\lWlllNlUlWlmNW“NLIHHHIWMHWNW 128 246 'THS 1lIIIIIIIHlIIIllIllllllllll/lllWHllH/l/I/Ill/"Willi 129 930 99257 THESIS | LIBRARY Michigan State Unlversity This is to certify that the thesis entitled Smoking, Advice to Quit, and Socioeconomic Status -- the Case of Pregnant Women in U.S., 1990 presented by Chiungyu Cheng has been accepted towards fulfillment of the requirements for MoA. degree inSOCiOlOEZ/Urban Studies Jew Major professor Date 3/7’i/¢r 0-7639 MS U i: an Affirmative Action/Equal Opportunity Institution PLACE ll RETURN BOX to remove this checkout from your record. TO AVOID FINES Mum on or him dd. duo. ATE DUE DATE DUE DATE DUE :JL l l MSU Is An Affirmative Action/Equal Opportunity Inflation WW1 SMOKING, ADVICE TO QUIT, AND SOCIOECONOMIC STATUS - THE CASE OF PREGNANT WOMEN IN US, 1990 By Chiungyu Cheng A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology/Urban Studies 1995 ABSTRACT SMOKING, ADVICE TO QUIT, AND SOCIOECONOMIC STATUS - THE CASE OF PREGNANT WOMEN 1N US, 1990 By Chiungyu Cheng This thesis discuss three questions. First, why do some women continue smoking after they learned that they are pregnant, but others stop? Second, would pregnant women comply with the advice to quit smoking fi'om doctors? And third, why do some women continue smoking after they learned that they are pregnant, but others stop? Three hypotheses were built. The three factors - doctor's advice to stop smoking, socioeconomic status, and amount smoked per day before pregnancy - were assumed to be important predictors of pregnant women's smoking behavior. The data is from "National Health Interview Survey, 1990: Health Promotion and Disease Prevention (HPDP) Pregnancy and Smoking Supplement”. All hypotheses were tested by ordinary least squares regression. The results indicate that socioeconomic status is not an important factor to predict whether a pregnant woman receive doctor's advice to stop smoking or not. The most important factor to predict whether a pregnant woman receive advice to quit is the amount she smoked per day before pregnancy. Moreover, socioeconomic status and doctor’s advice to quit smoking do not impact pregnant women's smoking behavior strongly. ACKNOWLEDGEMENTS I would like to express my appreciation to my advisor, Dr. Stan Kaplowitz, for his patience,valuable advice, and time. I also like to give my thanks to Dr. Harry Perlstadt and Dr. John Schweitzer for serving on my thesis committee. I really appreciate my friend Ryan Simmons and Yu-Hsi Lin for their assistance in data analysis. A The encouragement, support, and care that I received from my mother, my family, my boy fi'iend, Chung-Chin Cheng, and many of my friends are very important for the accomplishment of this thesis. Finally, a note of special thanks goes to God for everything He does for me. TABLE OF CONTENTS Page LIST OF TABLES ................................................................................ v 1. LITERATURE REVIEW AND PROBLEM STATEMENT .............. 1 SOCIOECONOMIC STATUS AND HEALTH .......... 6 SMOKING AND SES ................................................ 10 THE PURPOSE OF THIS PAPER ............................. 11 H. HYPOTHESES ................................................................................. 12 Hypothesis One .......................................................... 12 Hypothesis Two ......................................................... 13 Hypothesis Three ....................................................... 14 III. METHOD ....................................................................................... 16 IV. RESULTS ....................................................................................... 23 Hypothesis One ......................................................... 25 Hypothesis Two ........................................................ 26 Hypothesis Three ...................................................... 28 V. DISCUSSION AND CONCLUSION ............................................... 30 DISCUSSION .......................................................... 30 CONCLUSION ........................................................ 36 APPENDIX ........................................................................................... 38 REFERENCES ...................................................................................... 44 iv LIST OF TABLES (cont’d). Table Page A-l. Zero-order Correlations of Education, Income, Amount Smoked Per Day Before Pregnancy, Smoked During Pregnancy, Amount Smoked After Learning of Pregnancy, and Receiving Advice to Stop Smoking ................................................................................. 39 A-2. Efi'ects and Significance of Amomt Smoked Per Day Before Pregnancy, Education, and Income on Receiving Advice to Stop Smoking ......................................................................................... 40 A-3. Efi‘ects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy ........................................................................................ 40 A4. Effects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked Afier Learning of Pregnancy ................................. 41 A-5. Efl‘ects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy ........................................................................................ 42 A-6. Efi‘ects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked After Learning of Pregnancy ................................. 43 1. LITERATURE REVIEW AND PROBLEM STATEMENT It has been well accepted that maternal smoking is hazardous to the fetus. Many reports have documented that the increased frequency of spontaneous abortion, placentae previa, abruptio placentae, premature and prolonged rupture of membranes and low birthweight which are caused by maternal smoking. (Smoking and Health, 1990; Kleinmann, Pierre, Madans, Land, & Schramm, 1988; Stein, Kline, Levin, Susser, & Warburton, 1980). In 1957, Simpson and Lomal have found that smoking may increase the risk of low-weight babies (<2500 g), and this hypothesis has been well supported by a lot of following studies (Lumley & Astbury, 1989). Recent studies show that smoking increases the risk of late fetal and early neonatal death (I-Iaglund & Cnattingius, 1990; Cnattingius, Haglund, & Mein'k, 1988). Moreover, some researches have found that smoking may increase the risk of retarded learning and childhood cancer (John, Savitz, & Sandler, 1991; Naeye & some researches have found that smoking may increase the risk of retarded learning and childhood cancer (John, Savitz, & Sandler, 1991; Naeye & Peters, 1984). In sum, it is obvious that maternal smoking will seriously harm the health of fetus and newborns. Although it has been proven that smoking will cause negative health consequences to the fetus, many pregnant women still smoke. For example, in Norway, 35percent of women smokers continue to smoke after their first trimester (Peen et al., 1991); in America, over 60 percent of women smokers continue to smoke during their pregnancy (National Health Interview Survey, 1990). The main purpose of this paper is to discuss the question: why do some women continue smoking after they learn that they are pregnant, but others stop? A possible reason a woman continues to smoke during pregnancy is that she didn't get advice to stop smoking from any medical professional. Although the efficacy of physicians' advice to quit smoking has varied, most studies have shown that physicians' encouragement of smoking cessation can increase the patients' intention to stopping smoking (Glynn, 1990; Ockene, 1991; Commings et al., 1989; Pederson, 1982). For example, the 1979 Surgeon General's report on smoking and health found that 10 percent to 25 percent of smokers who were advised to quit smoking by doctors may quit or reduce the amount they smoke (Smoking and Health, 197 9). However, many studies indicate that most smokers didn't get advice to quit smoking from their doctors. In 197 5, a study found that only 35 percent of male smokers and 38 percent of female smokers have ever been cautioned to stopping smoking by their physicians (Adult Use of Tobacco, 1975). A 1980/1983 study found that only 42 percent male smokers and 46 percent female smokers had received advice to quit from physicians (Anda et al., 1987). Moreover, a recent 10 year study about smoking cessation carried out by the Stanford Center for Research in Disease Prevention found that among 2710 smokers, fewer than 50 percent said that their physicians ever advised them to stop (Hafi‘erty & Salloway, 1994). According to 1990 National Health Interview Survey, among the 2776 female smokers who were currently pregnant or had a child in the past five years, only 49 percent people reported that doctors had ever advised them to stop smoking. (National Health Interview Survey, 1990). Some studies also indicated that even though smokers have more doctor visits, most of them still didn‘t get advice to quit smoking from doctors. For example, Anda et a1. (1987) found that only 46 percent of the smokers who reported having a routine checkup with a physician in the last year reported having been told to quit, and only 53 percent of smokers who had ten or more physician contacts in the last year had been told to. Similarly, a 1993 report found that among the persons with four or more doctor visits in a year, only 45 .5 percent of them got advised to quit smoking from doctors (Epidemiology Br., 1993). Moreover, in the 1990 National Health Interview Survey, only 64 percent of pregnant smokers who had ten or more doctor visits in past 12 months reported that they had ever been advised to quit smoking by doctors (National Health Interview Survey, 1990). There are two interesting questions. First, for the women who have already received advice to stop smoking, did they comply with doctors? Second, why do some smokers get advised to quit smoking from doctors, but others don't? Some perspectives from Health Belief Model have been used to explain patient compliance (Becker, 197 4). Becker claimed that although a person recognizes personal susceptibility, he/she does not take any action rmless he/she also believes that becoming ill would cause serious organic and social repercussions (Becker, 1974). One thing should be emphasized here is that, susceptibility refers to the person's subjective perceptions rather than to medical estimate of how serious the illness may be .1 Many studies have proven that there is no association between medical views of the problems severity and patient compliance (Pragofl‘, 1962; MacDonald, Hagberg, Grossman, 1963; Johannsen, Hellmuth, Sorauf, 1966; Chamey et a1, 1967; Davis, 1968; Bonnar, Goldberg, Smith, 1969; Becker, 1974). The pattern of communication between physicians and patients would also influence patient compliance. Davis noted that "patterns of communication which deviate fi'om the normative doctor-patient relationship will be associated with patients' failure to comply with doctors' advice" (Davis, 1968). Such deviation includes that the physician is formal, rejecting, controlling, disagrees completely with the patients, or interviews the patient without feedback (Becker, 1974). Some studies reported that patients compliances are better when they are satisfied with intial contact, perceives the physician as fiiendly, and feels that the doctor understand the complaint (Francis, Korsch & Morris, 1969; Korsch, Gozzi, & Francis, 1968) W Toward the second question, some studies have found that heavier smokers were more likely to get advice to quit smoking than lighter smokers (Anda et al., 1987; Frank, Winkleby, Altman, Rockhill, & Fortmann, 1991). It seems that physicians will refer to the patient's smoking history to decide whether to give the patient quitting advice or not. However, there are very few studies dealing with this question from the view of social structure. Many studies have indicated that socioeconomic status has strong association with health status and health behavior (I will discuss this in next part). In this paper, I would like to see if socioeconomic status is an important predictor to explain the two questions I mentioned above. Socioeconomic status (SES) describes the position of a person in society. Social stratification is the assignment of persons into different positions in the social structure, and is the unequal distribution of social resources. Desirable social resources are usually owned by people in higher SES. Generally, the unequal distribution of social resources makes people in different SES have different living circumstances ,life style, and behavior. As Green wrote, conceptually, an index of socioeconomic status is intended to reflect the balance or net effect of social, environmental, situational, educational,financial, and other forces in the individual's personal world. (Green, 1971, p.54) Education, income, and occupation are three major indicators to measure SES (Liberatos Link, and Kelsey, 1988; Mueller, 1981; Hollingshead, 1971; Green, 1970; Duncan, 1961). Many studies have found the association between SES and health. SES could reflect somewhat difi‘erent individual and social forces associated with health and disease. For example, income reflects spending power, housing, diet, and medical care; occupation measures prestige, responsibility, physical activity, and work exposures; and education indicates skills requisite for acquiring positive social psychological, and economic resources. (Winkleby, 1992, p. 186) Generally, many health problems are related to some factors that differ by SES, such as persons' life-style (such as smoking, exercise, or diet), degree of compliance with advice of medical professionals, knowledge about health, and ability to access medical care. Among the three SES indicators, education is the best predictor of health status (Winkleby et al., 1992; Funchs, 1979; Kitagawa and Hauser, 1973). Winkleby et al said that compared to income and occupation, education has more influence on health since it may protect against disease or risk-health behavior by influencing life-style behaviors (such as smoking, exercise, and diet), problem-solving abilities, and values (Liberatos, Link, & Kelsey, 1988). Some other studies concluded that education is a good predictor of health behavior because more educated persons are more aware of health risks and more likely to change their behavior to reduce health risks (Fuchs, 1979; Grossman, 1976; Rosenstock, 1975)‘. Moreover, persons with higher education level are more attentive to health information and more trusting of the claims of science (Lefcowitz, 197 3). One explanation for the SES-health gradient is that persons in lower SES have fewer opportunities to access medical service (Wan and Gray, 197 8). However, Adler et a1. (1993) found that even if individuals have equal opportunities to access medical services, true access may still differ among persons in difierent SES. The reason is that "individuals with more education and income, who may be more skilled in dealing with bureaucracies and social systems, may be more efficient in determining who provides the best care and in obtaining care when needed". Moreover, many studies have ' Some studies found that for women, education has strong efl'ect on their health status, but for men, income is the most important predictor (Fuchs, 1979; Kitagawa and Hauser, 1973). A 1988 study indicates that poor pregnant women are less likely to receive prenatal care than the nonpoor (Nersesian, 1988). Compared to education and income, occupation has relatively little efi'ect on health status (Coburn and Pore, 1974). indicated that persons in higher SES receive better medical care than persons in lower SES (Dufi‘ and Hollingshead, 1968; Brotheston, 1976). The characteristics of the health care system (Rundall and Wheeler,197 9) and people's attitude toward health and health care (Suchman,1965) are also causes of unequal use of medical care. Medical services are for middle-class in origin (McKinlay, 197 5). Its culture is middle-class culture. As Adler et a1. (1993) said "Health promotion programs and physician recommendations are frequently designed with upper- individuals (higher SES individuals) in mind". Persons in the lower-SES feel that their interaction with the medical system is a dehumanizing experience (Rundall and Wheeler, 1979). Therefore, elimination of financial barriers cannot really resolve disparities in health care (Williams, 1990) Patients' social background should also be considered to resolve the disparities. As Alder et a1 suggested, physicians should pay attention to patients' background when they give health advice. He said, physicians need to recognize not only the importance of discussing risk behavior increases with patients who are lower on the SES ladder, but also that recommendations need to be tailored to the patient's life circumstance. (Alder et al., 1993, p.3144) 10 SMOKING AND SES Some smoking studies attempted to find the relationship between smoking behavior and SES. Some studies found that persons in lower SES are more likely to smoke (Alder et al., 1993). A study by Mattews et al. (1989) have found that the prevalence of smoking ranged from 45 percent of those with less than a high school education to 19 percent of those with an advanced degree. Winkleby et al. (1990) found that smoking has a negative strong association with years of education. Brownson et al. (1992) found knowledge about the health effects of smoking is lower among less-educated people. Besides these personal factors, there are rare studies dealing with the relationships between quitting behavior and SES. When the smokers meet the situation that something (such as their babies) may be hurt by their smoking behaviors, will they stop smoking? How does socioeconomic status impact on their behavior? In this paper, I will also discuss this question. Besides SES, some studies also indicated that heavy smokers have more difficulties to quit smoking. For example, some studies found that high daily cigarettes consumption is related to difficulty to quit smoking (Coambs & Kozlowski, 1989; Cherry & Kiemen, 1976). Similarly, heavy smokers are 11 less likely to quit smoking even if they participate in smoking cessation programs (Coambs & Kozlowski, 1989; Kozlowski, 1979). THE PURPOSE OF THIS PAPER The purpose of this paper is to discuss the questions mentioned above. First, why do some persons get advice to quit smoking fiom doctors, but others don't ? Second, would pregnant women comply with the advice to quit smoking from doctors? And third, why do some women continue smoking after they learned that they are pregnant, but others step? Based on the above discussions, three sets of hypotheses are developed to answer the two questions. Because many studies have found that only education and income have significantly effect on health behaviors, in this paper, I would use education and income as two SES indicators. Occupation wouldn’t be concerned here. II. HYPOTHESES Hypothesis One: Assuming that other things being equal: Because use of medical service is unequal among people in different socioeconomic status (based on the above discussions), l-A) pregnant women with higher education level and with more family income are more likely to get advice to quit smoking from doctors, than women with lower education level and with less family income. Because heavier smokers are more likely to get advice to quit smoking than lighter smokers (based on the above discussions), l-B) pregnant women who smoke more cigarettes per day are more likely to get advice to quit smoking from doctors than women who smoke fewer cigarettes per day. 12 13 Hmthesis Two: Assuming that other things being equal: Because people in higher SES are more aware of health risks, and more likely to change their behavior to reduce health risks (based on the above discussion), 2-A) women with higher education level and with more family income are more likely to stop smoking during their pregnancy than women with lower education level and with less family income. Because that physicians' encouragement of smoking cessation can increase the patients' intention to stop smoking (based on the above discussions), 2-B) women who are advised to quit smoking by doctors, are more likely to stop smoking during their pregnancy than women who are not advised to quit smoking by doctors. Because heavy smokers have more difficulty quitting (based on the above discussions), l4 2-C) women who smoked fewer cigarettes per day before learning of their pregnancy are more likely to stop smoking during their pregnancy than women who smoked more cigarettes per day. Hypothesis Three: Hypothesis three is very similar to hypothesis two. Hypothesis two predicts that some factors could influence pregnant smokers to continue or stop smoking after they learned that they were pregnant. However, in hypothesis three, the concern focuses on this: although some smokers continue smoking after they learned that they are pregnant, the amount they smoked may still be influenced by the factors mentioned in hypothesis two. As with hypotheses one and two, this set of hypotheses also assume that other things are equal. Because people in higher SES are more aware of health risks, and more likely to change their behavior to reduce health risks (based on the above discussions), 3-A) women with higher education level, and with more family income would smoke fewer cigarettes during pregnancy than women with lower education level and with less family income. 15 Because that physicians' encouragement of smoking cessation can increase the patients' intention to stopping smoking (based on the above discussions), 3-B) women who are advised to quit smoking by doctors would smoke fewer cigarettes during pregnancy than women who are not advised to quit smoking by doctors. Because heavy smokers have more difficulty with quitting (based on the above discussions), 3-C) women who used to smoke more cigarettes per day before learning of pregnancy would smoke more cigarettes during their pregnancy than women who used to smoke fewer cigarettes per day. III. METHOD The data used for testing above hypotheses is from "National Health Interview Survey, 1990: Health Promotion and Disease Prevention (HPDP) Pregnancy and Smoking Supplement" (National Health Interview Survey, 1990). The original collector of the data is ICPSR, Inter-University Consortium for Political and Social Research. The universe of the National Health Interview Survey (NHIS) is the civilian, noninstitutionalized population of the 50 states and the District of Columbia. The sampling method is a multistage probability sampling design. Four independent representative samples, which may be used in any combination, were drawn. Black persons were oversampled. The sample of this survey is every woman who is 18-44 years old and was included in the NI-IIS (total=25,83 9). Smoking questions were asked of 16 17 women who had a live-bom child in past five years or were currently pregnant (total=747 7). The response rate for the pregnancy and smoking supplement of the NHIS was 89.6 percent (household response rate from NHIS, 95.5% multiplied by the Supplement response rate, 92.8%). The final basic weight is included on the data file for accurate population estimate. However, instead of the final basic weight, all analyses in this paper are weighted by the modified weights to avoid incorrect significance results. The analyses weighted by the original final basic weight would appear to have much more cases than the sample size does. Due to the inflation of cases, the impact of an independent variable on a dependent variable may look very significant even though the standard partial regression coefficient is near to zero. In order to avoid this problem, modified weights are used in this paper. The formula of calculating the modified weights is : the modified weight=the final basic weight x (the sample size/the sum of the final basic weight). The sum of the modified weight is equal to the sample size. Therefore, the problem of the inflation of sample size would not occur. The results of the analyses without weights are listed in Appendix A. The subsample for testing hypothesis one in this paper is women who were currently pregnant or had child in the past five years, and had ever 18 smoked 100 cigarettes in their entire lives. The size of the subsample is 2,755. For hypotheses two and three, only the women who had live-born child in the past five years and smoked during the 12 months prior to birth were considered. The number of cases discussed in hypothesis two is 1645 and for hypothesis three is 1500. The women who were currently pregnant are excluded when testing hypothesis two and three. The reason is because they are not appropriate for testing hypotheses two and three. Hypotheses two and three predicted smoking behavior of pregnant women during their pregnancy, but the answers of the women who were currently pregnant when they were interviewed cannot represent their smoking behavior during the whole period of pregnancy. All hypotheses were tested by ordinary least squares (OLS) regression. Actually, for hypotheses one and two, the dependent variable is dichotomous. Therefore, OLS regression is not appropriate and other techniques such as logistic regression are more appropriate. I have, however, used OLS regression for two reasons. First, my statistics training has not included any of these other techniques. Second, when the two categories of the dependent variable are divided almost equally, the conclusions to be drawn from OLS are virtually the same as from logistic regression (Gillespie, 1977), and the l9 dependent variables of hypothesis one (advice) was divided into two almost exactly equal categories. For hypothesis two, 63% is in one category, and 37% in the other. The two categories are close enough to use OLS instead of other techniques (refer to table one). The explanation of the variables used in this paper is listed in Table 1. 20 Table 1: Variables List (In parentheses is the condition that had to be met for the question to be asked) Name of Variables lterns and Codes frequencies ADVICE Doctor Ever Advised Quitting Smoking (Currently pregnant or had child in past 5 years and ever smoked) 0. No 1395(51%) 1. Yes l360(49%) Non-missing total 2755(100%) SMOKE Smoked During Most of Pregnancy DURING PREG (Hadchildinpastfive yearsandsmoked during the 12 months prior to birth) 0. No“ 103 l(63%) 1. Yes 614(37%) Non-missing total 1645( 100%) " "No" was imputed for women who quit smoking after leaming of the pregnancy. SMOKE Amount Smoked After Learning of AFTER LEARN Pregnancy (Had child in past 5 years and smoked during the 12 months prior to birth) 00. less than one cigarette 377(25%) 01-14. 1-14 cigarettes 701(47%) 15-24. 15-24 cigarettes 330(22%) 25-60. 25-60 cigarettes 92(06%) 61+ more than 60 cigarettes 0(00%) Non-missing total 1500(100%) Mean=9.393 Standard Deviation=9.412 21 Table l (cont'd) EDUCATION Education of Individual-completed years 00. Never attended; Kindergarten only 0(00%) 01-11. Grades 1-11 597(22%) 12. Grade 12 1306(47%) College: 13. 1 year 228( 8%) 14. 2 years 230( 8%) 15. 3 years 91( 3%) l6. 4 years 212( 8%) l7. 5 years 29( 1%) 18. 6 years or more 79( 3%) Non-missing total 2772(100%) Mean=12.335 Standard Deviation=2.242 INCOME Family Income 0.5 Lessthan81000 23(1%) 1.5 31000-51999 28( 1%) 2.5 2000- 2999 34(1%) 3.5 3000- 3999 47( 2%) 4.5 4000- 4999 64(3%) 5.5 5000- 5999 41( 2%) 6.5 6000 - 6999 49( 2%) 7.5 7000 - 7999 46( 2%) 8.5 8000- 8999 42( 2%) 9.5 9000 - 9999 _ 49( 2%) 10.5 10000-10999 62( 2%) 11.5 11000-11999 38( 2%) 12.5 12000 - 12999 77( 3%) 22 Table 1 (cont’d) 13.5 13000 - 13999 44( 2%) 14.5 14000 - 14999 39( 2%) 15.5 15000 - 15999 50( 2%) 16.5 16000 - 16999 58( 2%) 17.5 17000 - 17999 50( 2%) 18.5 18000 - 18999 61( 2%) 19.5 19000 - 19999 71( 3%) 22.5 20000 - 24999 229( 9%) 27.5 25000 - 29999 232( 9%) 32.5 30000 - 34999 214( 9%) 37.5 35000 - 39999 167( 7%) 42.5 40000 - 44999 152( 6%) 47.5 45000 - 49999 116( 5%) 60.0 50000 and over 410(16%) Non-missing Total 2493(101%) Mean=28.862 Median=27.5 Standard Deviation=18.299 AMOUNT Number of Cigarettes Smoked Per Day SMOKED PER DAY (Ever smoked 100 regularly) BEFORE 00. Less than 1 cigarettes 54( 2%) PREGNANCY 01-14. 1-14 cigarettes l346(49%) 15-24. 15-24 cigarettes 1028(38%) 24-60. 24-60 cigarettes 293(11%) 61-90. 61-90 cigarettes 2( 0%) 91+ more than 90 cigarettes 0( 0%) Non-missing total 2723(100%) Mean=l4.593 Standard Deviation=7.336 Sourcezfrom codebook of ”National Health Interview Survey, 1990: Health Promotion and Disease Prevention (HPDP) Pregnancy and Smoking Supplement” IV. RESULTS Table 2: Zero-order Correlations of Education, Income, Amount Smoked Per Day Before Pregnancy, Smoked During Pregnancy, Amount Smoked after Learning of Pregnancy, and Receiving Advice to Stop Smoking Education Education Income 3901"" Amount -.03 25 Smoked Per Day Before Pregnancy Smoked -.1285"" During Pregnancy Amount -. 1394‘“ Smoked afier Learning Pregnancy F . . Adviceto -.0105 Stop Smoking Income .0197 -.ll78""‘l -.0857‘" .0198 Amount Smoked Amount Receiving Smoked Per During Smoked after Advice to Day Before Pregnancy learning Stop Pregnancy Pregnancy Smoking." 3290‘" 5905‘" .5736‘" 2198"" 2896“" 2107“" Sample Size=2755 '"' P<0.01 "W P<0.001 23 24 According to the above discussions, persons in lower SES are more likely to smoke. But the findings in Table 2 indicate that the relationships between SES (education and income) and smoking behavior (amount smoked per day before pregnancy, smoked during pregnancy, and amount smoked after learning of pregnancy) are all weak and negative (besides the relationship of income and amount smoked per day before pregnancy). And, as mentioned above, heavy smokers have more difficulty quitting. This idea is supported by the findings in Table 2. According to Table 2, the relationships between amount smoked per day before pregnancy and those variables, smoked during pregnancy, amount smoked after learning of pregnancy are all positive and quite strong. Moreover, the relationship between education and receiving advice to stop smoking is negative and very weak. The relationship between income and receiving advice to quit is positive and weak. Receiving advice to stop smoking has positive relationships smoking variables (amount smoked per day before pregnancy, smoked during pregnancy, and amount smoked after learning of pregnancy). 25 Hypgthesis One Table 3: Effects and Significance of Amount Smoked per Day Before Pregnancy, Education and Income on Receiving Advice to Stop Smoking Variables Std. Partial Partial Correlation Sig T Regression Coefficients Coeficients Amount Smoked .2185 .2202 .0000 Per Day Before Pregnancy Education -.0905 -.0842 .0000 Income -.0597 -.0558 .0059 Note: 1. Receiving Advice to Stop Smoking 1s coded 1, and not receiving it is coded O. 2. Multiple Regression Analysis 3. Sample Size=2755 Hymthesis l-A Hypothesis l-A is not supported by the results (please see Table 3). Opposite to the prediction that when the other variables are equal, pregnant women with higher education level and with more family income are more likely to get advice to quit smoking by doctors, the results indicate that pregnant women with lower education level and with lower family income are significantly more likely to be advised by their doctors to quit smoking. 26 However, the impacts of both education and income on receiving advice are quite weak. H thesis l-B Hypothesis l-B is supported by the findings (see Table 3). When the other variables are equal, pregnant women who used to smoke more cigarettes per day are significantly more likely to be advised to quit smoking by doctors. Hypothesis Two Table 4: Effects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy Variables Std. Partial Partial Correlation Sig T Regression Coefficients Coefiicients Education -.0856 -.0864 .0010 Income -. 1030 -. 1039 .0001 Advice .2414 .2522 .0000 Amount Smoked .2691 .2788 .0000 Per Day Before mane Notezl. Continuing Smoking During Most Pregnancy is coded 1. Stop Smoking During Most Pregnancy is coded 0. 2. Multiple Regression Analysis 3. Sample Size=1645 27 Hypothesis 2-A Hypothesis 2-A is supported by the results (see Table 4). The results indicate that when the other variables are equal, education and income have negative relations with continuing to smoke during pregnancy. That is, women with high education level and with more family income are significantly less likely to continue smoking after they learn that they are pregnant when the other variables are equal. However, the impact of both education and income on continuing smoking after learning of pregnancy are very weak. Hymthesis 2-B Hypothesis 2-B is not supported by the results. According to Table 4, receiving quitting advice has positive relationship with continuing smoking during pregnancy. That is, when the other variables are equal, pregnant women who are advised to quit smoking by doctors are significantly more likely to continue smoking during their pregnancy than are those who aren't advised to stop. Hypgthesis 2-C Hypothesis 2-C is supported by the results. According to Table 4, when the other variables are equal, the more cigarettes a pregnant woman 28 used to smoke per day before pregnancy, the higher the possibility that she continues to smoke during her pregnancy. Hypothesis Three Table 5: Efi‘ects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked after Learning of Pregnancy Variables Std. Partial Partial Correlation Sig T Regression Coefficients Coefficients Education -.0973 -.1 125 .0000 Income -.0601 -.0698 .0109 Advice .0854 . 1046 .0001 Amount Smoked .5700 .5743 .0000 Per Day Before Pregnancy Note:1.Amount Smoked alter Learning of Pregnancy is coded by the numbers of cigarettes pregnant women smoked after learning of pregnancy (0 to 95) 2. Multiple Regression Analysis 3. Sample Size=1500 Hypothesis 3-A Hypothesis 3-A is supported by the results. According to Table 5, when the other variables are equal, women with higher education level and with more family income would smoke significantly fewer cigarettes during 29 their pregnancy. However, the impact of both education and income on amount smoked per day after learning of pregnancy is very weak since the standardized partial regression coefficients of them are very close to zero. Hyp_othesis 3-B Hypothesis 3-B is not supported by the results. According to Table 5 , when the other variables are equal, women who are advised to quit smoking by doctors would significantly smoke more cigarettes during pregnancy than women who are not advised to quit smoking by doctors. However, the same as above, the impact of receiving advice to stop smoking on amount smoked after learning of pregnancy is very weak. Hymthesis 3-C Hypothesis 3-C is supported by the result. According to Table 5, controlling education. income, and advice, the more cigarettes a pregnant woman used to smoke per day, the more cigarettes she will smoke during her pregnancy. V. DISCUSSION AND CONCLUSION DISCUSSION Hypothesis 1-A is not supported by the findings. The findings indicate that when the other variables are equal, pregnant women who have lower education levels and lower family incomes have significantly more opportunities to be advised to quit smoking by their doctors. One possible reason is doctors may think that people with lower education level and with less family income usually smoked more cigarettes than women with higher education level and with more family income. Therefore, they have more opportunities to receive advice to quit smoking from doctors. Another reason may be that doctors think women with lower education level and with less family income might not recognize the danger of smoking and need doctors' advice to quit. 3O 31 Hypothesis l-B is supported by the findings since when the other variables are equal, pregnant women who used to smoke more cigarettes per day are significantly more likely to be advised to quit smoking from doctors. This findings are consistent with the results of some previous studies (Anda, et al in 1987, Frank, Winkleby, Altman, Rockhill, and Fortrnann in 1991). One possible explanation of this findings is doctors may think that heavier smokers are less aware of the negative influence of smoking on the health of fstus, and the amount of heavy smokers smoke every day would harm infant very seriously. However, if doctors only pay attention on heavier smokers, due to lack of quitting advice, lighter smokers may continue or even increase smoking after pregnancy to release some possible pressure caused by pregnancy. Among the three independent variables, income, education, and amount smoked per day before pregnancy, amount smoked per day before pregnancy has the strongest impact on receiving advice to quit smoking. The impact of education and income on receiving advice to quit smoking is significant but very weak. Hypothesis 2-A, when the other variables are equal, education and income have negative impacts on continuing to smoke during pregnancy, and 32 hypothesis 3-A, when the other variables are equal, women with higher education level and with more family income would smoke fewer cigarettes during pregnancy than women with lower education level and with less family income, are both supported. One possible reason is pregnant women with higher education level and with more family income are more aware of the danger while smoking during pregnancy and are more likely to take action to avoid the danger. However, as with the findings of hypothesis l-A, when the other variables are equal, the influences of education and income on both continuing to smoke during pregnancy and on amount smoked after learning of pregnancy are very low. It seems that education and, income are not really important factors to predict a pregnant woman's smoking behavior after learning of pregnancy. For hypothesis two and three, when subsample was replaced with the women who were currently pregnant, the results changed a little bit. When the subsample for testing hypothesis 2-A was replaced with the women who were currently pregnant’, the influence of education become stronger (please 2 In order to realize how the results would change when currently pregnant women are selected as sample, theanalysisusingcurrentlypregnantwomenassampletotesthypothesestwoandthreewerealsobeen performed. The results are in the Table A (for hypothesis Two) and Table B (for hypothesis Three). 33 compare Table A and Table 4). The difference may be caused by the error of recall. The pregnant women with higher education level may tend to recall Table A: Effects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy Variables Std Partial Regression Coeficients Partial Correlation Coefi'rcients Sig T Education -.2641 -.2607 .0000 Income -.1369 - -. 1404 .0149 Advice .3129 .3272 .0000 Amount Smoked .0033 .0037 .9489 Per Day Before W Note: 1.Continuing Smoking During Pregnancy is coded 1. Stop Smoking During Pregnancy is coded 0. 2.Tota1 cases=358(100%). The number of cases coded as 0 = l48(41%). The number of cases coded as 1=210(59%) 3.Multip1e Regression Analysis Table B:Efl‘ects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked after Learning of Pregnancy Variables Std Partial Regression Coefficients Partial Correlation Coefficients Sig T Education -.0506 -.0649 .3903 Income -.0275 -.0356 .6378 Advice .1679 .2231 .0028 Amount Smoked .6321 .6502 .0000 Per Day Before Pregnancy Note: 1. Amount Smoked after Learning of Pregnancy is coded by the numbers of cigarettes pregnant women smoked after learning of pregnancy (0 to 95) 2. Total cases=209(100%). 1-14 cigarettes=72(34%); 15-24 cigarettes=86 (41%); and 25-60 cigarettes= 51(24%) ; more than 60 cigarettes=0(0%) 3. Multiple Regression Analysis These results are somewhat different from the results measured by the sample of the women who had live-born child in past five years and smoked during the 12 months prior to birth (please refer to Table 4 and Table 5). 34 that they did not smoke cigarettes during their past pregnancy even though they actually did. Another possible reason for that change is that some pregnant women with higher education level may try to stop smoking after they realize that they are in pregnant. But because smoking is an addictive behavior, the quitting action may fail latter. Hypothesis 3-A (Women with higher education level and with more family income would smoke fewer cigarettes during their pregnancy.) is significantly supported by the findings listed in table five. However, when the subsample was replaced with the women who were currently pregnant, the two predictors, education and income, become not significant. One possible reason of the difference is the error of recall. For the women condoning to smoke after learning of their pregnancy, education and income may have no impact on how many cigarettes they smoked. But the women who were with higher education level and with more family income may tend to recall that they smoked fewer cigarettes during their past pregnancy. Hypothesis 2-B predicts that when the other variables are equal, pregnant women who are advised to quit smoking by doctors are more likely to stop smoking than those who are not. Hypothesis 3-B predicts that when the other variables are equal, pregnant women who are advised to quit 35 smoking by doctors would smoke fewer cigarettes during pregnancy than those who are not advised to quit smoking by doctors. Both of the two hypotheses are not supported by the results. According to the results, when the other variables are equal, pregnant women who are advised to quit smoking by doctors are significantly more likely to continue smoking and would smoke more cigarettes than those who are not. It seems that many pregnant women do not comply with the quitting advice from doctor. The finding reveals that doctors' quitting advice could not make pregnant women take action to quit. There are some possible reasons for that. For example, the communication patterns between doctors and patients are bad; doctors' advice did not give pregnant women enough motivation to quit; or besides the advice from doctors, pregnant women need other supports, such as quitting training programs, supports from farme and fiiends...etc., to help them quit, or doctors are likely to advise heavy smokers who are less likely to quit. Hypothesis 2-C states that when the other variables are equal, the more cigarettes a pregnant woman used to smoke per day before pregnancy, the higher possibility she will continue to smoke during her pregnancy. Hypothesis 3-C states that when the other variables are equal, pregnant 36 women who used to smoke more cigarettes per day before learning of pregnancy would smoke more cigarettes than women who used to smoke fewer cigarettes per day. Both hypotheses are supported by findings. However, interestingly, if the subsamaple for testing hypothesis 2-C is replaced with the women who were currently pregnant, the impact of amount smoked per day before pregnancy would become not significant (please refer to Table A in Footnote 2). One possible reason is that amount smoked per day before pregnancy hasn't efi‘ect on continuing to smoke in the early period of pregnancy but it does in late period. CONCLUSION Generally, according to the findings, socioeconomic status is not an important factor to predict whether a pregnant woman receive doctor' 8 advice to stop smoking or not. The most important factor to predict whether a pregnant woman receive advice to quit is the amount she smokes per day before pregnancy. Moreover, socioeconomic status and doctor' 5 advice to quit smoking haven't strong impact on pregnant women's smoking behavior. The most important factor to predict pregnant women's smoking behavior 37 after learning they are pregnant is how many cigarettes they smoke per day before learning they are pregnant. APPENDIX APPENDIX THE RESULTS REMOVED WEIGHTS The following tables present the results of the analyses removed weights. The conclusion of these tables would be the same as those weighted by the modified weights. 38 39 Table A-1: Zero-order Correlations of Education, Income, Amount Smoked Per Day Before Pregnancy, Smoked During Pregnancy, Amount Smoked afier Learning of Pregnancy, and Receiving Advice to Stop Smoking Education Income Amount Education Income 3852"“ Amount -.0274 .0168 Smoked Per Day Before Pregnancy Smoked During Pregnancy -.1272"" -.1129“‘ 3278“”I Amount Smoked afier Learning lfieenansy Receiving Advice to Stop Smoking -.l355““ -.0874'""' 6011*" -.0107 .0207 2151“" Smoked Smoked Per During Day Before Pregnancy Pregnancy Amount Receiving Smoked alter Advice to Learning Stop Pregnancy Smoking_ .5724m .2865‘" 2033“" Note: 1. Sample Size=2755 2. Compare with Table 2 “ P<0.01 "* P<0.001 40 Table A-2: Effects and Significance of Amount Smoked per Day Before Pregnancy, Education and Income on Receiving Advice to Stop Smoking Variables Std. Partial Partial Correlation Sig T Regression Coefiicients Coefficients Amount Smoked .2093 .2107 .0000 Per Day Before Presmncy Education -.0702 -.0654 .0013 Income -.0717 -.0668 .0010 Notezl. Receiving Advice to StOp Smoking is coded 1, and not receiving it is coded 0. 2. Multiple Regression Analysis ' 3. Sample Size=2755 4. Compare with Table 3 Table A-3: Effects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy Variables Std. Partial Partial Correlation Sig T Regression Coemcients 'Coeflicients Education -.0845 -.0853 .0012 Income -.0980 -.0988 .0002 Advice .2352 .2454 .0000 Amount Smoked .2668 .2758 .0000 Per Day Before Pregnancy Note: 1. Continuing Smoking During Most Pregnancy is coded 1. Stop Smoking During Most Pregnancy is coded 0. 2. Multiple regression analysis 3. Sample size=l645 4. Compare with Table 4 41 Table A-4 : Efi‘ects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked after Learning of Pregnancy Variables Std. Partial Partial Correlation Sig T Regression Coeficients Coemcients Education -.0961 -.1 124 .0000 Income -.0610 -.0716 .0000 Advice .0778 .0963 .0004 Amount Smoked .5832 .5872 .0000 Per Day Before .Iirsamcy Note: 1. Amount Smoked alter Learning of Pregnancy is coded by the numbers of cigarettes pregnant women smoked after learning of pregnancy (0 to 95) 2. Multiple Regression Analysis 3. Sample Size=1500 4. Compare with Table 5 42 Table A-5: Effects and Significance of Education, Income, Advice, and Amount Smoked Per Day Before Pregnancy on Continuing Smoking During Pregnancy Variables Std. Partial Partial Correlation Sig T Regression Coefiicients Coeficients Education -.2752 -.2795 .0000 Income -.1739 -.1833 .0014 Advice .3171 .3377 .0000 Amount Smoked -.0215 -.0244 .6733 Per Day Before We Note: 1. Continuing Smoking During Pregnancy is coded as 1. Stop Smoking During Pregnancy is coded as 0. 2. Total cases=358(100%). The number of cases coded as 0 = 148(41%). The number of cases coded as 1=210(59%) 3. Multiple Regression Analysis 4. Compare with Table A 43 Table A-6 :Efi‘ects and Significance of Education, Income, Receiving Advice to Stop Smoking and Amount Smoked Per Day Before Pregnancy on Amount Smoked after Learning of Pregnancy (compare with Table B) Variables Std. Partial Partial Correlation Sig T Regression Coeficients Coefficients Education -.0430 -.0570 . .4502 Income -.0486 -.0656 .3847 Advice .1379 .1884 .0118 Amount Smoked .6568 .6676 .0000 Per Day Before We Note: 1.Amount Smoked afier Leaming of Pregnancy is coded by the numbers of cigarettes pregnant women smoked after learning of pregnancy (0 to 95) 2.Total cases=209(100%). 1-14 cigarettes=72(34%); 15-24 cigarettes=86 (41%); and 25-60 cigarettes= 51(24%) ; more than 60 cigarettes=0(0%) 3.Multiple Regression Analysis 4.Compare with Table B REFERENCES REFERENCES Adler, N. E., Boyce, T., Chesney, M. A., Folkrnan, S., & Syrne, S. L. 1993. Socioeconomic Inequalities in Health. JAMA, 269(24): 3140-3145 Adult Use of T obacoo, 19 75. US. Dept of Health, Education, and Welfare, National Clearinghouse for Smoking and Health, 1976. Anda, RF, Remington P.L., Sienko, D.G., Davis, RM. 1987. Are Physicians Advising Smokers to Quit: the Patients' Perspective. Journal of American Medical Association, 257: 1916-1919. Becker, Marshall H. 1974. The Health Belief Model and Sick Role Behavior. Marshall H. Becker (Eds) The Health Belief Model and Personal Health Behavior. p.82-92 Bonnar J, Goldberg A, Smith J A 1969. Do Pregnant Women Take Their Iron? Lancet, 1 :45 1-45 8 Brotherston, R. 1976. Inequality: Is It Inevitable? C. 0. Carter and John Peel. (Eds). Equalities and Inequalities in Health London: Academic Press. pp 73-104 Brownson, R.C., Jackson, T.J., Wilkerson, J .C., Davis, J .R., Owens, N.W., & Fisher, E.B., Jr. 1992. Demographic and Socioeconomic Differences in Beliefs about the Health Efi‘ects of Smoking. American Journal of Public Health, 82(1): 99-103 Chamey E, et al. 1967. How Well Do Patients Take Oral Penicillin? A Collaborative Study in Private Pratice. Pediatrics, 40: 188-195. Cherry, N. & Kiemen, K. 1976. Personality Scores and Smoking Behavior: a longitudinal study. British Journal of Preventive and Social Medicine, 30: 123-131 Cnattingius, S., Haglund, B., & Meirik, O. 1988. Cigarette Smoking as a Risk Factor for Late Fetal and Early Neonatal Death. British Medical Journal, 297: 25 8-261. 44 45 Coambs, R.B., & Kozlowski, LT. 1989. The Future of Tobacco Use and Smoking Research. Tara Ney and Anthony Gale (Eds). Smoking and Human Behavior. New York: John Wiley & Sons. pp. 337-348 Cobum, D., Pope, C. R 1974. Socioeconomic Status and Preventive Health Behavior. Journal of Health and Social Behavior, 15: 67-77 Davis, MS. 1968. Physiologic and Demographic Factors in Patient Compliance with Doctors' Orders. 1968. Med Care, 6: 1 15-122. Davis, MS. 1968. Variations in Patients' Compliance with Doctors' Advice: An Empirical Analysis of Patterns of Communication. American Journal of Public Health, 582274-288 ’ Duncan, 0D. 1961. A Socioeconomic Index for All Occupations. Reiss, Al, Duncan O.D., Hatt, P.K., North,C.C. (Eds). Occupation and Social Status. New York: The Free Press of Glencoe. pp. 109-138 " Dufl', R. & Hollingshead, A. 1968. Sic/mess and Society. New York: Harper and Row. Epidemiology Br., Ofice of Smoking and Health, National Center for Chronic Disease Prevention and Health Prevention; Div of Health Interview Statistics, National Center for Health Statistics, CDC. 1993. Physician and Other Health Care Professional Counseling of Smokers to Quit-United States, 1991. The Morbidity and Mortality Weekly Report, 42: 854-857 Francis, V, Korsch, BM, Morris, MJ. 1969. Gaps in Doctor-Patient Communication: Patients’ Response to Medical Advice. N Engl J Med 280:535-540 Frank, E., Winkleby, M.A., Altman, D.G., Rockhill, B., Fortmann, SP. 1991. Predictors of Physicians’ Smoking Cessation Advice. Journal of American Medical Association, 266(22):3 139-3 144 Fuchs, V.R 1979. Economics, Health and Post-Industrial Society. Milbank Memorial Fund Quarterly Health and Society, 57(2): 153-182 Glynn, T.J., Manley, M.W., Pechacek, T.F. 1990. Physicians-Initiated Smoking Cessation Program: the National Cancer Institute Trials. Engstrom P.F., Rimer B., Mortenson L.E. (Eds) Advances in Cancer Control. New York, NY: John Wiley & Sons Inc. pp. 11-26. Green, L.W. 1970. Manual for Scoring Socioeconomic Status for Research on Health Behavior. Public Health Rep., 85: 815-827 46 Green, L.W. 1971. Status Identity and Preventive Health Behavior. Pacific Health Echacation Reports, 1. University of California, Berkeley, and University of Hawaii, Honolulu Grossman, M. 1976. The Correlation Between Health and Schooling. Nestor E. Terleckyj (Eds). Household Production and Consumption, Studies in Income and Wealth, Vol.4. DC: National Bureau of Economics Research. pp. 147-211 Hafl‘erty, F.W., Salloway, J.C. 1994. In Search of the Physician-Patient Interaction in Models of Health Behavior Change. Gary L. Albrecht (Eds). Advances in Medical Sociology: A Reconsideration of Health Behavior Charge Models, Volume 4. London England: JAI Press Inc. pp. 1 13-139. Haglund, B., & Cnattingius, S. 1990. Cigarette Smoking as a Risk Factor for Sudden Infant Death Syndrome: A Population Based Study. American Journal of Public Health, 80: 29-32. Hollingshead, AB. 1971. The Indiscriminate State of Social Class Measurement, Social Forces, 49: 563-567 Johannsen, WJ, Hellmuth, GA, Sorauf T. 1966. On Accepting Medical Recommendations. Arch Envir Health, 12:63-69 John. E. M., Savitz, D. A., & Sandler, D. P. 1991. Prenatal Exposure to Parents’ Smoking and Childhood Cancer. American: Journal of Epidemiology, 133: 123-132. Kitagawa, E. M. & Hauser, P. M. 1973. Difi'erential Mortality in the United States: A Satay in Socioeconomic Epidemiology. Cambridge: Harvard University Press. Kleinmann, Jc., Pierre, M.B., Jr., Mardans, J.H., Land, G.H., & Schramm, W.F. 1988. The Efl‘ects of Maternal Smoking on Fetal and Infant Mortality. American Journal of Epidemiology, 127: 274-281 . Korsch, BM, Gozzi, EK, Francis, V. 1968. Gaps in Doctor-Patient Communicationzl. Doctor-patient Interaction and Patient Satisfication. Pediatrics, 42:855-871. Kozlowski, LT. 1979. Psychological Influences on Cigarette Smoking. Smoking and Health. A Report of the Surgeon General. Washington, DC: US DHEW, Publication No. (PHS) 79-50066 Lefcowitz, M. J. 1973. Poverty and Health: A lie-Examination. Inquiry, 10(1): 3-13 47 Liberatos, P., Link, B., and Kelsey, J. 1988. The Measurement of Social Class in Epidemiology. Epidemiologic Review, 10: 87-121 Lumley, J ., & Astbury, J. 1989. Eflective Care in Pregnancy and Chilfiirth. I. Chalmers, M. Enkin, & M.J.N.C. Keirse (Eds). Oxford: Oxford University Press. pp. 237-254. MacDonald, ME, Hagberg, KL, Grossman, BJ. 1963. Social Factors in Relation and Participation in Follow-up Care of Rheumatic Fever. Journal of Pediatrics, 62:503- 5 13 Matthews, K., Kelsey, S., Meilahn, E., Kuller, L., & Wing, R. 1989. Educational Attainment and Behavioral and Biologic Risk Factors for Coronary Heart Disease in Middle-Aed Women. American Journal of Epidemiology, 129: 1132-1144 McKinlay, J .B. 1975. The Help-Seeking Behavior of the Poor. John Kosa, Aaron Antonovsky, & Irving Kenneth Zola (Eds). Poverty and Health: A Sociological Analysis. Cambridge: Harvard University Press Mueller, C.W., Parcel, TL. 1981. Measures of Socioeconomic Status: Alternatives and Recommendations. Child Development, 52: 13-30 Naeye, R. L., & Peters, E. C., 1984. Mental Development of Children Whose Mothers Smoked during Pregnancy. Obstetrics and Gynecology, 64: 601-607. National Health Interview Survey, 1990: Health Promotion and Disease Prevention (HPDP) Pregnancy and Smoking Supplement [Computer File]. Hyattsville, MD: US. Sept. of Health Statistics [Producer], 1990. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 1993. Nersesian, W. S. 1988. Infant Mortality in Socially Vulnerable Populations. Annual Review of Public Health, 92316-3 77 Ockene, J .K., Kristeller, J ., Goldberg, R. et al. 1991. Increasing the Efficacy of Physician- Delivered Smoking Interventions: a randomized clinical trial. J Gen Intern Med 6: l- 8 Peen, A., Berasio, P, Nesheirn, B.J.,Ullern, A.M., Heggelund, B.W., & Matheson, J. 1991. Characterization of Birth Populations in Two Norwegian Counties (Akershus and Hordaland). T idss'krifl for den Norslte La: geforening nr. 13, 111:1613-1616 Pragofl‘, H. 1962. Adjustment of Tuberculosis Patients One Year After Hospital Discharge. Public Health Rep, 77:671-679 48 Rosenstock, I. M. 1975. Prevention of Illness and Maintenance of Health. John Kosa, Aaron Antonovsky, and Irving K. Zola (Eds). Poverty and Health. Cambridge: Harvard University Press. pp. 193-223 Rundall, T. G., & Wheeler, John RC. 1979. The Effect of Income on Use of Preventive Care: An Evaluation of Alternative Explanations. Journal of Health and Social Behavior, 20: 397-406 Smoking and Health: A National Status Report. 1990. 2nd Edition. U.S. Departth of Health and Human Service, Ofice on Smoking and Health. DI-IHS Publication No. (CDC) 87-8396 (Revised 02-90). Smoking and Health: A Report of the Surgeon, Genera. 1979. U. S. Dept. of Health, Education, and Welfare Publication (PHS) 79-50066. Stein, 2., Kline, J ., Levin, B., Susser, M., & Warburton, D. 1980. Epidemiologic Studies of Environmental Exposures. G.G. Berg & D.D. Maille (Eds). Measurements of Risks pp. 170. Suchman, EA. 1965. Social Patterns of Illness and Medical Care. Journal of Health and Human Behavior, 6:2-16 Wan, T. T., & Gray, L. C. 1978. Differential Access to Preventive Services for Young Children in Low-Income Urban Areas. Journal of Health and Social Behavior, 19: 3 12-3 24 Williams, DR 1990. Socioeconomic Differentials in Health: A Review and Redirection, Social Psychology Quarterly, 53(2):81-89 Winkleby, M.A., Jatulis, D.E., Frank, E., Fortmann, SP. 1992. Socioeconomic Status and Health: How Education, Income, and Occupation Contribute to Risk Factors for Cardiovascular Disease. American Journal of Public Health, 82(6): 816-820 HICHIGRN STR [gr] "III! nuliruiiliflgtjyifllli“ 2930140