NESARC

Final Project - NESARC Dataset

The association between quantity of alcohol consumed and alcohol abuse and/or dependence among adults

Anuradha Muralidhar, Tampa FL, Course - Passion Driven Statistics (Wesleyan University through Coursera)

 Introduction

Alcohol dependence is substance related disorders in which an individual is addicted to alcohol with physically or mentally and continues to use alcohol despite significant areas of dysfunction. People become addicted to alcohol at some point in their life. Some depend on alcohol over a period of time and others immediately. Many surrounding factors contribute towards this association. Major Depression affects about 6.7% of US population over the age of 18. Furthermore, those with depression might consume more alcohol compared to those without. Alternatively alcohol is sometimes used to cope up with outside pressures and Depression as we know affects the person’s family and personal relationships, work or school habits, and general health.

 Research Questions:

Is drinking level associated with alcohol abuse and/or dependence?

Is the association between drinking and alcohol abuse and/or dependence similar for individuals with and without major depression?

 Methods

Sample:

Adults between the ages of 18 and 31 who consumed alcohol in past 12 months (n=7375) were considered for the analysis from the National Epidemiologic Study of Alcohol and Related Conditions (NESARC) Dataset.  The sample is from the first wave of the survey which was conducted in 2001-2002 and represents the civilian, non-institutionalized adult population of the United States.

Measures:

Major depression - MAJORDEPLIFE was assessed using the NIAAA, Alcohol Use Disorder and Associated Disabilities Interview Schedule – DSM-IV (AUDADIS-IV) with a value of 1 and 0.

Current alcohol (numalcyr_est) consumption was evaluated through quantity S2AQ8B (number of drinks of any alcohol usually consumed on days when drank alcohol in last 12 months. Furthermore, created drink category based on cumulative distribution of numalcyr_est.

Alcohol abuse and/or Dependence variable - ALCABDEP12DX was re-categorized to have 2 values as ALC12DEP with a value of 1 and 0.

 image

image

Rates of past year alcohol abuse and/or dependence (ALC12DEP) and lifetime major depression (MAJORDEPLIFE) Proc Freq used.

image

Rate of Alcohol abuse and/or dependence (categorical response variable) by major depression (present or absent, categorical third variable). Chi Square analysis, Proc freq procedure. Interpret column percents (18.56% and 26.72%).

image

Examining the association of Alcohol abuse and/or dependence (categorical response variable) and Drink Category (categorical explanatory variable) using Chi Square analysis, Proc freq procedure used. Interpret column percents (0.52%, 3.91%, 13.80%. 25.52%, 53.30%).

 Results

Univariate:

When examining the association between number of alcohol drinks consumed (quantitative response variable) and past year alcohol abuse and/or dependence (categorical explanatory variable), an Analysis of Variance (ANOVA) revealed that among adult drinkers, those with alcohol abuse and / or dependence reported drinking significantly more alcohol per day (Mean=746.69, s.d. ±1088.60) compared to those without alcohol abuse and/or dependence (Mean=151.77, s.d. ±337.67), F(1, 1497)=1274.49, p <.0001.

More than 20% of daily adults who consumed alcohol met the criteria for Alcohol abuse and/or Dependence in the past year.

A total of 21% met criteria for major depression at some point in their life.

image

Bivariate:

When examining the association between past year Alcohol abuse and/or dependence (categorical response variable) and Drink Category (categorical explanatory variable), a chi-square test of independence revealed that among adults drinkers, those who consumed alcohol almost every day were more likely to have alcohol abuse and/or dependence, X2 =1699.12, 4 df, p=0001. That is, more the alcohol consumed, the more likely they are to abuse and/or dependent on alcohol.

Adults who consumed alcohol daily with major depression (26.66%) were significantly more significantly more likely to meet criteria for alcohol abuse and/or dependence than those without major depression (18.51%), X2 =  422.6258, 4 df, p<.0001

image

Examining the association of Alcohol abuse and/or dependence (categorical response variable) and Drink Category (categorical explanatory variable) for those without major depression (=0)using Chi Square analysis, Proc freq procedure used. Interpret column percents (0.47%, 2.50%, 11.48%. 22.67%, 50.36%).

image

Examining the association of Alcohol abuse and/or dependence (categorical response variable) and Drink Category (categorical explanatory variable) for those with major depression (=1) using Chi Square analysis, Proc freq procedure used. Interpret column percents (0.71%, 8.31%, 21.89%. 37.50%, 64.51%).

Moderation

Major depression did not moderate the association between alcohol consumption and alcohol abuse and/or dependence (i.e. alcohol consumption was positively and significantly associated with alcohol abuse and/or dependence for those with and without major depression). Instead, among adults with daily alcohol consumption, at each level, the probability of alcohol abuse and/or dependence was significantly higher among those with major depression than those without major depression (See Bar Graph below)

image

Discussion

What might the results mean?

Individuals with major depression seem to be more sensitive to Alcohol Abuse and/or Dependence across the levels of alcohol consumed.

Strengths

Results are based on a large nationally representative sample of U.S. adults who consumed alcohol daily.

Limitations

Analysis was done on only major depression lifetime variable. The findings are based on cross-sectional data and do not reflect the alcohol consumption at which alcohol abuse and/or dependence emerges among those with and without major depression.

Recommended Future Research

Further research is needed to determine whether individuals with major depression are more sensitive to alcohol abuse and/or dependence then without major depressions. Other physical and/or psychological factors should be considered. Also the analysis should be expanded to the remaining ages in the population.

Resources

NESARC Codebook

Alcohol Abuse: http://en.wikipedia.org/wiki/Alcohol_abuse

Major Depression: http://en.wikipedia.org/wiki/Major_depressive_disorder

Statistical Program:

/*NESARC - SAS CODE*/

libname mydata “/courses/u_coursera.org1/i_1006328/c_5333” access=readonly;

data nesarc;

 set mydata.nesarc_pds;

label alcabdep12dx=’Alcohol abuse or dependence in last 12 months’

 s2aq2=’Drank at least 12 drinks in last 12 months’

 s2aq3=’Drank at least 1 drink in last 12 months’

 s2aq8a=’Frequency  of alcohol comsumption’

 s2aq8b=’Quantity of alcohol drinks consumed’

 usfreqyr=’No of days of alcohol consumption in last 12 months’

 numalcyr_est=’Estimate the number of alcohol drinks consumed in last 12 months’;

/*code missing data*/

if s2aq2=9 then s2aq2=.;

if s2aq8a=99 then s2aq8a=.; 

if s2aq8b=99 then s2aq8b=.; 

/*Skip pattern Code Valid Data*/

if s2aq8a=. then s2aq8a=11; 

if s2aq8b=. then s2aq8b=0; 

/*Create measurement in meaningful way*/

/*get the number estimated days where alocohol was consumed in last 12 months*/

if s2aq8a=1 then usfreqyr=365; /*Every Day which in turn equals to 365 in yr*/

 else if s2aq8a=2 then usfreqyr=286; /*Nearly Every Day 5.5*52*/

else if s2aq8a=3 then usfreqyr=182; /*3 to 4 times a week 3.5*52*/

 else if s2aq8a=4 then usfreqyr=104; /*2 times a week 2*52 */

else if s2aq8a=5 then usfreqyr=52; /* once a week 1*52*/

 else if s2aq8a=6 then usfreqyr=30;  /*2 to 3 times in a month 2.5 *12months*/

else if s2aq8a=7 then usfreqyr=12; /*once a month —> 12 times in a yr*/

 else if s2aq8a=8 then usfreqyr=9; /* 7 to 11 times in a yr—> 9 */

else if s2aq8a=9 then usfreqyr=4.5; /*3 to 6 times in a yr —> 4.5*/

 else if s2aq8a=10 then usfreqyr=1.5; /*1 to 2 times in a yr—> 1.5*/

else if s2aq8a=11 then usfreqyr=0; /*NA*/

/*Create Secondary Variables - Estimate the number of drinks in last 12months*/

numalcyr_est=usfreqyr*s2aq8b;

/*Create drink category based on cumulative distribution of numalcyr_est*/

if numalcyr_est >0 and numalcyr_est LE 5 then DrkCategory=1; /*3*/

 else if numalcyr_est > 5 and numalcyr_est LE 30 then DrkCategory=2; /*17*/

else if numalcyr_est > 30 and numalcyr_est LE 100 then DrkCategory=3; /*35*/

  else if numalcyr_est > 100 and numalcyr_est LE 360 then DrkCategory=4; /*130*/

else if numalcyr_est GT 360 then DrkCategory=5; /*35410*/

/*Create 2 groups for Dependent or response variable*/

if alcabdep12dx = 0 then alc12dep=0;

 else if alcabdep12dx in (1,2,3) then alc12dep=1;  /*Alcohol abuse and / or Depeendence*/

/*Create categories for age*/

/*creating grouping variables by spliting the observations in 25, 50, 75 percentiles*/

/*if age <32 then agegrp=1;*/

/* else if age >=32 and age <= 43 then agegrp=2;*/

/*else if age >=44 and age <= 59 then agegrp=3;*/

/* else if age >= 60 then agegrp=4;*/

if s2aq3=1; /*retain who drank in last 12 months*/

if age <32;  /*Adults less than 32,used only agegrp=1*/

proc sort data=nesarc;

by idnum;

/*Association between alcohol consumption and alcohol abuse and or alcohol dependence*/

proc anova; 

class alc12dep;

model numalcyr_est=alc12dep;

means alc12dep;

/*Chisq Statistic*/

proc freq data=nesarc;

tables  alc12dep*DrkCategory /chisq;

/*Pearson Coeffiecient test for Age, number of drinks consumed and */

/*largest number of drinks of any alcohol consumed*/

proc corr;

var age numalcyr_est s2aq8c;

/*sort by the third car varaible*/

proc sort data=nesarc;

by majordeplife; 

/*Association between alcohol consumption and alcohol abuse and or alcohol dependence along */

/*with moderating third cat varaible majordeplife*/

proc anova; 

class alc12dep;

model numalcyr_est=alc12dep;

means alc12dep; by majordeplife;

/*Chisq statistics for alc12dep by majordeplife*/

proc freq;

 tables alc12dep*majordeplife/ chisq;

/*Chisq with third cat variable majordeplife*/

proc freq;

 tables alc12dep*DrkCategory /chisq;

 by majordeplife;

/*Pearson Coeffiecient test for Age, number of drinks consumed and */

/*largest number of drinks of any alcohol consumed*/

/*with moderating third cat varaible majordeplife*/

proc corr;

var numalcyr_est s2aq8c;

by majordeplife;

proc freq;

tables  age majordeplife alcabdep12dx alc12dep DrkCategory;

/*s2aq8b s2aq2 s2aq3 s2aq8a s2aq8b sex usfreqyr numalcyr_est*/

run;

/*Bonferroni AdjustmentPair-Wise comparisons*/

/*Bonferroni Correction = 0.05/10=0.005*/

data comparison1; 

 set nesarc;

 if DrkCategory=1 or DrkCategory=2;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

data comparison2; 

 set nesarc;

 if DrkCategory=1 or DrkCategory=3;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison3; 

 set nesarc;

 if DrkCategory=1 or DrkCategory=4;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison4; 

 set nesarc;

 if DrkCategory=1 or DrkCategory=5;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison5; 

 set nesarc;

 if DrkCategory=2 or DrkCategory=3;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison6; 

 set nesarc;

 if DrkCategory=2 or DrkCategory=4;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison7; 

 set nesarc;

 if DrkCategory=2 or DrkCategory=5;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison8; 

 set nesarc;

 if DrkCategory=3 or DrkCategory=4;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison9; 

 set nesarc;

 if DrkCategory=3 or DrkCategory=5;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

 data comparison10; 

 set nesarc;

 if DrkCategory=4 or DrkCategory=5;

 proc sort; by idnum;

 proc freq;

 tables alc12dep*DrkCategory /chisq;

 run;

Output from Statistical Program:

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image


image

image

Assignment 9 - Moderation

Evaluation of potential third variable moderator - major depression in the context of ANOVA, Chi Square or Correlation and interpretation of the results.

Major depression did not moderate the association between alcohol consumption and alcohol abuse and/or dependence (i.e. alcohol consumption was positively and significantly associated with alcohol abuse and/or dependence for those with and without major depression). Instead, among adults with daily alcohol consumption, at each level, the probability of alcohol abuse and/or dependence was significantly higher among those with major depression than those without major depression (See Bar Graph above)

Assignment 8 - Correlation

Pearson Correlation Test using SAS and interpretation of the results

Among young adult drinkers, the correlation between number of alcohol drinks consumed in last 12 months (quantitative explanatory variable) and the largest number of alcohol drinks consumed (quantitative response variable) was 0.28 (p<=.0001), suggesting that only 8.3% (i.e. 0.28 squared) of the variance in largest number of alcohol drinks consumed can be explained by number of alcohol consumed in last 12 months.

Assignment 7 - CHI Square

Chi Square Test of Independence using SAS and interpretation of the results

When examining the association between past year Alcohol abuse and/or dependence (categorical response variable) and Drink Category (categorical explanatory variable), a chi-square test of independence revealed that among young adults drinkers, those who consumed alcohol almost every day were more likely to alcohol abuse and/or dependence, X2 =1699.12, 4 df, p=0001.

Assignment 6 - ANOVA

Analysis of Variance using SAS and interpretation of the results

When examining the association between number of alcohol drinks consumed (quantitative response variable) and past year alcohol abuse and/or dependence (categorical explanatory variable), an Analysis of Variance (ANOVA) revealed that among adult drinkers below age 32, those with alcohol abuse and / or dependence reported drinking significantly more alcohol per day (Mean=746.69, s.d. ±1088.60) compared to those without alcohol abuse and/or dependence (Mean=151.77, s.d. ±337.67), F(1, 1497)=1274.49, p=0001.

Assignment 5: Graphing 2

For this assignment using SAS graphing wizard, here is the visualization which shows an association between two variables.