Research Methods

Association Between Childhood Residential Mobility and Non-medical Use of Prescription Drugs Among American Youth

Meagan E. Stabler1 • Kelly K. Gurka2 • Laura R. Lander3

Published online: 9 July 2015

� Springer Science+Business Media New York 2015

Abstract

Introduction Prescription drug abuse is a public health

epidemic, resulting in 15,000 deaths annually. Disruption

of childhood residence has been shown to increase drug-

seeking behavior among adolescents; however, little

research has explored its association specifically with non-

medical use of prescription drugs (NMUPD). The objective

of the study was to measure the association between resi-

dential mobility and NMUPD.

Methods The 2010 National Survey on Drug Use and

Health data were analyzed for 15,745 participants aged

12–17 years. NMUPD was defined as self-report of any

non-medical use (i.e., taking a prescription drug that was

not prescribed to them or consumption for recreational

purposes) of tranquilizers, pain relievers, sedatives, or

stimulants. Logistic regression for survey data was used to

estimate the association between residential mobility and

NMUPD, adjusting for potential confounders.

Results After controlling for demographic, intrapersonal,

interpersonal, and community factors, adolescents with low

mobility (1–2 moves in the past 5 years) and residential

instability (C3 moves) were 16 % (OR 1.16, 95 % CI 1.01,

1.33) and 25 % (OR 1.25, 95 % CI 1.00, 1.56) more likely

to report NMUPD compared to non-mobile adolescents (0

moves). Low-mobile adolescents were 18 % (OR 1.18,

95 % CI 1.01, 1.38) more likely to abuse pain relievers,

specifically. No relationship was found between moving

and tranquilizer, stimulant, or sedative use.

Discussion Increasing childhood residential mobility is

associated with NMUPD; therefore, efforts to prevent

NMUPD should target mobile adolescents. Further exam-

ination of the psychological effects of moving and its

association with pain reliever abuse is indicated.

Keywords Residential mobility � Adolescents � Health behavior � Prescription drug misuse � Opioids

Significance

What is already known on this subject? Among adoles-

cents, childhood residential instability is linked to devel-

opmental and social-emotional issues, including substance

use. However, the relationship between residential mobility

and non-medical use of prescription drugs (NMUPD) in

particular is unknown.

What this study adds? Residentially mobile adolescents

are more likely to use prescription drugs non-medically

than non-mobile adolescents. Drug-specific associations

indicate a relationship between low-mobile adolescents

and pain reliever abuse. This suggests that increasing

childhood residential mobility is associated with increased

NMUPD.

& Meagan E. Stabler [email protected]

1 Department of Epidemiology, Robert C. Byrd Health

Sciences Center, West Virginia University School of Public

Health, P.O. Box 9190, Morgantown, WV 26506-9190, USA

2 Department of Epidemiology, Injury Control Research

Center, West Virginia University School of Public Health,

Research Ridge, 3606 Collins Ferry Road, Suite 201,

Morgantown, WV 26505, USA

3 Department of Behavioral Medicine and Psychiatry, Chestnut

Ridge Center, West Virginia University School of Medicine,

930 Chestnut Ridge Road, Morgantown, WV 26505, USA

123

Matern Child Health J (2015) 19:2646–2653

DOI 10.1007/s10995-015-1785-z

Introduction

Prescription drug abuse is a public health epidemic [24].

One in every 20 Americans take prescription drugs for non-

medical use, resulting in 15,000 deaths annually [4].

American youth have also felt the impact of this epidemic.

One in every three people aged 12 years or older who ini-

tiated drug use in 2009 did so with non-medical use of

prescription drugs (NMUPD) [24]. Prescription medications

are second to cannabis as the most commonly abused illicit

drugs among youth aged 12–17, with an estimated 759,000

individuals who reported current NMUPD [13, 30].

In addition to high rates of NMUPD, the US has con-

sistently high residential mobility rates. In 2012, about 36.5

million Americans, 1 year of age or older, moved [32].

Adverse developmental-, behavioral-, social-, emotional-,

and health-related outcomes are attributed to childhood

residential mobility. Long-term effects that continue into

adulthood include depression, lack of continuity of health

care, poorer well-being, psychosocial stress, exhaustion,

and lack of consistency in personality characteristics [3,

18–20, 25]. Childhood residential mobility is a risk factor

for late-adolescent substance use disorders, particularly

alcohol, marijuana, and nicotine-dependence [3, 17, 31].

An increased number of residential moves in childhood and

adolescence is also associated with early initiation of illicit

drugs [7]. However, it is not known specifically how resi-

dential mobility influences adolescent NMUPD. Thus, the

purpose of this study was to explore the association

between childhood residential mobility and NMUPD

among American adolescents.

Methods

Data Source

The National Survey on Drug Use and Health (NSDUH) is

a nationally-representative survey that provides country

and state-level information on the distribution and deter-

minants of tobacco, alcohol, and illicit drug use (including

NMUPD) among non-active-duty Americans aged

12 years and older. NSDUH utilizes a complex, multi-level

sampling strategy described in detail elsewhere [22, 30,

33]. The survey includes a 1-h interview that is conducted

in each participant’s home using a laptop computer into

which most responses are entered by the participant.

Audio, computer-assisted, self-interviewing is utilized for

sensitive questions. Participants are provided $30 for

completing the interview.

For the current study, publicly-available, de-identified

data from the 2010 survey were analyzed. Thus, ethics

approval was not required as this research was not considered

human subjects research. A total of 68,487 participants

completed interviews. Of those, 18,614, were aged

12–17 years and completed the youth experiences ques-

tionnaire. Topics in this questionnaire include participant

physical and social environments; legal and illegal activities

and behaviors; accessibility of substances and substance

prevention programs; and personal attitudes, perceived

norms, and risk related to drug use.

Independent Variable

Residential mobility was assessed during the interview by

asking the question ‘‘How many times have you moved in

the past 5 years’’. In the NSDUH dataset, this categorical

variable has eight response options ranging from ‘‘none’’ to

‘‘six or more times’’, including ‘‘I don’t know’’. From this

question, the independent variable was created where

respondents were classified as ‘residentially immobile’ if

they reported zero moves, ‘low mobility’ if they reported

one or two moves, and ‘residentially unstable’ if they

reported three or more moves. This coding is consistent

with other research [1, 10, 28] in which residential insta-

bility is defined as three or more childhood or adolescent

moves [1, 2, 15, 18, 35].

Dependent Variables

The primary outcome of interest for this study was

NMUPD. During the NSDUH interview, participants were

asked if they had ever used tranquilizers, pain relievers,

sedatives, and/or stimulants non-medically. Participants

were instructed to only report NMUPD if the drug was not

prescribed to them or if they ‘‘took the drug for the expe-

rience or feeling it caused’’ [33]. They were asked to only

report on drugs that require a doctor’s prescription, not

over-the-counter drugs. A picture and name of each pill/

brand was shown to the participant to improve participant

recall. For example, non-medical use of prescription pain

relievers is ascertained from the following question:

‘‘[Have you] ever used pain reliever non-medically?’’ The

outcome variable was created from responses to questions

regarding NMUPD that were asked throughout the survey.

Participants were classified as ‘yes’ for NMUPD if they

endorsed ever using a prescription drug non-medically at

least once during the interview or ‘no’ for NMUPD if they

never endorsed using a prescription drug non-medically.

In addition to the overall NMUPD variable (any pre-

scription drug, regardless of class), individual variables

were created for each of the four NSDUH categories of

prescription drugs, i.e. tranquilizers, pain relievers, seda-

tives, and stimulants. Participants were classified as ‘yes’

for drug-specific NMUPD if they endorsed ever using any

of the drugs in the specific category non-medically at least

Matern Child Health J (2015) 19:2646–2653 2647

123

once during the interview and ‘no’ for drug-specific

NMUPD if they never endorsed using any drugs in that

category non-medically during the interview.

Covariates

To calculate the least-biased estimate of the effect of

NMUPD on childhood residential mobility, confounding

by a number of factors was explored. Socio-demographic,

intrapersonal, interpersonal, and community factors likely

to influence adolescent prescription drug misuse were

identified utilizing the social-ecological model as a theo-

retical framework. This model conceptualizes the interplay

between intrapersonal-, interpersonal-, and community-

level factors and their effect on health outcomes [21]. The

current study utilized a three-level, modified social-eco-

logical framework to select potential confounders a priori.

Socio-demographic factors used in the analysis included

race/ethnicity, sex, age, and financial assistance. Race/

ethnicity was categorized as ‘non-Hispanic white’; ‘non-

Hispanic black’; ‘Native American, Alaskan, Hawaiian, or

other Pacific Islander’; ‘non-Hispanic Asian’; ‘non-His-

panic, mixed race’; or ‘Hispanic, any race’. Participants

were classified ‘yes’ for financial assistance if they

endorsed participating in one or more government assis-

tance programs, including supplemental security income,

food stamps, cash assistance, and/or non-cash assistance.

Intrapersonal-level factors included in the analysis were

perception of health, perceived risk of drug use, sensation

seeking behavior, bonding to school, average grades,

delinquent behavior, and other drug use. Using a 5-point

scale ranging from ‘‘excellent’’ to ‘‘poor’’, participants

were asked to rate their current health. Responses of ‘‘ex-

cellent’’, ‘‘very good’’, and ‘‘good’’ were classified as

positive perceived health. Responses of ‘‘fair’’ and ‘‘poor’’

were classified as negative perceived health. Perceived risk

of drug use was classified as ‘present’ if respondents

answered ‘‘great risk’’ to eleven drug related activities

(e.g., ‘‘smoke 1? packs of cigarettes per day’’). Sensation-

seeking behavior was classified as ‘present’ if the partici-

pant endorsed that they agree with the following statements

‘‘sometimes’’ or ‘‘always’’: ‘‘I get a real kick out of doing

dangerous things’’ and ‘‘I like to test myself by doing risky

things’’. School bonding was classified as ‘high’ if

respondents reported participating in more than one

school/community-based youth activity within the previous

school year or if they positively responded to questions

regarding attitudes and feelings towards school. Partici-

pants were also classified based on their average grades as

passing (C or above) or failing (below a C). Delinquent

behavior was considered ‘present’ if the participant

reported activity in six antisocial scenarios (e.g., ‘‘had a

serious fight at school/work’’). Use of other substances was

classified as ‘present’ if the participant reported ‘‘yes’’ to

smoking part or all of a cigarette in the past 30 days,

having ever had a drink of alcohol (excluding sips or a

couple of drinks from someone else’s alcoholic drink), and

having ever used crack, heroin, LSD, PCP, mescaline,

psilocybin (i.e., mushrooms), ecstasy, or any other

hallucinogens.

Interpersonal-level covariates include parental monitor-

ing, parental disapproval of drug use, perceived peer drug

use, and peer disapproval of drug use. Parental monitoring

was classified as ‘present’ if the participant responded

‘‘always’’ or ‘‘sometimes’’ for six example scenarios

reflecting parental monitoring in the previous year (e.g.,

‘‘my parents check if homework is done’’). Parental dis-

approval of drug use was classified as ‘present’ when

participants perceived their parents to ‘‘strongly disap-

prove’’ of four drug-related activities (e.g., ‘‘smoking a

pack of cigarettes per day’’). Peer disapproval of drug use

was classified as ‘present’ if participants reported one of

their close friends ‘‘strongly’’ or ‘‘somewhat’’ disapproving

of four drug-related activities (e.g., ‘‘have one to two

alcoholic drinks per day’’). Peer drug use was classified as

‘yes’ if the participant reported that ‘‘most’’ or ‘‘all’’ stu-

dents in their grade smoke cigarettes, use marijuana, drink

alcoholic drinks, or get drunk at least once a week.

Community-level factors were availability of drugs,

school norms against drug use, and community norms

against drug use. Availability of drugs was classified as

‘yes’ if participants reported being approached by someone

selling drugs or if they perceived their ability to obtain

marijuana, cocaine, crack, heroin, or LSD as ‘‘fairly easy’’

or ‘‘very easy’’. School and community norms against drug

use were classified as ‘yes’ if participants reported ‘‘yes’’ to

having had any drug education in school or if they saw

drug prevention messages outside of school.

Statistical Analysis

Descriptive statistics and Chi square tests were utilized to

describe the sample and compute population estimates of

predetermined variables by drug use. To measure the

association between residential mobility and NMUPD,

multivariable logistic regression models were fit, account-

ing for the complex survey design. Due to oversampling in

the survey design, weights were utilized during the anal-

ysis. A sensitivity analysis was conducted to explore

missing data among the covariates for participants with

complete data and participants with incomplete data; there

were no significant differences in the exposure variable,

outcome variable, or exposure-outcome association

between the complete and incomplete cases. Then a com-

plete-case, multivariable analysis was performed, during

which only participants without missing data (n = 15,745)

2648 Matern Child Health J (2015) 19:2646–2653

123

for all of the factors under study were included in all of the

descriptive and multivariable analyses. When assessing for

confounding, conducting a complete-case analysis enables

attributing changes in the effect estimates to confounding

rather than differences in the participants included in the

analysis.

Four models were fit to estimate the association between

residential mobility and NMUPD for all prescription drug

types, while adjusting for the different groups of social-

ecological covariates. The first model included the expo-

sure and outcome variables of interest, adjusting for only

demographic variables. In the second model, the intraper-

sonal-level variables were added to the model with the

demographic variables. In the third model, the interper-

sonal-level variables were added to the model with the

demographic and intrapersonal-level variables. In the final

model, all of the variables, including the community-level

variables were incorporated. All analyses were conducted

using SAS� 9.3.

Results

Of the 15,745 participants aged 12–17 years included in

this study, 54 % (n = 8594) were non-mobile, 33 %

(n = 5081) were low-mobile, and 13 % (n = 2070) were

residentially instable (Table 1). Among the 10 %

(n = 1642) of adolescents who reported using drugs for

non-medical purposes, most moved, were female, older,

relied on financial assistance, and were Native American,

Alaskan, Hawaiian or other Pacific Islander or non-His-

panic white, compared to adolescents who did not report

NMUPD. Adolescents who reported NMUPD were sig-

nificantly more likely to perceive their health to be fair or

poor, engage in sensation seeking behaviors, report low

school bonding, have lower grades on average, exhibit

delinquent behavior, use other drugs, lack parental moni-

toring, perceive parents and peers to not disapprove of drug

use, have friends who use drugs, have access to drugs, and

do not perceive school norms to be against drug use,

compared to adolescents who did not report NMUPD.

Adolescents with low mobility were 26 % (unadjusted

OR 1.26, 95 % CI 1.20, 1.41) and adolescents with resi-

dential instability were 81 % (unadjusted OR 1.81, 95 %

CI 1.49, 2.21) more likely to use prescription drugs for

non-medical purposes than their non-mobile counterparts

(Table 2). This association persists after controlling for

socio-demographic and social-ecological factors. When

controlling for socio-demographic factors only, the odds of

NMUPD among low-mobile and residentially unstable

adolescents were 32 and 94 % greater, respectively, than

the odds of NMUPD among non-mobile adolescents. When

controlling for social-ecological factors via a four step

Table 1 Individual-, intrapersonal-, interpersonal-, and school-level characteristics of study sample, NSDUH, 2010 (n = 15,745)

Total Drug use No drug use

N (%) N (%) N (%)

Number of moves*

0 8594 (54) 747 (46) 7847 (55)

1–2 5081 (33) 582 (35) 4499 (33)

C3 2070 (13) 313 (19) 1757 (13)

Individual characteristics

Race/ethnicity*

Non-Hispanic 9767 (60) 1007 (61) 8760 (60)

Non-Hispanic 1964 (14) 182 (12) 1782 (14)

Native American 271 (1) 43 (1) 228 (1)

Non-Hispanic 487 (4) 23 (2) 464 (4)

Non-Hispanic 653 (2) 90 (3) 563 (2)

Hispanic 2603 (19) 297 (21) 2306 (19)

Sex*

Male 7949 (51) 744 (47) 7205 (52)

Female 7796 (49) 898 (53) 6898 (48)

Age*

12 2133 (14) 78 (5) 2055 (15)

13 2439 (16) 114 (7) 2325 (17)

14 2558 (16) 204 (12) 2354 (17)

15 2772 (18) 345 (21) 2427 (17)

16 2901 (18) 411 (23) 2490 (18)

17 2942 (19) 490 (32) 2452 (17)

Financial assistance*

Yes 3550 (23) 492 (29) 3058 (22)

No 12,195 (77) 1150 (71) 11,045 (78)

Intrapersonal characteristics

Perception of overall health*

Fair/poor 547 (3) 101 (6) 446 (3)

Good/excellent 15,198 (97) 1541 (94) 13,657 (97)

Perceived risk of drug use

Present 15,202 (97) 1586 (97) 13,616 (97)

Absent 516 (3) 55 (3) 461 (3)

Sensation seeking behavior*

Present 4121 (26) 790 (47) 3331 (23)

Absent 11,624 (75) 852 (53) 10,772 (77)

Bonding to school*

High 15,645 (100) 1608 (99) 14,037 (100)

Low 100 (1) 34 (1) 66 (0)

Average grades*

A, B or C 14,915 (95) 1445 (89) 13,470 (96)

D or F 830 (5) 197 (11) 633 (4)

Delinquent behavior*

Present 4806 (30) 967 (58) 3839 (27)

Absent 10,939 (70) 675 (42) 10,264 (73)

Other drug use*

Yes 6440 (40) 1377 (84) 5063 (35)

No 9305 (60) 265 (16) 9040 (65)

Matern Child Health J (2015) 19:2646–2653 2649

123

modeling process, the association again persists but is

attenuated with odds ratios ranging from 1.32 (controlling

for socio-demographic and intrapersonal-level factors) to

1.16 (controlling for socio-demographic, intrapersonal-,

interpersonal-, and community-level factors) for adoles-

cents with low mobility; in contrast, residentially unstable

adolescents had attenuated odds ratios ranging from 1.94 to

1.25. After controlling for all social-ecological factors,

low-mobile adolescents were 16 % (OR 1.16, 95 % CI

1.01, 1.33) and residentially unstable adolescents were

25 % (OR 1.25, 95 % CI 1.00, 1.56) more likely to use

prescription drugs for non-medical purposes than their non-

mobile counterparts.

With regard to specific categories of prescription drugs,

low-mobile adolescents were 18 % (OR 1.18, 95 % CI

1.01, 1.38) more likely to abuse pain relievers than their

non-mobile counterparts; this association was not signifi-

cant for residentially unstable adolescents. The association

between moving and NMUPD was not significant for

tranquilizers, stimulants, or sedatives alone (Table 3).

Discussion

In a nationally-representative sample of US adolescents,

low-mobile and residentially unstable adolescents were

more likely to use prescription drugs for non-medical

purposes than non-mobile adolescents. The magnitude of

the association increased with more moves and persisted

even after controlling for traditional confounders [5, 34],

which were selected and organized using the social-eco-

logical model. Pain relievers were the only specific cate-

gory of prescription drugs that low mobility adolescents

were significantly more likely to abuse than non-mobile

adolescents.

These results are congruent with existing literature. The

current study, along with previous studies [8, 14], found

adolescents who are female, white, have depressed socioe-

conomic status, poor perceived health, poor academic

Table 1 continued

Total Drug use No drug use

N (%) N (%) N (%)

Interpersonal characteristics

Parental monitoring*

Present 14,499 (92) 1447 (88) 13,052 (92)

Absent 1246 (8) 195 (12) 1051 (8)

Parental disapproval of drug use*

Present 15,271 (97) 1468 (91) 13,803 (98)

Absent 474 (3) 174 (9) 300 (2)

Perceived peer drug use*

Yes 7705 (49) 1277 (78) 6428 (46)

No 8040 (51) 365 (22) 7675 (54)

Peer disapproval of drug use*

Present 14,503 (92) 1290 (79) 13,213 (94)

Absent 1242 (8) 352 (21) 890 (6)

School-level characteristics

Availability of drugs*

Yes 8779 (56) 1387 (84) 7392 (52)

No 6966 (44) 255 (16) 6711 (48)

School norms against drug use*

Yes 11,902 (76) 1135 (68) 10,767 (76)

No 3843 (24) 507 (32) 3336 (24)

Community norm against drug use

Yes 12,193 (77) 1232 (76) 10,961 (77)

No 3497 (23) 408 (24) 3089 (23)

n, number of participants; %, weighted percent (takes into account

NSDUH’s complex survey design and does not include missing

values)

* Significant (p\ 0.05) Rao–Scott Chi Square value

Table 2 Association between childhood residential mobility and non-medical use of prescription drugs, NSDUH 2010 (n = 15,745)

OR (95 % CI) Model 1b Model 2c Model 3d Model 4e

aOR (95 % CI) aOR (95 % CI) aOR (95 % CI) aOR (95 % CI)

Moves in previous 5 yearsa

1–2 versus 0 1.26 (1.20, 1.41) 1.32 (1.17, 1.50) 1.18 (1.04, 1.35) 1.16 (1.01, 1.34) 1.16 (1.01, 1.33)

C3 versus 0 1.81 (1.49, 2.21) 1.94 (1.57, 2.39) 1.35 (1.09, 1.68) 1.27 (1.02, 1.60) 1.25 (1.00, 1.56)

NSDUH National Study on Drug Use and Health, n number of participants, OR unadjusted odds ratio, aOR adjusted odds ratio, CI confidence

interval a Respondents were classified based on 0 moves (non-mobile), 1–2 moves (low mobility), and C3 moves (residential instability) b Model 1 adjusts for race, sex, age, and financial assistance c Model 2 adjusts for Model 1 covariates ? perceived overall health, bonding to school, sensation seeking behavior, average grades, delinquent

behavior, and other drug use d Model 3 adjusts for Model 2 covariates ? parental monitoring, parental disproval of drug use, peer drug use, and peer disapproval of drug use e Model 4 adjusts for Model 3 covariates ? availability of drugs and school norms against drug use

2650 Matern Child Health J (2015) 19:2646–2653

123

performance, delinquent behavior, use other drugs, lack peer

and parental disapproval of drug use, and have drugs

available to them are more likely to use prescription drugs

non-medically. Also, the increasing magnitude of the asso-

ciation between residential mobility and NMUPD with

increased instability is similar to those observed of other

illicit drug use [7].

Short- and long-term effects of childhood residential

mobility have been linked to adolescent alcohol, marijuana,

and nicotine use [3, 7, 17, 31]. Additionally, increased

numbers of moves before 16 years of age have been shown

to be associated with early initiation of illicit drugs,

including marijuana, hallucinogens, crack/cocaine, and illi-

cit use of prescribed drugs [7]. Yet in previous research,

prescription drug misuse was not broken down into drug

types and illicit use of prescription drugs was only signifi-

cantly associated with moving four or more times before the

age of 16 years. The current study, however, showed an

association between different levels of residential mobility

and NMUPD via drug-specific categories; i.e., low-mobile

adolescents were at increased odds of prescription pain

reliever abuse compared to their non-mobile counterparts.

After controlling for potential confounders, this relationship

was not significant for other drug-specific categories nor for

residentially unstable adolescents. This study may have been

underpowered to detect small differences in the other drug

types due to the lower prevalence of abuse of these classes

of drugs, compared to pain relievers, and the lower preva-

lence of frequent residential moves within our sample.

The influence of moving on adolescent drug use could

have numerous explanations. Usually children have little

control over the decision to move. This lack of control over

a major life event could result in adolescents feeling pow-

erless, frustrated, and lonely [23]. Loneliness resulting from

moving is common and can have serious implications for

adolescents who need supportive social and normative

structures [29]. Close social structures aid in adolescent self-

esteem and social competence [6, 16]. Absence of support to

deal with stress and anxiety could make these adolescents

more vulnerable to deviant behavior, such as drug use ini-

tiation [7]. Many changes at once (e.g., schools, friends,

neighborhood, neighborhood surroundings) could be stress-

ful. These adolescents are forced to adjust quickly. Many

studies have shown that youth who transfer to a new school

are more likely to exhibit academic and behavioral problems

[11, 26, 27]. Another reason relocating is associated with

adolescent drug use could be the time and energy demands

placed on the parents associated with moving. DeWit [7]

purport that parental distraction, resulting from moving,

could result in less parental supervision, which could prompt

children to seek acceptance from peers. Peer delinquency

and best-friend delinquency are strong risk factors for ado-

lescent substance abuse [12].

These findings suggest that social-ecological factors

confound the association between residential mobility and

NMUPD. There was a substantial increase in the odds ratio

between models 1 and 2, which is evidence of confounding

by intra-personal factors. Future studies, therefore, should

further establish the association between intra-personal

factors and NMUPD.

This study has several strengths and limitations. This

analysis utilizes a large, nationally-representative sample

of US adolescents. The interviewing methodology has been

tested and utilized over several years. Because participants

are asked to report their own historic drug use, their

responses are dependent upon accurate recall and honest

answers. Though these data were obtained through self-

report, which has the above-mentioned limitations, these

data are not available from an alternative source. Further-

more, the manner in which sensitive data are collected, by

way of direct entry into a study computer and the different,

repeated ways in which questions are asked about the same

topic, ensure that the data are as accurate as possible. This

study is the first to examine the association between resi-

dential mobility and different categories of prescription

drugs. In addition, the participants were only asked to

Table 3 Association between childhood residential mobility and NMUPD by drug class, NSDUH, 2010 (n = 15,745)

NMUPDb Pain relieversb Tranquilizersb Stimulantsb Sedativesb

OR (95 % CI) aOR (95 % CI) aOR (95 % CI) aOR (95 % CI) aOR (95 % CI)

Moves in previous 5 yearsa

1–2 versus 0 1.16 (1.01, 1.33) 1.18 (1.01, 1.38) 1.07 (0.79, 1.45) 1.08 (0.75, 1.55) 1.33 (0.71, 2.50)

C3 versus 0 1.25 (1.00, 1.56) 1.12 (0.89, 1.42) 1.28 (0.92, 1.79) 1.35 (0.85, 2.13) 1.03 (0.54, 1.96)

NMUPD non-medical use of prescription drugs, NSDUH National Study on Drug Use and Health, n number of participants, aOR adjusted odds

ratio, CI confidence interval a Respondents were classified based on 0 moves (non-mobile), 1–2 moves (low mobility), and C3 moves (residential instability) b Model is adjusted for race, sex, age, financial assistance, perceived overall health, bonding to school, sensation seeking behavior, average

grades, delinquent behavior, other drug use, parental monitoring, parental disproval of drug use, peer drug use, peer disapproval of drug use,

availability of drugs, and school norms against drug use

Matern Child Health J (2015) 19:2646–2653 2651

123

recall residential mobility that took place within the past

5 years, thus potentially reducing errors in reporting of

residential mobility. A limitation of using secondary data

to answer this research question, however, is that the def-

inition of residential mobility employed was dependent on

the questions asked during the survey, preventing the

consideration of the effect of the distance between moves

on the outcome. Furthermore, because the data are cross-

sectional in nature, the temporal relationship between

moving in the last 5 year and lifetime use of NMUPD is

unclear. Given the age range of the participants, however,

the likelihood of initiation of drug use prior to any moving

in the previous 5 years is less likely than in an older

population.

In addition, we were unable to control for adverse

childhood experiences (ACE), which could confound the

association between frequency of moves and NMUPD.

Potentially important ACE factors (e.g., parental divorce,

abuse, and family dysfunction) were not included in the

NSDUH interview. ACE have been studied with regard to

their influence on child health, their continued impact on

adult health, alcohol initiation, drug addiction, and lifetime

drug use [8, 9]. Though the influence of ACE on the

relationship between frequent childhood and adolescent

moves and NMUPD has not been studied, the current study

did explore numerous variables under the application of the

social-ecological model. Variables exploring financial sta-

bility, perceived health and risk of drug use, parent moni-

toring, delinquent behaviors, and parental/peer use and

disapproval of drugs may serve as proxy measures for

certain ACE factors, for which the current study was

unable to control.

Conclusion

Based on these findings, childhood residential mobility is

associated with adolescent NMUPD, with the magnitude of

the association increasing with more moving. Future

research should further examine the association between

moving and NMUPD, particularly with regard to the

direction of the association and the types of drugs involved.

In addition, interventions to prevent prescription drug

misuse and abuse among adolescents should be developed

for and targeted towards mobile adolescents and the par-

ticular challenges faced by this group. Given that pre-

scription drugs are most commonly obtained from family

and friends, and the influence that parents have on ado-

lescent initiation of drug use, future interventions should

also target parents of residentially mobile adolescents.

Acknowledgments Kelly Gurka was partially supported by the Centers for Disease Control and Prevention, National Center for

Injury Prevention and Control, Grant R49CE002109. The content is

solely the responsibility of the authors and does not necessarily rep-

resent the official views of the CDC.

Conflict of interest The authors declare no conflict of interest.

References

1. Brown, D., Benzeval, M., Gayle, V., Macintyre, S., O’Reilly, D.,

& Leyland, A. H. (2012). Childhood residential mobility and

health in late adolescence and adulthood: Findings from the West

of Scotland Twenty-07 Study. Journal of Epidemiology and

Community Health, 66(10), 942–950.

2. Bures, R. M. (2003). Childhood residential stability and health at

midlife. American Journal of Public Health, 93(7), 1144–1148.

3. Buu, A., diPaazza, C., Jing, W., Puttler, L. I., Fitzgerald, H. E., &

Zucker, R. A. (2009). Parent, family, and neighborhood effects on

the development of child substance use and other psychopathol-

ogy from preschool to the start of adulthood. Journal of Studies

on Alcohol and Drugs, 70(4), 489–498.

4. Centers for Disease Control and Prevention. (2012). Policy

impact: Prescription painkiller overdoses. Retrived from http://

www.cdc.gov/homeandrecreationalsafety/rxbrief/

5. Collins, D., Abadi, M. H., Johnson, K., Shamblen, S., &

Thompson, K. (2011). Non-medical use of prescription drugs

among youth in an appalachian population: Prevalence, predic-

tors, and implications for prevention. Journal of Drug Education,

41(3), 309–326.

6. Cornille, T. A. (1993). Support systems and the relocation pro-

cess for children and families. Marriage and Family Review,

19(3/4), 281.

7. DeWit, D. J. (1998). Frequent childhood geographic relocation:

Its impact on drug use initiation and the development of alcohol

and other drug-related problems among adolescents and young

adults. Addictive Behaviors, 23(5), 623–634.

8. Dong, M., Anda, R. F., Felitti, V. J., Williamson, D. F., Dube, S.

R., Brown, D. W., & Giles, W. H. (2005). Childhood residential

mobility and multiple health risks during adolescence and

adulthood: The hidden role of adverse childhood experiences.

Archives of Pediatrics and Adolescent Medicine, 159(12),

1104–1110. doi:10.1001/archpedi.159.12.1104.

9. Dube, S. R., Felitti, V. J., Dong, M., Chapman, D. P., Giles, W.

H., & Anda, R. F. (2003). Childhood abuse, neglect, and

household dysfunction and the risk of illicit drug use: The

adverse childhood experiences study. Pediatrics, 111(3),

564–572.

10. Ersing, R., Sutphen, R., & Loeffler, D. (2009). Exploring the

impact and implications of residential mobility: From the

neighborhood to the school. Advances in Social Work, 10(1),

1–18.

11. Felner, R. D., Ginter, M., & Primavera, J. (1982). Primary pre-

vention during school transitions: Social support and environ-

mental structure. American Journal of Community Psychology,

10(3), 277–290.

12. Fite, P. T., Vitulano, M., Elkins, S., Grassetti, S., & Wimsatt, A.

(2012). Perceived best friend delinquency moderates the link

between contextual risk and factors and juvenile delinquency.

Journal of Community Psychology, 40(6), 747–761.

13. Forum on Child and Family Statistics. (2013). Pop1 child popu-

lation: Number of children (in millions) ages 0–17 in the United

States by age, 1950–2013 and projected 2014–2050. ChildStats.

Retrived from http://www.childstats.gov/americaschildren/tables/

pop1.asp

14. Gasper, J., DeLuca, S., & Estacion, A. (2010). Coming and

going: Explaining the effects of residential and school mobility

2652 Matern Child Health J (2015) 19:2646–2653

123

on adolescent delinquency. Social Science Research, 39(3),

459–476. doi:10.1016/j.ssresearch.2009.08.009.

15. Gilman, S. E., Kawachi, I., Fitzmaurice, G. M., & Buka, L.

(2003). Socio-economic status, family disruption and residential

stability in childhood: Relation to onset, recurrence and remission

of major depression. Psychological Medicine, 33(8), 1341–1355.

16. Hendershott, A. B. (1989). Residential mobility, social support

and adolescent self-concept. Adolescence, 24(93), 217–232.

17. Hoffmann, J. P. (2002). The community context of family

structure and adolescent drug use. Journal of Marriage and

Family, 64(2), 314–330.

18. Jelleyman, T., & Spencer, N. (2008). Residential mobility in

childhood and health outcomes: A systematic review. Journal of

Epidemiology and Community Health, 62(7), 584–592. doi:10.

1136/jech.2007.060103.

19. Lin, K.-C., Twisk, J. W. R., & Huang, H.-C. (2012). Longitudinal

impact of frequent geographic relocation from adolescence to

adulthood on psychosocial stress and vital exhaustion at ages 32

and 42 years: The Amsterdam Growth and Health Longitudinal

Study. Journal of Epidemiology, 22(5), 469–476.

20. Lin, K. C., Twisk, J. W., & Rong, J. R. (2011). Longitudinal

interrelationships between frequent geographic relocation and

personality development: Results from the Amsterdam Growth

and Health Longitudinal Study. American Journal of Orthopsy-

chiatry, 81(2), 285–292. doi:10.1111/j.1939-0025.2011.01097.x.

21. McLaren, L., & Hawe, P. (2005). Ecological perspectives in

health research. Journal of Epidemiology and Community Health,

59(1), 6–14. doi:10.1136/jech.2003.018044.

22. Morton, K. B., Aldworth, J., Chromy, J. R., Foster, M. S., Hirsch,

E. L., & Kott, P. (2009). 2010 National Survey on Drug Use and

Health: Sample design plan. Prepared for the Substance Abuse

and Mental Health Services Administration, Office of Applied

Studies, under Contract No. 283-08-0210, Phase I, Deliverable

No. 7, RTI/0211838.103. Research Triangle Park, NC: RTI

International.

23. Newcomb, M. D., & Harlow, L. L. (1986). Life events and

substance use among adolescents: Mediating effects of perceived

loss of control and meaninglessness in life. Journal of Personality

and Social Psychology, 51(3), 564–577.

24. Office of National Drug Control Policy. (2013). Prescription

drug abuse. Retrived from http://www.whitehouse.gov/ondcp/

prescription-drug-abuse

25. Oishi, S., Lun, J., & Sherman, G. D. (2007). Residential mobility,

self-concept, and positive affect in social interactions. Journal of

Personality and Social Psychology, 93(1), 131–141. doi:10.1037/

0022-3514.93.1.131.

26. Seidman, E., Allen, L., Aber, J. L., Mitchell, C., & Feinman, J.

(1994). The impact of school transitions in early adolescence on

the self-system and perceived social context of poor urban youth.

Child Development, 65(2 Spec No), 507–522.

27. Simmons, R. G., Burgeson, R., Carlton-Ford, S., & Blyth, D. A.

(1987). The impact of cumulative change in early adolescence.

Child Development, 58(5), 1220–1234.

28. Simpson, G. A., & Fowler, M. G. (1994). Geographic mobility

and children’s emotional/behavioral adjustment and school

functioning. Pediatrics, 93(2), 303–309.

29. Stack, S. (1994). The effect of geographic mobility on premarital

sex. Journal of Marriage and Family, 56(1), 204–208.

30. Substance Abuse and Mental Health Services Administration.

(2011). Results from the 2010 National Survey on Drug Use and

Health: Summary of national findings. HHS Publication No.

(SMA) 11-4658, NSDUH Series H-41. Rockville, MD: Substance

Abuse and Mental Health Services Administration.

31. Trim, R. S., & Chassin, L. (2008). Neighborhood socioeconomic

status effects on adolescent alcohol outcomes using growth

models: Exploring the role of parental alcoholism. Journal of

Studies on Alcohol and Drugs, 69(5), 639–648.

32. US Census Bureau. (2012). Census Bureau reports national

mover rate increases after a record low in 2011. Retrieved from

http://www.census.gov/newsroom/releases/archives/mobility_of_

the_population/cb12240.html

33. US Department of Health and Human Services, Substance Abuse

and Mental Health Services Administration, Center for Behav-

ioral Health Statistics and Quality. (2013). National Survey on

Drug Use and Health, 2010. Retrieved from doi:10.3886/

ICPSR32722.v4

34. Viana, A. G., Trent, L., Tull, M. T., Heiden, L., Damon, J. D.,

Hight, T. L., & Young, J. (2012). Non-medical use of prescription

drugs among Mississippi youth: Constitutional, psychological,

and family factors. Addictive Behaviors, 37(12), 1382–1388.

doi:10.1016/j.addbeh.2012.06.017.

35. Wood, D., Halfon, N., Scarlata, D., Newacheck, P., & Nessim, S.

(1993). Impact of family relocation on children’s growth,

development, school function, and behavior. Journal of the

American Medical Association, 270(11), 1334–1338.

Matern Child Health J (2015) 19:2646–2653 2653

123

Copyright of Maternal & Child Health Journal is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.