Skip To Content

Go to the Table Of Contents

Click for DHHS Home Page
Click for the SAMHSA Home Page
Click for the OAS Drug Abuse Statistics Home Page
Click for What's New
Click for Recent Reports and HighlightsClick for Information by Topic Click for OAS Data Systems and more Pubs Click for Data on Specific Drugs of Use Click for Short Reports and Facts Click for Frequently Asked Questions Click for Publications Click to send OAS Comments, Questions and Requests Click for OAS Home Page Click for Substance Abuse and Mental Health Services Administration Home Page Click to Search Our Site

2003 State Estimates of Substance Use

7. Discussion

This report presents estimates for the 50 States and the District of Columbia based on the combined National Survey on Drug Use and Health (NSDUH) data for 2002 and 2003 utilizing the small area estimation (SAE) methodology. Chapters 2 through 6 of this report describe the variations across the States in each of the 21 measures for which estimates were produced based on combined data from the 2002–2003 NSDUHs. This chapter provides a more comprehensive analysis of the results, considering similarities and differences in patterns across different measures, in the context of prior research on the relationships between these measures at the individual level. Also included in this chapter is a discussion of how these results compare with prior State estimates from NSDUH, as well as information on the quality of these estimates.

7.1. 2002–2003 State Rankings for Substance Use and Other Measures

State estimates of the prevalence of substance use can provide, among other things, information on the regional clustering of these rates. Many factors can influence State prevalence rates, including local culture and social norms, State and local policies, and the sources, supply, and marketing of drugs. The findings in this report reveal varying degrees of clustering of substance use among States depending on the substance.

States with the highest prevalence of illicit drug use for persons aged 12 or older were mostly in the West (six States) or in the Northeast (three States). The lowest fifth was comprised primarily of States in the South (five States) or the Midwest (three States) (Figure 2.1). There was similar clustering associated with alcohol use among the States, with the highest rates for persons aged 12 or older reflected mostly in the Northeast (six States). However, all of the States in the lowest fifth for alcohol use among persons aged 12 or older, except for Utah, were in the South (Figure 3.1). Cigarette use among persons aged 12 or older was clustered mostly in the South and a few Midwestern States (South Dakota, Missouri, and Ohio) (Figure 4.5). The highest rates of both binge alcohol use and general alcohol use among persons aged 12 or older were found mostly in Northeastern or Midwestern States (Figures 3.1 and 3.5). The highest rates of past month cigarette and tobacco use among persons aged 12 or older were in the South (Figures 4.1 and 4.5).

Substance use epidemiology has documented the inverse relationship between the perceptions of risk in using a substance and the actual use of the substance at the individual level (e.g., Bachman et al., 1998). The lower the perception that use involves risk, the higher the probability of use. This relationship at the individual level is reflected to varying degrees in correlations at the State level. The relationship between risk of binge use of alcohol and actual use is evident at the State level. Eight out of ten States with the lowest percentages of perceived risk of binge drinking among persons aged 12 or older had the highest levels of binge alcohol use (Figures 3.5 and 3.9). A similar (but weaker) relationship occurred between past month binge use of alcohol and past month use of alcohol in general among persons aged 12 or older, with six of the States that ranked highest in past month binge alcohol use also ranking highest in past month alcohol use (Figures 3.1 and 3.5).

Eight of the States with the lowest perceived risk of occasional marijuana use also had the highest rates of past month use of marijuana among persons aged 12 or older (Figures 2.9 and 2.13). The strength of the relationship between the perception of risk and the prevalence of use of cigarettes among persons aged 12 or older was slightly weaker than that between perceived risk of binge use of alcohol and actual binge use. Six States that had high rates of cigarette use also had the lowest rates of perceived risk of heavy use of cigarettes; five States that had low rates of cigarette use also had the highest rates of perceived risk of heavy cigarette use (Figures 4.5 and 4.9). Because marijuana is the most commonly used illicit drug, 8 out of the 10 States with the highest rates of illicit drug use also were the States with the highest rates of past month marijuana use in the 12 or older population (Figures 2.1 and 2.9). States where the rate of first-time use of marijuana was high also tended to be States with the highest rates of past month marijuana use (Figures 2.9 and 2.17). States with the highest rates of past year marijuana use were mostly in the West (five States) and the Northeast (four States) (Figure 2.5).

Of the 10 States in the top fifth with respect to past month use of an illicit drug for persons aged 12 or older, 5 were in the top fifth for past month use of an illicit drug other than marijuana (Figures 2.1 and 2.20). Only five of the States with the highest levels of past month use of illicit drugs other than marijuana for persons aged 12 or older also had the highest rates of past year use of cocaine (Figures 2.20 and 2.24). In general, a State that had a high level of use of one substance also tended to have high levels of use of related substances.

States that ranked high for substance use by all persons aged 12 or older also ranked high in the use of substances by the population aged 26 or older. This relationship derives from the fact that the latter group represents 77 percent of the total population aged 12 or older. Although the 26 or older age group often drove the prevalence rates in the 12 or older population in a State, rates among the 12 to 17 and 18 to 25 age groups may not have followed the same pattern. For example, Louisiana and the District of Columbia had rates in the top fifth for past year use of cocaine among persons aged 26 or older, but they were ranked in the lowest fifth both in the 12 to 17 age group and in the 18 to 25 age group. On the other hand, Colorado and Arizona ranked in the highest fifth for past year cocaine use among all three age groups (12 to 17, 18 to 25, and 26 or older) (Figures 2.24 to 2.27).

The relationship of past month use of alcohol to past year alcohol dependence or abuse was not particularly strong due in part to the widely different prevalence levels of the measures. For example, among the States with the highest rates of past month alcohol use for those aged 12 or older (States ranged from 57.7 to 59.8 percent), only three States were in the highest fifth for past year dependence on or abuse of alcohol (rates ranged from 9.2 to 10.8 percent) (Tables B.8 and B.14, Figures 3.1 and 5.1). The relationship between past month binge use of alcohol (about 23 percent nationally) and past year alcohol dependence or abuse was substantially stronger, showing 6 States in the top 10 for binge alcohol use also present in the top fifth for alcohol dependence or abuse in the past year among persons aged 12 or older (Table  B.9, Figures 3.5 and 5.1).

The majority of States with high prevalence rates for alcohol dependence or abuse were not the same States that had high prevalence rates for illicit drug dependence or abuse. Only three of the States in the top fifth with the highest rates of alcohol dependence or abuse (Arizona, New Mexico, and Rhode Island) among persons aged 12 or older also were in the group of States with the highest levels of illicit drug dependence or abuse (Figures 5.1 and 5.9). Most of the States with the highest levels of illicit drug dependence or abuse were in the Northeast (Massachusetts, New Hampshire, Rhode Island, and Vermont) or the West (Arizona, Colorado, New Mexico, and Washington). The top fifth also included the District of Columbia and Louisiana from the South. There were three States (Massachusetts, New Mexico, and Rhode Island) common in the top fifth for all three age groups (12 to 17, 18 to 25, and 26 or older) for past year illicit drug dependence or abuse (Figures 5.9 to 5.12).

There was a strong relationship between high rates of past year illicit drug dependence or abuse and high rates of past year cocaine use at the State level. Nine out of ten States were ranked among the highest for persons aged 12 or older for both measures: Arizona, Colorado, District of Columbia, Louisiana, Massachusetts, New Hampshire, New Mexico, Rhode Island, and Vermont (Figures 2.24 and 5.9).

Not only did geographic clustering of States occur among those with high prevalence rates, but similar clustering also was evident among the States with the lowest rates. For example, nine Southern States were in the lowest fifth for past month use of alcohol, eight Southern States were in the lowest fifth for past month binge use of alcohol, and seven Southern States were among those indicating a high risk of binge drinking (population aged 12 years or older). By contrast, no Southern States were in the top fifth for current use of alcohol, no Southern State had the highest rates of binge alcohol use, and no Southern State was in the lowest fifth for perceived risk of binge drinking for persons aged 12 or older (Figures 3.1, 3.5, and 3.9). Similarly, nine Southern States comprised the category of States with the highest perceived risk of using marijuana occasionally, five Southern States had the lowest rates of past month marijuana use, and seven Southern States had the lowest rates of marijuana incidence. No Southern State was in the group of States with the lowest perceived risk of marijuana for persons aged 12 or older (Figures 2.9, 2.13, and 2.17).

States with the lowest rates of serious mental illness (SMI) represented an even mixture of all four regions for persons aged 18 or older: three from the Northeast, three from the South, three from the Midwest, and one from the West (Figure 6.1). The State with the lowest rate was Hawaii (7.2 percent). States in the highest fifth were somewhat more clustered geographically in the South (four States) and the West (4 States). Rhode Island had the highest rate of SMI (11.0 percent) in 2002–2003. Six of the eight most populous States were ranked in the lowest or next-to-lowest fifth: Pennsylvania (7.6 percent), Illinois (7.8 percent), Florida (7.9 percent), California (8.2 percent), Texas (8.4 percent), and Michigan (8.4 percent). Persons aged 18 to 25 had higher rates of SMI than did the 26 or older age group. In the 18 to 25 age group, Illinois had the lowest rate (11.8 percent), and Rhode Island had the highest rate (16.8 percent) (Table B.21).

7.2. Comparisons with Prior Estimates and Rankings

As discussed in Chapter 1, the 2002–2003 State rankings for each of the 21 substance use measures should not be compared with the rankings in 1999 through 2001 due to significant NSDUH methodological changes implemented in the survey in 2002. Because the State sample sizes in each survey year are generally not large enough to detect changes at the State level from 1 year to the next, the data for 2 years were combined to provide improved State estimates of prevalence levels and rankings. State estimates are available using the combined 1999–2000, 2000–2001, and 2002–2003 data. Combined-year estimates were presented in the 2000 and the 2001 State reports (Wright, 2002a, 2002b, 2003a, 2003b).

It is important to note, however, that although a State's rank from 2001 or earlier is not comparable with its rank for 2002 and later, there is a degree of consistency in overall rankings. This may be expected because of the consistency of the data collection methodology and the SAE methodology across those States and years. For example, comparing the estimates for past month use of marijuana for 1999–2000, 2000–2001, and 2002–2003, it can be seen that for the 10 States ranked in the highest fifth for 1999–2000, 9 of them were in the top fifth in 2000–2001 and 8 were in the top fifth in 2002–2003. For the lowest fifth, the relationship was somewhat weaker. Eight States in the lowest fifth in 1999–2000 were in the same fifth in 2000–2001, but only five States were ranked similarly in 2002–2003. For past month use of cigarettes, the relative ranking was similar for the highest fifth. Of the 10 States ranked highest in 1999–2000, 9 also were present in 2000–2001 and 8 were present in 2002–2003. For less prevalent substances like cocaine, there was less continuity both in the highest fifth and the lowest fifth. Of the nine States ranked in the top fifth for 1999–2000, eight were the same in 2000–2001, but only five were ranked similarly in 2002–2003. In the fifth with the lowest prevalence rates of past year use of cocaine (11 States) in 1999–2000, only 3 were ranked similarly in 2000–2001 and only 4 were still present in 2002–2003.

Because the 2002–2003 estimates were partly based on data for 2002, in conjunction with the fact that the national prevalence of use of most illicit drugs remained stable between 2002 and 2003, the rankings of States based on the 2002–2003 data were similar to the rankings based only on the 2002 data. With a combined sample over 2 years, the estimates tended to be more precise than those based on 2002 data alone as reflected in smaller prediction intervals (PIs). Estimates for the States with the largest populations that had larger sample sizes (between 3,600 and 3,800 for a single year, including California, Florida, Illinois, Michigan, New York, Ohio, Pennsylvania, and Texas) tended to be more consistent between 2002 and 2002–2003 than were estimates for the other 43 States. For example, in large-sample States, estimates of the prevalence rates of past month use of an illicit drug among persons aged 12 or older only showed an average absolute difference of 0.18 percent between 2002 and 2002–2003, but the same measure among small-sample States was 0.42 percent.

7.3. Validation

Given the unique NSDUH design and limited availability of independent data sources that provide State-level estimates, it is difficult to validate NSDUH State estimates using external sources. State estimates from prior years of NSDUH have been compared with estimates from the Behavioral Risk Factor Surveillance System (BRFSS) and the Youth Risk Behavior Survey (YRBS) sponsored by the Centers for Disease Control and Prevention (CDC, 2004a, 2004b). However, these CDC surveys (a) did not focus extensively on substance use, (b) employed different data collection methods, (c) did not cover all of the States on an annual basis, and (d) had varying degrees in potential response and nonresponse bias. It is, therefore, difficult to know how much confidence should be placed on comparing the results of surveys that are so different in design and implementation.

Although external validation of NSDUH findings is problematic, internal validation of the State estimates with NSDUH data can be useful. Because the State prevalence levels for 2002–2003 were estimated in the same manner as they were for earlier years, the procedures and the results of the validation done for prior estimates apply to these estimates.7 The average relative absolute bias (RAB) values from the 2000 State report (produced by pooling the 1999 and 2000 NHSDAs) that compared large-sample benchmark values with small-sample hierarchical Bayes estimates (see Tables A.22 to A.25 of the 2000 State report) were as follows (Wright, 2002b):

These results suggest that, if the true value of past month use of marijuana for persons aged 12 or older in a State with a sample of about 1,800 persons was 5 percent, the small area estimate would, on average, fall within 0.2 percent (4.07 x 5 percent) of the true value. The precision of these estimates was better than that from corresponding design-based estimates of the same sample size. By combining 2 years' data, the PIs for the hierarchical Bayes estimates were about 25 to 35 percent shorter, depending on the substance, than the corresponding design-based intervals.

As noted in past State reports, the models may not adequately adjust for differential nonresponse and bias effects at the State level. Any such bias resulting from nonresponse that varied in relation to the prevalence rates would raise concerns about comparisons among States.8 For such bias to exist after nonresponse adjustments have been made requires that the true probabilities for persons to respond to the survey still depend to some degree on whether they used a substance or not.


End Notes

7 For details, see Appendix A, Section A.4.2, of the 2000 State report (Wright, 2002b).

8 Tables A.1 through A.6 of Appendix A in this report provide response rates for 2002, 2003, and 2002–2003 combined.

Go to the Table of Contents

This page was last updated on June 03, 2008.