conduct grades, d) violent
behavior and gang membership, and e) arrests, juvenile detentions, and convictions,
especially for violent crimes?

2. To what degree does this program reduce subsequent children's risk for HIV
infection, including a) use and abuse of substances; b) risky sexual behaviors; c)
sexually transmitted infections (STIs) and d) pregnancies?

3. To what degree are the program effects on subsequent children more pronounced for
a) males, b) those born to high-resource mothers, and c) those living in the most
disadvantaged neighborhoods at registration?

4. To what degree are program effects on subsequent children's functioning explained
by its earlier impact on a) the timing and rates of subsequent births; b) families'
use of welfare-related services; c) stability in partner relationships; d)
improvements in neighborhood contexts; and e) antisocial behavior among the
first-borns?

Aims of the Benefit-Cost Analysis

With an administrative supplement, we conducted a benefit-cost analysis of NFP in
Memphis. The benefit-cost study was designed to:

1. Estimate return on investment in Memphis NFP from the perspectives of government,
society and individual participants.

2. Estimate the quality-adjusted life year (QALY) savings produced by the Memphis NFP.

3. Combine effectiveness estimates from Memphis NFP with those from other NFP
evaluations and produce a combined estimate.

4. Develop a model that states can use to estimate the value of funding NFP programs.

5. Compare the cost-effectiveness of the NFP to other commonly delivered childhood
interventions.

Inclusion Criteria:

- Women who were enrolled in the New Mothers Study and their children as described in
Study Population Description.

Exclusion Criteria:

- Women who were not enrolled in the New Mothers Study and their children as described
in the Study Population Description.