Sample

The study data were drawn from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) [23], [24]. The NESARC is a nationally representative survey of 43093 adults in the United States. The nationwide face-to-face survey was conducted in 2001 and 2002 by the US Census Bureau and the National institute on Alcohol Abuse and Alcoholism. The overall response rate for the 2001–2002 NESARC was 81 percent. The survey provides data on alcohol and drug use, psychiatric classification of SUDs and other psychiatric disorders, treatment utilization, and sociodemographic information. The NESARC procedures were reviewed and approved by the US Census Bureau and the US Office of Management and Budget. Informed consent was obtained from all NESARC participants.

Measures

Lifetime substance use disorders.

Diagnoses of lifetime SUDs were made by the Alcohol Use Disorder and Associated Disabilities Interview Schedule–DSM-IV Version (AUDADIS-IV) [25], which was developed based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) [26]. The AUDADIS-IV has shown to be reliable and valid for assessing SUDs [27]–[30]. In a general population sample testing the AUDADIS-IV, the reliabilities were good to excellent for DSM-IV lifetime alcohol abuse and dependence (kappa = 0.70), lifetime nicotine dependence (kappa = 0.60), any drug abuse and dependence (kappa = 0.66–0.79), cannabis abuse and dependence (kappa = 0.71–0.78), cocaine abuse and dependence (kappa = 0.68–0.91), and heroin abuse and dependence (kappa = 0.66–0.80) [28], [29].

Lifetime SUDs evaluated were as follows: (1) nicotine dependence; (2) alcohol abuse or dependence; (3) cannabis abuse or dependence; (4) cocaine abuse or dependence; (5) opioid (painkillers or heroin) abuse or dependence; (6) amphetamine abuse or dependence; (7) hallucinogen abuse or dependence; and (8) other drug (sedatives, tranquilizers, inhalants, solvents, or others) abuse or dependence. Although sedatives, tranquilizers, inhalants, and solvents were independently assessed, we have lumped into the category of “other drug” because of the small percentages of these drug disorders.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Logistic regression of adoption-specific risk factors associated with any lifetime SUD in all individuals (n = 42881).

https://doi.org/10.1371/journal.pone.0049655.t003

Statistical Analyses

Participants were classified into two groups based on adoption status (adopted and nonadopted). Prevalence rates of demographic variables and lifetime SUDs were obtained using cross-tabulations. Comparison of the two groups was performed for demographic variables to estimate odds ratios (OR) and 95% confidence intervals using a series of logistic regression. The variables (1) age, (2) race, (3) education, (4) personal income, and (5) marital status were coded as indicator variables using (1) ages of 45 or older, (2) white, (3) education of 0–11 years, (4) personal income of $0–19999, and (5) married/cohabitation as the reference category.

For main outcome variables (prevalence rates of lifetime SUDs), we conducted logistic regression to examine the effects of adoption status on lifetime SUDs. First, unadjusted odds ratios (OR) and 95% confidence intervals were estimated. Second, adjusted odds ratios (AOR) and 95% confidence intervals were estimated while adjusting for covariates (gender, age, race, education, and marital status). The covariates were selected based on the known correlates of adoption or of SUD.

To identify adoption-specific risk factors for any lifetime SUD, a series of logistic regression analyses were conducted using the entire sample. Interaction effects between adoption status and other predictor variables (gender, age, race, education, and marital status) were examined in logistic models to determine if any of these variables modified the effect of adoption status on SUD.

Demographic Characteristics

Of the 43093 adults in the NESARC, 42881 participants (99.5%) were included for data analysis after excluding 212 individuals who did not know their adoption status. The rate of adoptees (n = 378) in this study sample (n = 42881) was 0.88%.

Table 1 describes the demographic characteristics of two groups based on the adoption status. Adoptees were more likely than nonadoptees to be younger (ages of 30–44), Asian/Pacific Islanders, and to have higher levels of education. Adoptees were less likely than nonadoptees to be Hispanic, black, or widowed. The remaining demographic characteristics (gender, personal income) failed to show statistical differences.

Prevalence and Odds Ratios of Substance Use Disorders

Table 2 presents prevalence rates of lifetime SUDs in the two adoption groups. Prevalence of any lifetime SUD was 43% higher among adoptees (50.5%) compared to nonadoptees (35.4%). Adoptees had higher prevalence rates for every SUD compared to nonadoptees. Lifetime prevalence rates of licit SUDs were 41.0% (alcohol) and 25.4% (nicotine) among adoptees. In nonadoptees, the rates dropped to 27.5% (alcohol) and 16.1% (nicotine). Lifetime prevalence rates of illicit SUDs ranged from 2.9% (opioid) to 13.2% (cannabis) among adoptees and from 1.3% (opioid) to 7.6% (cannabis) among nonadoptees.

Table 2 also shows unadjusted odds ratios (OR), adjusted odds ratios (AOR), and 95% confidence intervals (95% CI). Logistic regression analyses showed that the odds of every lifetime SUD were statistically higher among adoptees than nonadoptees after adjusting for confounding variables. Overall, adoptees had a 1.87-fold increased risk of any lifetime SUD than nonadoptees (AOR 1.87, 95% CI 1.51–2.31). For licit SUD, adoptees had a 1.88-fold increased risk than nonadoptees (AOR 1.88, 95% CI 1.52–2.32). For illicit SUD, adoptees had a 1.89-fold increased risk than nonadoptees (AOR 1.89, 95% CI 1.43–2.50). The adjusted odds ratios of SUDs ranged from 1.77 (cannabis) to 3.14 (amphetamine).

Adoptees also had higher odds of both diagnostic categories of abuse and dependence compared to nonadoptees as follows: alcohol abuse (AOR 1.39, 95% CI 1.08–1.77), alcohol dependence (AOR 1.83, 95% CI 1.40–2.39), any illicit drug abuse (AOR 1.86, 95% CI 1.39–2.50), and any illicit drug dependence (AOR 2.08, 95% CI 1.30–3.32).

Adoption-Specific Risk Factors for Any Substance Use Disorders

Table 3 presents adoption-specific risk factors for any lifetime SUD among all adopted and nonadopted individuals (n = 42881). Logistic regression examined interaction effects between adoption status and risk factor variables (gender, age, race, education, and marital status). Two adoption-specific risk factors were significantly associated with any lifetime SUD: (1) Hispanic (OR 2.52, 95% CI 1.30–4.89) and (2) never married (OR 1.71, 95% CI 1.02–2.84). The strongest adoption-specific risk factor was Hispanic race-ethnicity, which was 2.52 times more likely associated with any lifetime SUD. The second adoption-specific risk factor was being never married, which increased the odds of any lifetime SUD by a factor of 1.71.

Odds Ratios for Substance Use Disorders

Our findings from the population-based NESARC data showed that adoptees have higher rates of lifetime SUDs than nonadoptees. These findings are consistent with two large European studies [6], [7], although the European data were based on intercountry adoptees and our data were based on predominantly domestic adoptees. The American NESARC rate of any lifetime SUD among adoptees (AOR 1.87, 95% CI 1.51–2.31) overlap with the rates of alcohol abuse (AOR 2.6, 95% CI 2.0–3.3) among Swedish intercountry adoptees [7] and the rate of substance abuse or dependence (AOR 2.05, 95% CI 1.32–3.17) among a cohort of intercountry adoptees in the Netherlands [6]. These comparisons among three countries indicate that the increased prevalence for SUD tends to hold true regardless of country of adoption (Sweden, Netherlands, United States) and mode of sampling adoptees (intercountry vs. all adoptees; cohort vs. cross section of all adoptees in one time period). One possible discrepancy could be drug abuse. The overall rate of illicit SUD in the NESARC data (AOR 1.89, 95% CI 1.43–2.50) did not overlap with the Swedish drug abuse rate (AOR 5.2, 95% CI 2.9–9.3). Concurrently, the NESARC rates for cocaine, amphetamine, hallucinogen, opioid, other drug abuse/dependence, and any illicit drug dependence did overlap with Swedish drug abuse rate, suggesting that type of drug, type of adoptee, and/or sampling method could produce even higher odds ratios for SUD. In particular, intercountry adoptees in the European studies could have higher odds of mental health problems than domestic adoptees in the NESARC study because of complicated intercountry adoption process, acculturation, and racial issues [31], as well as delays in adoption so that children are older. Although some disagreement in specific rates suggests national/cultural or methodological differences, these differences are more one of extent than of general directionality.

Importantly, adoptees’ odds ratios were high for both abuse and dependence (not just dependence alone). The odds ratios were slightly higher for dependence than abuse, but these differences were not significant on the 95% CI. This findings applied to both alcohol abuse/dependence as well as drug abuse/dependence. From a public health perspective, these data indicate that adoptees have a wide range of SUD including abuse and dependence.

These adoption studies convey both theoretical and practical implications. On a theoretical level, they indicate a linkage between being adopted and having a propensity to SUD years later in adulthood. The most likely linkage is a genetic one [32]–[35] with parental SUD genes (or perhaps related genes such as impulsivity [36]) favoring (1) a genetic propensity for SUD in their offspring and/or (2) selection of more severe parental SUD due to their inability to raise their children. Other biological factors include paternal fetal damage from impaired spermatogenesis, detrimental maternal intrauterine or childbirth events [21], and – in those adopted later – childhood exposure to drug use and adverse child raising [37]. On the contrary, adopted children tend to have adoptive families with social characteristics better than families at large, including higher education and higher socioeconomic status. In sum, adopted children tend to have a unique admixture of greater at-risk genetic factors and enhanced environmental factors in the adoptive family [17], [38]. Whatever the cause, being adopted should be recognized as a risk factor to SUD.

Adoption-Specific Risk Factors for Any Substance Use Disorders

Our study identified two risk factors increasing the prevalence of lifetime SUD specific to adoption status. The first and largest risk factor was Hispanic race-ethnicity (AOR 2.52, 95% CI 1.30–4.89). Factors in the Hispanic adoptees that might account for this greater vulnerability to SUD include a higher prevalence of foreign-born adoptees with more adverse intrauterine factors (poor nutrition, impoverished family-of-origin) and/or environmental factors in the family-of-adoption (e.g., adopted by Hispanic relatives with lower socioeconomic status as opposed to the usual families-of-adoption with higher socioeconomic status). Or conversely, Hispanic nonadoptees may possess protective factors against SUD, such as higher levels of parental monitoring against SUD [39].

The second risk factor was single marital status (AOR 1.71, 95% CI 1.02–2.84). However, we have no previous studies pointing to a higher single marital status as observed in this study. Other remaining demographic characteristics (gender, age, education) failed to show statistical differences. Although male gender was a risk factor for SUD in population studies [40], [41], it was not an “adoption-specific” risk factor for SUD in this study.

Limitations

Several limitations should be considered in interpreting our findings. First, the NESARC data did not provide information regarding the status of international or domestic adoption [10], [42]. We assumed it would be similar to percentages of international (15%) and domestic (85%) adoptees estimated previously in the United States [43]. Second, the NESARC also did not provide information on the age at adoption [3] or resilient factors such as self-esteem [44], which could play intermediary roles between adoptees and later SUD. Third, the NESARC did not collect information on SUD and other behaviors in adoptees’ biological parents. Fourth, the sample size of adoptees was disproportionately smaller than that of nonadoptees.

Conclusions

Adoptees had higher odds for lifetime SUDs than nonadoptees in this study using NESARC data. Despite the advantages of adoptees’ higher educational levels probably due to being raised by higher educated, higher income adopting parents, adoptees are still at higher risk to lifetime SUD. Awareness of adopted persons and their adoptive parents to this risk may help in primary prevention (never using substances; having conservative rules about doses and frequency of use) and in secondary prevention (being alert to early signs and symptoms; timely intervention to reduce damage and increase the chance of recovery). The findings can also be useful for clinicians and policy makers to provide education, prevention, and support for adoptees and their families.