Trends in Alcohol, Drug and Cigarette Use Among Haitian Youth in Miami-Dade County, Florida

Abstract

The objective of this article is to report on prevalence of drug and cigarette use among a segment of Haitian youth in the United States. The article is an argument in favor of contextualizing knowledge about drug use among young people across socioethnic lines. Because initiation of licit and illicit drugs tends to occur during adolescence, ethnic differentiation is crucial if we are to understand the drug experience among young people in the United States. Immigration, acculturation, and identity processes are critical in refuting the conventional racial categorization commonly used for interpretation of risks and behaviors among youth in the United States. The task of bringing empirical evidence to bear on drug use and drug choices by young people from different contexts will lead to the re-examination of patterns of drug use as well as to creative ways of conceptualizing these patterns.

INTRODUCTION

Epidemiological data on drug, alcohol and cigarette (ATOD) use among young people are mostly reported within the conventional classification of Blacks, Whites, and Hispanics, with the predominant emphasis on the distribution of substance use for Blacks and Whites. In turn, these epidemiological profiles provide the basis for designing prevention and intervention programs that address risks for subgroups within these dominant categories. The categories are problematic, however, in that there may be marked variability of substance use within dominant racial/ethnic categories. Therefore, the design of intervention and prevention models tailored for specific ethnic groups within a larger racial category also become problematic. The underlying premise that appears to govern the use of these conventional racial categories in drug studies and public health is that “race,” not sociohistorical and cultural processes, shapes choices of drugs or specific patterns of use.

For the Hispanic population in the United States, several investigators (Canino et al., 2003; Wallace, Bachman, O’Malley, Johnston, Schlenberg, & Cooper, 2002) have shown that data formulated in terms of broad “race/ethnic” categories tend to obscure variation by immigration status and national origin. This obscuring of variation is also true for data reported under the category “Black” or “African American,” in which ethnic, sociocultural, and national variability epistemologically challenges the applicability of programs designed after these reports for specific populations such as Haitian youth (for a case of alcohol use among Haitian and African American youth, see Strunin, 1999; Strunin & Demissie, 2001).

The objective of this article is to report on prevalence of drug and cigarette use among a segment of Black youth in the United States: Haitian/Haitian American youth (hereafter, Haitian youth). The article introduces descriptive data that aim to strengthen the argument in favor of contextualizing knowledge about drug use among young people across socioethnic lines. Because initiation of licit and illicit drugs tends to occur during adolescence, ethnic differentiation is crucial if we are to understand the drug experience among young people in the United States. Immigration, acculturation, and identity processes are critical in refuting the conventional racial categorization commonly used for interpretation of risks and behaviors among youth in the United States.

The task of bringing empirical evidence to bear on drug use as well as on choices of drugs by young people from different contexts will lead to re-examination of patterns of drug use among them as well as to creative ways of conceptualizing these patterns. Despite its limitations, this article reports that, while use of drugs among Haitian youth is relatively low when compared with regional and national data (Florida Youth Behavior Survey, 2000; Monitoring the Future, 2002; National Survey of Drug Use History, 2003), risks for Haitian youth to be involved in drug use are increasing with compounding marginalization and the power of the street to attract and engage youth (Anderson, 1994). Indeed, drug use among young people of Haitian descent has received no attention in the literature. While the present article does not intend to generalize the findings to all Haitian youth in the United States, it does provide important answers to significant questions such as the choice of drugs among Haitian youth, contexts of use and associated factors related to prevalence of drug use. It provides a basis not only for large-scale quantitative study of drug behaviors related to the study population but also for early prevention and intervention programs that may help mitigate the increasing risks for advanced stage in substance use. The data presented here resulted from two studies of Haitian youth in Miami-Dade County, Florida.

Study Background and Sites

Miami-Dade County is a large and highly diverse region with multiple ethnic enclaves. According to the US census, in 2000 Miami-Dade County had over 2,000,000 residents, 57.3% of whom were Hispanic and 19% of whom were Non-Hispanic Black. With a population estimated at 1,018,222 or 46.3% of Miami-Dade residents, North Miami comprises more than 10 incorporated cities including the City of North Miami with various levels of economic and socio-demographic indicators and has the highest concentration of Blacks of various national origins–31.6% compared to 9.8% throughout the rest of the county–(Census, 2000).

Immigrants of Haitian descent and their children comprise one of the largest ethnic groups in Miami-Dade. As for other immigrant communities in Miami-Dade, this population is highly diverse and segmented in terms of class, color, space, language and other forms of valued social capital (Marcelin, 2004). While some Haitians have integrated into the American socioeconomic mainstream, the majority of this population continues to experience economic hardship, partly because of their immigration trajectories and their mode of insertion within the local society (Marcelin, 2004; Portes & Stepick, 1994). Most of this immigrant population arrived in South Florida between 1978 and 1992, from predominantly rural and inner city Haiti, with very few resources and in extremely adverse conditions documented by Laguerre (1988; Lawless, 1992; Portes & Stepick, 1994; and Stepick, 1992).

The driving forces of this relatively recent Haitian migration to South Florida have been the severe sociopolitical and economic crisis and accelerated environmental degradation that have been destabilizing urban and rural Haiti continuously since the mid 1970s. There is no exact count of Haitian immigrants living in Florida. Laguerre (1998) suggests a range of 500,000 to 1,000,000 Haitians in South Florida alone. Official census estimates have not been in a position to compute and report about important segments of the Haitian community that are largely undocumented. Additionally, historical stigmatization and racialization of Haitians in the United States have further limited likelihood of many Haitian immigrants of participating in the censuses, which, in turn, have resulted in massive undercounts (Levitt & Waters, 2002; Marcelin & Marcelin, 2001; Stepick, 1992). Moreover, the generic classification which lumps different ethnic groups into single labels such as “African American” or “Hispanic” has resulted in neglecting specific risks associated with particular conditions in American society.

The studies that produced the data presented here took place in three main geographical areas in Miami-Dade County: North Miami/Little Haiti, Kendall, which is comprised of suburban neighborhoods, and the Homestead/Florida City area, which is located in the south section of the county. Two of the three sites–Little Haiti and Homestead/Florida City–are heavily populated by Haitian immigrants. Little Haiti’s and Homestead/Florida City’s environments are among the most marginalized communities in South Florida (Stepick, 1998). Haitian immigrants’ families’ prospects for survival in these milieus require a massive participation in the informal economy such as clandestine migrant work, domestic services, and low paid services. In their social milieu, normal child rearing is constantly undermined by social forces that interfere with healthy child and adolescent development (Marcelin & Marcelin, 2001). While Homestead/Florida City constitutes a semi-rural environment that shapes social context of risk (80% of Haitian families living in Homestead/Florida City are seasonal migrant farm workers), Little Haiti seems to present different characteristics in terms of urban life and risk for delinquency and crime. In sharp contrast, Kendall is an ethnically mixed middle class community in Miami-Dade County, where more than 60% of the Haitian middle class families reside. The youth that were chosen in Kendall served also as a control group to inform the investigators about the extent to which social contexts and cultural processes shape drug and alcohol use among immigrant youth.

METHODOLOGY

Sampling

We present results from two studies, the Haitian Youth Gang Study and the Haitian Adolescents Lifestyles and HIV Study (both are labeled in this paper under the title Haitian Adolescent Study), funded by the National Institutes of Health/National Institute on Drug Abuse. The studies were designed to explore the sociocultural contexts within which Haitian youth are living in Miami-Dade, including challenges and issues related to identity and gang activity as well as drug use associated with lifestyles, immigration experience and contexts of marginalization or integration of Haitian community into the wider local society. Both studies used snowball community-based samples of Haitian youth (296 participants from the Gang Study and 300 from the HIV Study) selected in neighborhoods predominantly inhabited by Haitian immigrants. The process of establishing a presence in these neighborhoods helped to assure that the field recruiters had developed rapport with the Haitian young people in their natural habitats by the time they began to recruit participants in the structured interview phase of the study. Having established rapport in the neighborhood, the field recruiters could rely on the strength of their introduction to each successive candidate by an already enrolled participant. That introduction assured them the rapport necessary to collect reliable and valid information, despite their relatively brief acquaintance with the candidate. The studies were designed in a way that would make questions related to drug and alcohol use comparable or integrated into one data set for analytical purposes. The merging of the drug and alcohol related data is, in our view, synergistic: the effectiveness and the analytical potential and understandings that result from the combined data are greater than if the two were undertaken separately.

Participants

The 296 participants in the Haitian Youth Gang Study were between the ages of 12 and 26 from two sites in Miami-Dade, while those in the Haitian Adolescents Lifestyles and HIV (300) were between 13 and 18 from three sites in Miami-Dade. The third site included in the second study was Kendall, which yielded a relatively low proportion of participants. Participants’ level of education in both studies ranged between 7 to 12th grades, with a few in first year college. For the purpose of this article, data collected from these questions have been merged into a single set, excluding age range beyond 22 and below 13. After the merging and exclusions, the population totals 557 young people of ages 13 to 22. The results have been grouped under three different age categories: 13–14; 15–17; and 18–22. This categorization reflects developmental differences and the opportunity to compare data within the same age range across ethnic groups in Florida (Florida Youth Substance Abuse Survey, 2000) and in the nation (National Survey on Drug Use History: Monitoring the Future, 2000; and the NIDA Drug Use Among Racial/Ethnic Minorities, 2003).

Instrument

The research design involves three integrated data collection procedures. Participant observations of Haitian youth in their environments (neighborhoods, parks, streets and their homes) where conversations on related drug or alcohol use were conducted. These observations were recorded in field notes. An interview schedule was administered to the 557 study participants (257 participants aged 13–22 for the Haitian Youth Gang Study, between October 1999 and December 2001, and 300 aged 13 through 18 for the Haitian Adolescents Lifestyles and HIV Study between May 2002 and September 2003) that comprised questions focused on sociodemographics, family history, religious activity and schooling, community, and drug and alcohol use. Questions related to drug and alcohol use focused on the first time participant used cigarettes, marijuana, cocaine, heroin, amphetamines, crack, pills/prescription drugs, or other drugs with possibility of multiple choices. Questions also focused on the last time participant used any of these drugs and regularity of use for each of these drugs (Almost everyday; Once a week; Two or three times a month; Other (specify); DK; Refused), the dosage or amount, the route of ingestion (Smoke; Snort; Inject; Eat; Other); friendship and drug use and how they pay for the drugs.

The design also provided for continuous in-depth interviews and intensive observations of family contexts of a subsample of youth from each site. Data collection took place in the youth community environment, including neighborhoods, parks and homes in the three sites. This strategy allowed us to capture contexts of drug and alcohol-related practices or avoidance of drugs and alcohol as well as the meanings attached to these practices.

The interview schedule was designed to capture self-reported behaviors related to drug and alcohol use. In this sociocultural context where there is virtually no discussion of these themes, children may refrain from breaking cultural taboos or other issues related to their conditions as children of immigrants or immigrants themselves. Thus, a critical question consists of the validity of these reports. As is the case for other studies and surveys dealing with use of illicit drugs, we were faced with the reality of the limitations of self-reports (Matt et al., 2003). Two factors, however, have contributed to our high confidence in the quality of the data. The first stems from the fact that the qualitative approach, which consists mostly of systematic and long term observations, in-depth interviews and social mapping over time, has led to strengthening of relationships between investigators and participants, thereby establishing a relatively high level of trust between them. The interviewers who elicited the quantitative data were not the same people who conducted the in-depth study, and we therefore had the advantage of trust for our research project but no observer bias in the collection of the quantitative data. Second, the participants in the in-depth studies were selected from the respondents to the structured interview schedule. The ongoing observation and in-depth investigation of their behavior in natural habitat enabled the investigators to crosscheck responses to items on the interview schedule with observational and in-depth data gathered on the same individuals. Although not all individuals’ data came under this examination, we found sufficient corroboration for the responses to the interviews in the qualitative data to reinforce confidence in the quantitative survey, consistent with the concept of triangulation (Russell, 2000).

Statistical Methods

We used the SPSS statistical program to analyze the data. Because the nature of the study was primarily exploratory, quantitative analyses of the sociodemographic data were predominantly descriptive and non-parametric. Tables of frequencies and percentages provide us with a detailed picture of this not well-understood segment of the population. Bivariate chi-square analyses were utilized to pursue specific hypotheses on the relationship among alcohol, cigarette and other drug (ATOD) use and sociodemographic characteristics (e.g., migration sequences or family disruption, exposure to environment of severe marginalization). Regular use of alcohol and drugs (i.e., marijuana) was further investigated with logistic regression to help identify the relative risk of critical demographic factors (gender, age, place of residence, family structure, attitudes towards religion, and friends’ drug use).

RESULTS

Sociodemographics

Sociodemographic characteristics of the study participants are reported in Table 1. Participants’ ages range from 13 to 22 with 308 (55.3%) male and 249 (44.7%) female. The sample can be characterized as “late adolescent” (median age = 17), predominantly US and Haitian-born, respectively 61.3 and 32.4 percents.

TABLE 1.

Participants’ Sociodemographic Characteristics

Demographic Characteristics (N = 557)	Number	Percent
Gender
Male	308	55.3
Female	249	44.7
Age
13–14 years old	84	15.1
15–17 years old	257	46.1
18–22 years old	216	38.8
Country of Birth1
Haiti	179	32.4
United States	339	61.3
Bahamas	29	5.2
Other	6	1.1
Ethnic Identity2
Haitian	274	49.3
Haitian American	251	45.1
Black	12	2.2
African American	2	0.4
Bahamian	5	0.9
Other	12	2.2
Place of Residence
North Miami/Little Haiti	296	53.1
Homestead	182	32.7
Kendall	79	14.2
Religious Affiliation3
Catholic	199	37.6
Protestant	125	23.6
No Religious affiliation	94	17.8
Other	111	21.0
Family Structure (“Who raised you?”)
Both biological parents	83	14.9
One biological/One stepparent	122	21.9
(Biological) Mother only	101	18.1
(Biological) Father only	35	6.3
With at least one stepparent4	39	7.0
Other combinations	216	38.8

¹4 missing cases.

²1 missing case.

³DK (13) and 15 missing cases leaves (n = 529) for percentage values.

⁴Note that this category overlaps with the second category which includes one step and one biological parent.

Country of birth and ethnic identification

The majority (61.3%) were born in the United States while most of the remainder were born in Haiti (32.4%). Parents of participants were predominantly born in Haiti (93% of both mothers and fathers). Most participants self-identified as either “Haitian” (49.3%) or “Haitian-American” (45.1%). Among those born in Haiti, 83% described ethnic identification as “Haitian” while 16% self-described as “Haitian-American.” In contrast, among those born in the United States, 63% described ethnic identification as “Haitian-American” while 32% selected “Haitian” as their ethnic designation.

Country of birth and residence

Overall, approximately one-half of the respondents lived in North Miami/Little Haiti (53.1%), one-third lived in Homestead (32.7%), and the remainder (14.2%) lived in Kend-all at the time of the survey. Distribution of residence differed by country of birth. For those born in Haiti, almost one-half lived in Homestead (47%) and about one-third (34%) lived in North Miami/Little Haiti. In contrast, for those born in the United States, the majority (62%) lived in North Miami/Little Haiti, while only one-in-four (25%) lived in Homestead.

Religious affiliation

The majority identifies either Catholic (37.6%) or Protestant (23.6%) religious affiliation and only 18% indicate no religious affiliation. A clear majority considers religion either “very important” (48%) or “a little important” (29%).

Family structure

Regarding family structure, for approximately one-half of the participants (53%), parents were either currently married (43.5%) or living together in common law (9.5%). For more than one-third of the participants (34.3%), parents were either separated (21.3%) or divorced (13.0%); 6.9% of participants lived with a parent who was identified as “single,” while 5.8% lived with a parent who had been widowed. About 42 percent of the respondents answered the question, “Who raised you?” with the response, “both parents.” Many others indicated that they were raised by their mother only (18.1%). Father (6.3%) and stepfather were also reported as important primary caretakers. Household sizes vary from 5.6 in Little Haiti and Homestead to 3.4 in Kendall with a general mean of 3.8 for the overall sample.

Drug Use

As indicated in Table 2, lifetime (“ever use”) use of ATOD was reported in the following proportions: cigarettes (27.4%), alcohol (50.5%), and illicit drugs (25%). Overall, 57.9% of participants reported ever using either alcohol or drugs during their lifetimes. Lifetime use of marijuana was reported by 134 participants (24%), and cocaine use by 14 participants (2.5%). Two participants reported lifetime use of crack cocaine. Use of heroin and amphetamines were each reported by a single participant. It is clear, therefore, that of the various illicit drugs mentioned, marijuana is by far the most likely to have been tried or used with anything approaching regularity. For this reason, subsequent analyses of “regular” drug use are restricted to the use of marijuana specifically. Also, because daily use of alcohol or drugs is infrequent in this sample, “regular” use is conservatively defined as “at least once per month” ensuring a distribution of responses that is more appropriate for the types of analyses presented (see Table 3).

TABLE 2.

Lifetime Use (“Ever Use”) of Alcohol, Cigarettes, and Other Drugs (ATOD)

Demographic Characteristics	Cigarettes	Alcohol	Drugs	Alcohol or Drugs
	Percent	Percent	Percent	Percent
All Participants	27.4	50.5	25.0	57.9
Gender
Male	28.2	50.3	30.0	59.8
Female	26.3	50.6	18.9	55.6
P value	n.s.	n.s.	.003	n.s.
Age
13–14 years old	12.2	36.9	8.6	40.7
15–17 years old	22.3	49.2	16.9	55.6
18–22 years old	39.3	57.3	40.7	67.1
P value	.0001	.006	.0001	.0001
Place of Residence
North Miami/Little Haiti	28.4	48.5	33.3	60.5
Homestead	22.1	40.3	13.3	43.6
Kendall	35.4	81.0	21.5	81.0
P value	n.s.	.006	.0001	.0001
Place of Birth
Haiti	29.0	48.3	20.7	55.2
United States	26.5	49.7	24.5	63.9
P value	n.s.	n.s	n.s.	.06
Family Structure
Two parents in home	33.1	47.1	26.4	56.2
Two biological parents	29.3	45.1	22.0	51.2
1 or 2 step parents	41.0	51.3	35.9	66.7
One biological parent in home	32.1	49.3	40.6	62.4
Other combinations	22.9	52.3	17.2	56.6
P value	< .04	n.s.	.0001	n.s.
Importance of Religion
Not important	34.8	58.9	43.2	71.2
A little important	36.5	55.8	32.1	63.2
Very important	18.6	44.3	12.1	49.2
P value	.002	.03	.0001	.001
Friend’s History of Drug Use
Yes	41.5	62.0	44.2	75.3
No	21.1	45.3	15.1	54.6
P value	< .0001	< .0001	< .0001	< .0001

TABLE 3.

Regular/Recent ⁵ Use of Alcohol and Marijuana

Demographic Characteristics	RegularAlcohol Use	Regular Marijuana Use	RecentAlcohol Use	Recent Marijuana Use
	Percent	Percent	Percent	Percent
All Participants	45.4	14.2	16.2	10.4
Gender
Male	46.4	20.5	19.2	15.6
Female	44.2	6.4	12.4	4.0
P value	n.s.	.0001	.03	.0001
Age
13–14 years old	31.0	2.4	8.3	6.0
15–17 years old	43.6	6.2	9.7	3.9
18–22 years old	53.2	28.2	26.9	19.9
P value	.002	.0001	.0001	.0001
Place of Residence
North Miami/Little Haiti	45.9	21.6	14.2	16.2
Homestead	33.0	6.0	11.5	4.4
Kendall	72.2	5.1	34.2	2.5
P value	.0001	.0001	.0001	.0001
Place of Birth
Haiti	41.9	10.6	20.1	7.8
United States	45.4	15.3	13.6	10.9
P value	n.s.	n.s.	< .05	n.s.
Family Structure
Two parents in home	45.1	17.2	12.3	11.5
Two biological parents	43.4	16.9	13.3	8.4
1 or 2 step parents	48.7	17.9	10.3	17.9
One biological parent in home only	47.1	27.9	16.9	21.3
Other combinations	44.8	6.7	17.4	5.0
P value	n.s.	.0001	n.s.	< .0001
Importance of Religion
Not important	51.3	26.5	21.2	23.0
A little important	52.9	17.2	18.5	12.1
Very important	39.5	6.8	12.0	3.8
P value	.02	.0001	< .05	< .0001
Friend’s History of Drug Use
Yes	57.6	30.2	26.7	22.1
No	40.0	7.0	11.4	5.2
P value	< .0001	< .0001	< .0001	< .0001

⁵“Regular” use indicates that participant reports using at least once per month. “Recent” indicates that they have used within the past 30 days.

Gender

Similar proportions of males and females reported lifetime use of cigarettes or alcohol. A significant gender disparity was noted for lifetime use of drugs, however; 30% of males reported having used drugs in contrast to 18.9% of females.

Age

Significant increases in lifetime prevalence of ATOD use were noted for older youth for cigarettes, alcohol, and drugs, with increasing age. For the combination, alcohol-and/or-drug use, lifetime prevalence rose from 40.7% among 13–14 year-olds, to 55.6% among 15–17 year-old youth, and 67.1% among those in the age range 18–22.

Place of residence

Significant contrasts in lifetime prevalence of ATOD were noted by neighborhood of residence. The lowest prevalence for all substances was found in participants from Homestead. The highest prevalence for alcohol was observed among the sample from Kendall while the highest drug use prevalence was observed in the North Miami/Little Haiti sample.

Family structure

Lifetime prevalence of alcohol use did not differ between participants from homes with two parents compared with those from single-parent homes. Drug use prevalence, however, was lower for those youth having two (especially biological) parents in the home rather than one. Within two-parent households, lifetime ATOD prevalence was numerically lower for those with two biological parents compared with those living with at least one step parent for all comparisons–cigarettes, alcohol, drugs, alcohol-and/or-drugs.

Place of birth

Though not significant in any case, trends in the data suggest that those born in the United States have higher rates of alcohol and drug use. Overall ATOD use was marginally higher for US-born (63.9%) than for Haitian-born (55.2%) participants.

Importance of religion

Lowest lifetime prevalence rates were consistently found for participants who indicated that religion is “very important.” Generally, intermediate prevalence rates were found for those who identified religion as “a little important.” With the exception of cigarette use, highest lifetime prevalence rates were found for those who labeled religion as “not important.” Comparisons across the three response categories were highly significant for cigarette, alcohol, drugs, and alcohol-and/or-drugs.

Friends’ use of drugs

Participants with friends who use drugs had consistently higher rates of cigarette, alcohol, drug and overall ATOD use. The fact that substantial differences were found across all substances suggests that friends’ use of drugs, together with age and importance of religion, may be one of the primary influences on Haitian adolescent ATOD use.

Regular (past 30 days) use of marijuana and alcohol are reported in Table 3. In this sample, 45.4% reported at least monthly use of alcohol and 14.2% reported monthly or more frequent use of marijuana.

Gender and age

While similar proportions of male and female youth reported regular drinking, the prevalence of regular marijuana use was more than three times higher for males compared with females. Prevalence of regular and recent use rose significantly with age for both alcohol and marijuana, with a particularly sharp upward surge noted for regular marijuana use between age categories 15–17 years (6.2%) and 18–22 years (28.2%).

Place of residence

Highly significant contrasts in alcohol and marijuana use were seen by place of residence. Highest rates of regular drinking were found among participants from Kendall. In contrast, highest rates of marijuana use were observed in the North Miami/Little Haiti sample.

Place of birth

Regular alcohol and marijuana use and recent use of marijuana were not significantly different for those born in Haiti or the United States. Nevertheless, differences were obtained on recent alcohol use, indicating that a higher percentage of Haitian-born (20.1%) than US-born (13.6%) participants drank within the past 30 days.

Family structure

Rates of drinking were very similar for participants in two-parent households, one biological parent households, and “other combinations” households. Regular and recent use of marijuana, however, was found to be significantly higher in households with only one biological parent than in other households, especially those with both biological parents and in households with “other combinations.”

Importance of religion

Prevalence of regular and recent drinking was significantly lower for those who rated religion as “very important” while prevalence rates were higher and very similar for those indicating that religion was either “a little important” or “not important.” For regular and recent marijuana use, a similar significant trend was noted with the highest rate found among those indicating that religion was “not important” (26.5%; 23.0%)–approximately four times greater than the rate for those who rated religion as “very important.” (6.8%; 3.8%). Those indicating that religion was “a little important” had an intermediate prevalence rate (17.2%; 12.1%).

Friends’ use of drugs

As with lifetime ATOD use, those with friends who use drugs were consistently more likely to report regular and recent use of both alcohol and marijuana than participants without drug-using friends.

The preceding analyses (Tables 2 & 3) give us preliminary indication of the factors associated with drug (i.e., marijuana) and alcohol use among the Haitian adolescent sample. While many of the sociodemographic variables were significantly associated with at least some aspect of ATOD use when considered alone (i.e., with bivariate analyses), it remained to be seen whether any given association retained significance when the influence of all sociodemographics are taken into account (i.e., controlled for). To this end, a series of logistic regression analyses were run to determine the relative risk of each variable for alcohol or marijuana use. The dependent variables (categorical) of these analyses (lifetime, regular, and recent use of alcohol and marijuana) were thus regressed on the seven demographic factors of interest (Gender, Age, Place of residence, Place of birth, Family structure, Importance of religion, and Friends’ drug use). Results are summarized in Table 4.

TABLE 4.

Relative Risks for Alcohol and Marijuana Use Among Haitian Adolescents

	Ever used alcohol6	Ever used marijuana7	Regular use of alcohol8	Regular use of marijuana9	Recent use of alcohol10	Recent use of marijuana11
Overall Rates	50.5	24.1	45.4	14.2	16.2	10.4
Variablea	OR/CIb	OR/CI	OR/CI	OR/CI	OR/CI	OR-CI
Male	.91 (.61−1.34)	1.38 (.82−2.31)	.94 (.63−1.39)	2.21* (1.09−4.47)	1.41 (.79−2.52)	2.53* (1.11−5.78)
Older Age	1.17 ** (1.05−1.29)	1.46** (1.27−1.67)	1.18** (1.07−1.31)	1.54** (1.30−1.82)	1.37** (1.19−1.59)	1.32** (1.10−1.57)
Resident ofc
Kendall	6.04** (3.08−11.83)		4.69** (2.51−8.76)		5.32** (2.47−11.47)
Homestead		0.33** (.14−.75)12
Born in . . .
Haiti	.78 (.51−1.20)	.60 (.34−1.06)	.78 (.50−1.20)	.50 (.24−1.03)	1.09 (.62−1.94)	.64 (.29−1.41)
Not raised by both parents	1.10 (.85−1.41)	.97 (.70−1.34)	.96 (.75−1.23)	.91 (.61−1.35)	1.18 (.81−1.73)	.91 (.59−1.41)
Religion deemed unimportant	1.93** (1.16−3.23)	3.19** (1.70−6.00)	1.70* (1.02−2.83)	1.82 (.83−3.99)	1.82 (.90−3.69)	2.79* (1.15−6.75)
Friend’s use drugs	1.58* (1.03−2.45)	2.77** (1.68−4.55)	1.58* (1.03−2.42)	3.45** (1.86−6.41)	2.10** (1.20−3.67)	3.07** (1.54−6.11)

^a)Logistic regressions were run on ever used (lifetime), regular use (at least monthly) and recent use (within past 30 days) of alcohol and marijuana. Each of the variables listed were included as predictors in all analyses.

^b)OR = Odds Ratios and CI = 95% Confidence Intervals.

^c)Little Haiti/N. Miami not significant for any outcome.

⁶The overall model correctly classified 64.1% of cases accurately; Chi-square(9) = 66.92, p < .0001.

⁷The overall model correctly classified 79.5% of cases accurately; Chi-square(9) = 116.17, p < .0001.

⁸The overall model correctly classified 64.4% of cases accurately; Chi-square(9) = 65.08, p < .0001.

⁹The overall model correctly classified 88.7% of cases accurately; Chi-square(9) = 112.19, p < .0001.

¹⁰The overall model correctly classified 85.1% of cases accurately; Chi-square(9) = 66.86, p < .0001.

¹¹The overall model correctly classified 90.9% of cases accurately; Chi-square(9) = 77.25, p < .0001.

¹²The Odds Ratio < 1.00 indicates that living in Homestead is a mitigating, as opposed to a risk factor, for ever having used marijuana.

^*p < .05

^**p < .01

Alcohol Use

With respect to alcohol use, a clear and consistent pattern emerges in the data. For lifetime and regular use of alcohol, results indicate that users are more likely to be older (adolescent) residents of the neighborhood of Kendall who consider religion unimportant and who have friends who use drugs. For recent use of alcohol, defined as use within the past 30 days, the picture is similar though importance of religion no longer retains significance. Of these risk factors for alcohol use, it is clear that neighborhood of residence (Kendall) has the most powerful effect on alcohol use [Odds ratios range from 4.69 (regular use) to 6.04 (lifetime use)], indicating that residents of Kendall are anywhere from 369% to 504% more likely to use alcohol than residents of the other Miami neighborhoods.

Marijuana Use

The picture that emerges of marijuana users in the sample is somewhat, though not substantially, different from that of the alcohol users. On all three measures of marijuana use (lifetime, regular, recent), age and friends’ drug use consistently predict use of marijuana. As with alcohol, this indicates that older adolescents in the sample with drug-using friends are significantly more likely to use marijuana than others. While gender did not appear to influence alcohol use in these analyses, males appear to be more at risk for regular and recent use of marijuana. Odds ratios of 2.21 (regular use) and 2.53 (recent use) indicate that males are from 121% to 153% more likely to use marijuana than females in the sample. The only neighborhood associated with marijuana use was Homestead. Nevertheless, whereas Kendall was associated with increased risk for alcohol use, Homestead residents were actually 67% less likely to have a history of ever trying marijuana (OR = .33). Finally, the most powerful predictor of marijuana use in the Haitian adolescent sample is friends’ use of drugs. Odds ratios for all three measures of marijuana use (2.77, 3.45, 3.07) are consistently higher than odds ratios for the other factors and suggest that participants with friends who use drugs are anywhere from 177% to 245% more likely to use marijuana, depending on how it is being measured.

It is also worth pointing out that while all six logistic regression models are statistically significant overall, those predicting marijuana use are consistently more accurate at classifying cases as users or non-users than those predicting alcohol use (see footnotes, Table 4). The overall models accurately classified 79.5%, 88.7% and 90.9% of lifetime, regular, and recent marijuana users/non-users, while models predicting lifetime, regular, and recent alcohol use were generally less accurate overall (64.1%, 64.4%, 85.1%). Nevertheless, it is clear that models predicting use of both drugs are reliable, especially with respect to recent alcohol and marijuana use (85.1%, 90.9% classified correctly) which is typically taken as a marker for “active” use of the drug in question.

DISCUSSION

Comparison with National Surveys

The Haitian Adolescent Study (HAS) sample, in comparison to both national and regional samples, is not particularly high-risk with respect to ATOD use. The following sections present various levels of comparison between the HAS data and national surveys.

Three reports (FYSAS, NSDUH and MTF) present data compiled from adolescents of varying ages/grade levels, e.g., the NSDUH includes data from adolescents aged between 12 and 17 years, whereas the MTF data presents data restricted to adolescents surveyed in grades 8 and 10. The NIDA survey presents data for each grade (8th through 12th) level separately. Not surprisingly, 12th grade students had higher lifetime and recent use rates for all four of the substances reported in Table 5, than did eighth grade students. This age graded response is consistent with the findings of HAS and thus caution must be exercised in making comparisons with surveys that do not report data on adolescents of comparable ages.

TABLE 5.

A Comparison of Haitian Adolescent ATOD Use with National and Regional Data

		Data Source
	HAS13 (ages 13−18)	FYSAS14 (overall high school, through age 18)	NSDUH15 Ages 12−17	MTF16 8th and 10th grade	NIDA17: (8th−12th grades)			HAS
					White	Hispanic	Black	Haitian
Cigarettes
Lifetime	23.9	52.3	33.3	39.4	(42.6−67.2)	(45.7−66.1)	(39.8−45.5)	(17.4−37.6)
Past month	5.0	21.7	13.0	14.2	(17.7−37.9)	(16.6−27.7)	(9.6−14.4)	(0.0−9.1)18
Alcohol
Lifetime	48.6	68.9	43.4	57.0	(52.1−82.0)	(57.7−84.3)	(47.7−70.3)	(40.0−57.0)
Past month	13.1	43.4	17.6	27.5	(24.7−55.1)	(26.7−51.2)	(16.0−30.0)	(16.0−25.2)
Marijuana
Lifetime	18.0	36.6	20.6	29.0	(19.0−49.4)	(27.1−55.0)	(23.8−45.7)	(10.0−33.0)
Past month	5.9	18.3	8.2	13.1	(8.4−22.7)	(12.7−24.6)	(9.3−19.0)	(0.0−10.7)
Cocaine
Lifetime	< 1.0	6.5	2.7	4.9	(4.2−9.9)	(8.9−13.3)	(1.4−1.9)	(0.0−1.0)
Past month	< 1.0	2.0	< 1.0	1.4	(1.1−2.5)	(2.7−3.6)	(< 1.0)	(0.0−1.0)

¹³In the HAS study, recent ATOD used was assessed in two ways. One question asked participants “how long has it been since you used” the substance in question while another asked, “how often do you use” the substance in question. Percentages reported for past month use among HAS participants reflect the number indicating recollection of use within a month’s time. HAS participants were not directly asked, “Have you used within the past month?”

¹⁴Florida Youth Substance Abuse Survey, 2000.

¹⁵National Survey of Drug Use History, 2002.

¹⁶Monitoring the Future, 2002.

¹⁷Drug Use Among Racial/Ethnic Minorities (Revised), 2003.

¹⁸Data concerning “recent use of cigarette” were only available on a subset (n = 300) of the overall sample.

When compared to the national and Florida surveys, Haitian adolescents were much less likely to have engaged in lifetime or recent use of cigarettes or cocaine. For cigarette use, the Haitian adolescents were only half as likely to have ever smoked in their lifetimes (23.9%) as high school adolescents surveyed in Florida (FYSAS) (52.3%) and two thirds as likely as students in the NSDUH (33.3%) and MTF (39.4%) surveys. Haitian adolescents were also less likely to have used marijuana (lifetime or recent use) than other adolescents. Haitian adolescents were only half as likely (18%) as the adolescents in the FYSA study (36.6%) and just over two-thirds as likely as adolescents in the NSDUH (20.6%) and MTF (29%) surveys to report marijuana use. Recent use of marijuana by Haitian adolescents (5.9%) was only one-third of that of the FYSAS adolescents (18.3%) but closer to rates reported in the NSDUH (8.2%) and MTF (13.1%) surveys. The percentage of Haitian adolescents reporting lifetime use of alcohol (48.6%) was more similar to that of the other adolescents reported in the Florida (FYSAS) (68.9%) and MTF (57%) surveys. Lifetime alcohol use among Haitian adolescents was actually higher than that reported among adolescents in the NSDUH survey (43.4%). Nevertheless, recent use of alcohol was consistently lower for Haitian adolescents (13.1%) than that reported in the FYSAS (43.4%), NSDUH (17.6%) and MTF (27.5%) studies. The Haitian adolescents had very low use of cocaine (just under1%) when compared to all other adolescent populations.

Of all the variables investigated in the HAS study, apart from age comparisons which were possible for the NIDA survey, gender was the only other variable for which a comparison was possible in Florida (FYSAS study). In the HAS study, males were more likely than females to have used marijuana in either their lifetime or recently. This pattern of higher male use of marijuana in adolescents was also evident from data presented from Florida (FYSAS study).

A more detailed examination of the progression of the use of cigarettes, alcohol, and marijuana in Florida adolescents and the Haitian adolescents revealed some interesting and consistent differences. At each age (13 through 18 years), Haitian adolescents had lower lifetime and recent use rates than other Florida adolescents. The adolescent Haitians appear to have begun using cigarettes and marijuana at a later age because, for example, no 13-year-old Haitian adolescent in this study had ever used marijuana or smoked cigarettes in the past month, while 12% of 13-year-old Florida adolescents had tried marijuana and smoked cigarettes recently. For Florida adolescents, the greatest increase in the number having tried or currently using cigarettes and marijuana appears to have been between the ages of 13 and 15 years while the greatest increase in the numbers of Haitian adolescents having ever tried them was between the ages of 17 and 18 years. Differences in the use of alcohol between the Haitian and other Florida adolescents were not as marked.

Florida adolescents showed the largest difference in the proportion using alcohol between the ages of 13 and 14 years, whereas the difference between the proportion of younger Haitian adolescents using alcohol and the older Haitian youth age were more modest than differences among Florida age groups. There was, however, a larger proportion of lifetime use for Haitian youth at around 15 years, and regular use between the ages of 17 and 18 years. This age-related progression in the use of alcohol, marijuana, and cigarettes is summarized in Figure 1.

FIGURE 1.

ATOD Use Among Haitian and Florida Adolescents Ages 13 to 18

Ethnicity and Drug Use

The NIDA report and the regional Florida FYSAS also examined ATOD use in the different ethnicities. Comparisons were made between White, Black and Hispanic adolescents. The data from the NIDA are presented in Table 5 and the data from the Florida survey are presented in Figures 1 and 2. In the national (NIDA) survey, White and Hispanic students reported the highest lifetime and recent use of cigarettes, alcohol and cocaine but Black students had the highest use for lifetime but not recent use of marijuana. When the data from the HAS were compared with these two surveys it was clear that Haitian adolescents had relatively low lifetime and recent (where comparison is possible) use of cigarettes. While alcohol use was lower than reported for White and Hispanic adolescents it approached the level of use for the Black adolescents surveyed for the NIDA study and exceeded the lifetime use for Black adolescents in the Florida survey and was only a little lower for recent use. Similarly, marijuana use was lower for Haitian adolescents than for White, Hispanic, and Black adolescents in the national NIDA survey. Nevertheless, while marijuana use (both lifetime and recent) use was lower among Haitian adolescents when compared to the Black adolescents surveyed nationally, the Haitian adolescents had a higher lifetime but not recent use incidence in Florida. Haitian adolescents also showed much lower use of cocaine (both lifetime and recent) than White and Hispanic adolescents and slightly lower lifetime use and comparable recent use to the Black adolescents surveyed (NIDA study only; not presented for FYSAS study in Figure 2).

FIGURE 2.

ATOD Use Among Haitian Adolescents and Other Ethnic Groups in Florida

Haitian Youth and ATOD Use: A Perspective on Contexts and Immigration

Haitian respondents to the interview schedules used in two companion studies of youth at risk indicated that their patterns of drug use differed somewhat from those of school age youth all over the United States (cf. Johnston, O’Malley, & Bachman, 2003). Because one data source involved a large probability sample of high school students throughout the United States, while the HAS involved a mix of 13–18-year-old Haitian respondents recruited from the communities in Miami/Dade County, we cannot infer that Haitian youth have less drug use than do young people of all race/ethnic categories from all over the United States. The data gathered in Miami/Dade County do, however, suggest that the process of adjusting to life as first or second generation immigrants to the United States has not yet produced the same patterns of drug use found in “native” populations of young people. The “protective” impact of recent immigration on patterns of drug use (see Amaro, Whitaker, Coffman, & Heeren, 1990; Boles, Casa, Furlong, Gonzalez, & Morrison, 1994; Burnam, Huogh, Karno, Escobar, & Telles, 1987; Vega, Alderete, Kolody, & Aguilar-Gaxiola, 1998) seems to obtain in general in the Haitian respondents to our interviews. Some of them, usually the older ones, have incorporated marijuana and cocaine use into their drug use repertoires, but they have not yet become regular users of these drugs. When we pursue the “protective” impact of recent immigration, however, the story becomes somewhat unclear. Those Haitian youth born in the United States were no more likely to have used marijuana or other illegal drugs than were the ones born in Haiti or elsewhere. Further consideration of the protective hypothesis must take into account the following factors: (1) developmental stage at the time of arrival will affect acculturation of the individual arrivee; (2) acculturative experience may lead to a strong need to identify oneself as Haitian or Haitian American; (3) recently discovered need to identify oneself as culturally distinct may cause a person to identify himself as Haitian; (4) faced with possible recrimination or stigma, an individual may choose not to identify himself as Haitian.

Drugs that generally have bad reputations, such as crack cocaine and heroin, had low proportions of users among the young Haitian respondents. Interviewees expressed very negative opinions about use of certain drugs:

INTERVIEWER: And what about crack, I mean . . .?

INTERVIEWEE: Crack is the same thing . . . I’m not messing with all that stuff. It fuck you up, mess up your mind . . .

INTERVIEWER: But do you think that most of young Haitians now, because they are getting more American they don’t . . .

INTERVIEWEE: No, they don’t . . . they don’t do . . . I don’t know none of my Haitian friends that do crack, they smoke weed but....all that heroin and stuff like that . . .

INTERVIEWER: But what about the injection . . .

INTERVIEWEE: Yeah, that’s heroin . . . I don’t . . . I don’t know ‘bout all that. I’m not that much into they life to know what they do . . .

INTERVIEWER: But the young kids, they don’t . . . they are not fascinated by that, they don’t want to go . . .

INTERVIEWEE: Yeah . . . they don’t want to do none of that stuff.

INTERVIEWER: Why is it they don’t want to?

INTERVIEWEE: Cause we hear all the stuff heroin make crack heads . . . spend all your money . . . you get . . . addicted to it on your first try you know . . . you lose everything, because it . . . nobody wanna try that. I wouldn’t try it. I don’t know about anybody else, but me, I wouldn’t try it.

Apparently, this respondent had taken to heart the warnings he had heard on the street about drugs such as crack and heroin. At the same time, he acknowledges that his friends smoke marijuana. The attitudes toward that drug, as reflected in far higher percentages of users among the respondents, appear to reflect the benign view of “weed” articulated by rappers and other African Americans:

INTERVIEWER: But, how was it for you the first time you smoke weed?

INTERVIEWEE: Just . . . it was just a high . . . it wasn’t nothing, it’s not all it’s cut up to be. Everybody talking about it feels good and all that stuff, it’s not all that . . . I don’t think it’s all that. It’s just . . . I don’t know why people do the coughing, you know . . . it’s not something that I would . . . always wanted to do.

INTERVIEWER: Yeah, but the first experience, you how, how did it focus?

INTERVIEWEE: First experience . . . it was . . . the first time, it’s ain’t really good, cause you cough and you not used to it. For you to get high, you have to try it at least like three times . . . to get that high.

INTERVIEWER: Yeah. But do they . . . it, does it make you more, I mean . . . alert or . . .

INTERVIEWEE: Weed slows you down, so you know, it . . . it slows you . . . slows you down, you have a calm to you or something . . . it’s not really alert. I don’t know if . . . for me, I can still control myself, but I don’t know about other people . . . other people have different side effects to it.

This self report is very typical of early experimentation with marijuana, especially the uncertainty about whether or not the effects were worth the trouble. It also features an objective, self-aware assessment of how the individual performs in street settings under the influence of Cannabis.

The data presented here may give some insight into the process of adopting certain kinds of drug use first. The Haitian youth who responded to the interviews clearly preferred marijuana over most other drugs except alcohol. Furthermore, their self-reported alcohol use tended not to involve heavy drinking, unlike the MTF twelfth-graders, of whom nearly thirty percent reported recent heavy drinking. The intercultural context in which young Haitians in Miami find themselves does not emphasize drinking as a social event. Ideal behaviors and expectations in that context involve the acquisition and display of material goods and an alert demeanor to defend one’s personal image. Drunkenness is the object of derision there, so the drugs that one uses must not bring stupors or somnolence. Marijuana, and to a far lesser extent, powder cocaine are the drugs of choice in this kind of social context, because the Haitian youth take only light doses that do not produce effects that could interfere with their alertness. This principle contrasts sharply with youths’ drinking patterns in much of the rest of the United States, possibly explaining the discrepancy between MTF figures and those found in the present data. In many parts of the United States, young people primarily have access to alcohol, and they gather in grocery store parking lots and other gathering places to drink. This pattern has not been found even among Haitian youth who drink fairly heavily.

Crack has a particularly low reputation among youth in Miami/Dade County, where young people see it as the drug of homeless people and middle-aged burnouts. The small numbers (2) of self-reports elicited from 557 Haitian youth indicate that these respondents likewise had internalized that perception of crack cocaine. Heroin hardly appears at all among youth generally, and it is especially rare among Haitian youth in the present study.

CONCLUSION

The data presented here have clear limitations in delineating the contrast between drug use patterns of Haitian youth and the rest of the adolescents in the United States. Nevertheless, they provide the opportunity to interpret the patterns of drug use found among Haitian youth. These patterns suggest that the process of adjustment to life in the United States is at a crucial juncture for young Haitian-Americans. They have not plunged fully into the youthful ethos of risk-taking that seems to draw second- and third-generation children of immigrants, but they seem to be on the verge of adopting these aspects of North American life. Their risk of involvement in drug use seems to be increasing, making the present a very opportune time for preventive intervention. All of our observations and data suggest that large-scale involvement in the consumption of illegal drugs has not yet occurred among Haitian youth. Rather, some individuals have begun to use marijuana and taste other illegal drugs. The rest of the Haitian youth seem naïve about drugs in general, showing familiarity with only alcohol and cigarettes. This naiveté would require prevention efforts to take a very different approach than that used in the “general population,” or that taken by prevention programs that focus on minority populations.