Psychosocial stressors among depressed adolescent mothers.

Prodromidis, Margarita ; Abrams, Sonya ; Field, Tiffany 等

Depression is a major problem for adolescent mothers. In a recent study on postpartum depression, mother's age was necessarily covaried as a result of the disproportionate number of adolescent mothers in a sample of randomly recruited depressed mothers (Field et al., 1988). Early sexual activity, pregnancy, childbirth, and motherhood exemplify stressful situations that may trigger or exacerbate depression. Limited educational and financial resources, single parenthood, and lack of social support networks may also relate to adolescent depression. In turn, the infants of adolescent mothers are more likely to have cognitive, emotional, and physical problems (Baldwin & Cain, 1980; Field, Widmayer, Stringer, & Ignatoff, 1980; Furstenburg, Brooks-Gunn, & Chase-Lansdale, 1989). Further the various combinations of psychosocial stressors may predispose some adolescents to poorer functioning than others (Donovan & Jessor, 1985; Ensminger, 1990).

The problem of identifying depressed adolescent mothers for early intervention underscores the importance of using screening instruments to determine the particular profile of problems experienced.

Assessing adolescent psychosocial stressors requires a measure that takes into account several of life's activities. To aid in the assessment and referral of adolescent problem behavior, the National Institute on Drug Abuse developed a questionnaire called The Problem Oriented Screening Instrument for Teenagers (POSIT) (1987). The POSIT is comprised of 139 items addressing various problem areas. The items are summarized by a total score and grouped into 10 target problem areas relevant to adolescent life experiences: substance use/abuse, physical health, mental health, family relations, peer relations, educational status, vocational status, social skills, leisure and recreation, and aggressive behavior/delinquency. This instrument was selected because of its comprehensive nature and because depressed adolescent mothers in particular have problems in these areas. For example, the use of drugs and alcohol is problematic because depressed adolescents (including those who are pregnant) may use them as self-medication for their depression (Adams & Adams, 1991; Zuckerman, Amaro, Bauchner, & Cabral, 1989). Depression also contributes to problem behaviors such as drinking, early sexual activity, and subsequent pregnancy.

High intercorrelations have been found between alcohol and marijuana use and sexual intercourse among adolescents (Jessor & Jessor, 1977). In another study, juvenile offenses involving formal contact with the authorities were correlated with pregnancy; in addition, offenders were likely to mate with other offenders (Elster, Lamb, Peters, Kahn, & Tavare, 1987).

In a recent study, Ensminger (1990) assessed four patterns of behavior among adolescent males and females (N = 705): no problem behaviors; sexual activity; substance abuse; and assaults. The findings showed that behavior problems were intercorrelated. In fact, there were no cases in which adolescents were involved in assaultive behavior and substance abuse, but not sexual activity. There were, however, cases in which sexual activity occurred alone. For females classified as having sexual activity only, sexual activity was associated with the girls having been parented by adolescent mothers with little education. Ensminger (1990) cautioned that a variety of factors were not considered in the study which are likely to contribute to early sexual activity and pregnancy, such as disturbed family relations and social isolation.

Family structure and social support problems are the two most commonly addressed psychosocial stressors related to adolescent pregnancy. Of these, level of parental education, socioeconomic status, and family constellation are among the most frequently examined. Adolescent mothers are more likely to come from disadvantaged socioeconomic backgrounds (Barnett, Papini, & Gbur, 1991; Haggstrom, Kanouse, & Morrison, 1987; Hofferth, 1984); further, families of adolescent mothers are also less likely to encourage educational or vocational goals.

Unfortunately, only a few studies have considered the effects of the quality of family relations on pregnant adolescent mothers, particularly communication between adolescents and their parents which affects adolescent sexual behavior. For example, adolescents who talk openly to their parents about sexual intercourse and contraception are more likely to delay intercourse and use contraception (Furstenburg, 1971; Lewis, 1978; Shah, Zelnik, & Kantner, 1973). The quality of family relations may also affect adolescent mothers' psychological well-being. Open communication about the pregnancy, labor, and childrearing issues may alleviate stress.

Studies indicate that social support affects adolescent mothers' behavior with their infants (Colletta & Gregg, 1981; Crnic, Greenberg, & Slough, 1986; Levine, Garcia, & Oh, 1985). Adolescent mothers who perceived having had adequate social support during their pregnancy were more likely to have fewer prenatal complications, more positive childcare commitments, and higher self-esteem (Dunst, Vance, & Cooper, 1986). Although social support can extend beyond the family to friends and professionals in the community, adolescents are often ashamed of their pregnancy, lose contact with friends (Bolton, 1980) and frequently are unaware of the professional support systems that are available to them (Crockenberg, 1986).

The present study attempted to identify the psychosocial stressors that affect depressed adolescent mothers. The sample consisted of adolescent mothers classified as depressed and nondepressed postpartum. The question of whether depressed mothers were more likely to have experienced more behavior problems and psychosocial stressors than did nondepressed mothers was addressed. The specific aims of the study were to determine whether there were significant differences on the POSIT scores for mothers who were identified as depressed or nondepressed, which of the POSIT subscales best predicted depression, and whether other measures of stress and mental health were associated with the POSIT.

METHOD

Sample

Adolescent mothers (N = 154) were interviewed at a hospital maternity unit between one and three days postpartum. The selection criteria were: (a) age between 14 and 21 years, and (b) a Beck Depression Inventory (BDI: Beck, Rush, Shaw, & Emery, 1979) score between 3 and 9 (nondepressed, N = 49) or greater than 13 (depressed, N = 105). The sample was representative of the delivery population at the hospital which consists predominantly of single, black, and Hispanic mothers of lower socioeconomic status who averaged a tenth grade education. Mothers and infants received several gifts for their participation (e.g., rattles, pacifiers, pictures of their infant, infant books).

Table 1

BDI, PBF, and Background demographic variables for depressed

and nondepressed mothers

Variables Depressed Nondepressed

 (N=105) (N= 49)

 Mean Mean pl

BDI 18.0 (5.0) 6.4 (2.4) .00

PBF 7.6 (2.4) 5.8 (3.2) .00

Age 17.7 (1.9) 18.2 (1.9) NS

Education 10.0 (1.4) 10.6 (1.4) .01

Socioeconomic status 4.4 (.8) 4.4 (.7) NS

Marital status (%): S 75.4 88.2 NS

 M 24.4 11.5

Race (%): Caucasian 25.0 15.0 NS

Black 50.2 70.0

Hispanic 25.8 15.0

Number of children %: 1 63.5 72.0 NS

 2 30.5 24.0

 3 5.0 3.0

 4 1.0 1.0

Substance use (%)

Cigarettes Yes 5.3 2.0 NS

 No 95.4 98.0

Alcohol: Yes 12.0 14.0 NS

 No 88.0 86.0

Marijuana: Yes 3.0 2.0 NS

 No 97.0 98.0

Caffeine: Yes 92.0 50.0 .01

 No 8.0 50.0

Note: Standard deviations are indicated in parentheses.

1 p values are derived from t-test analyses for BDI, PBF, age, education and
socioeconomic background, and p values for the remaining variables are derived
from chi-square analyses

2 Hollingshead SES -- lower scores are optimal (range = 1 to 5)

Measures

Because of potential reading-level differences among the mothers, questionnaires were administered in an interview format. In addition to the Problem Oriented Screening Instrument for Teenagers (POSIT), the primary measure of this study, additional questionnaires included a drug history, the Beck Depression Inventory, the Paranoid-Border-line Features Scale of the Millon Clinical Multiaxial Inventory, and the Maternal Stress Interview. They were given in the order in which they are described below.

Drug/alcohol history. Mothers were asked to identify substances used prior to or during pregnancy. The list included caffeine, cigarettes, alcohol (beer, wine, liquor), nonprescription medications, prescription medications (valium, codeine), and illicit drugs (marijuana, cocaine, heroin, methadone). Mothers were subsequently asked to estimate the date on which they last used any of these substances and the frequency (daily, weekly) over the past year and throughout the pregnancy.

Beck Depression Inventory (BDI: Beck, Rush, Shaw, & Emery, 1979). This 21-item questionnaire uses a 4-point scale (0 to 3) to indicate the presence or absence and severity of depressed feelings, behaviors, and symptoms. The inventory has been frequently used for assessing non-clinically depressed populations and has good construct validity (Coyne & Gotlib, 1983). Based on the level of severity of depression, mothers in the present study were assigned to either a "nondepressed" group (BDI scores between 3 and 9), or a "depressed" group (BDI scores of 13 and above). Total scores on the BDI range from 0 to 60. Mothers scoring between 0 and 2 on the BDI were omitted since previous research suggested that they may be more depressed than high BDI mothers and may be "denying" their depressed feelings (Field et al., 1991).

Paranoid-Borderline Features (PBF) (Millon, 1982). To determine the presence of paranoid and borderline features, 15 questions from a subset of 75 from the Millon Clinical Multiaxial Inventory (MCMI) were administered (Millon, 1982). The questions were read to the subjects who responded by indicating "true" or "false." The total score can range from 0 to 15, with higher scores representing greater paranoid and borderline thoughts.

Maternal Stress Interview (Field, 1984). This interview examines various demographic and life experiences. It consists of 42 questions that provide a total score and 5 composite scores including Socioeconomic Status, Social Support, Child Rearing, Own Childhood, Stress, and Crowding.

The Problem Oriented Screening Instrument for Teenagers (POSIT: National Institute on Drug Abuse, 1987) was administered to screen for social, emotional, and behavioral problems. The 139-item questionnaire requires "yes" or "no" responses and measures various social, familial, physical, and mental stressors. The items have been grouped into 10 subscales (functional areas) in order to identify specific problem areas: Substance Use/Abuse, Physical Health Status, Mental Health Status, Family Relations, Peer Relations, Educational Status, Vocational Status, Social Skills, Leisure and Recreation, and Aggressive Behavior/Delinquency.

Clinical cut-off points are given for each of the functional areas. In addition, "flagged" items alone indicate the need for further assessment. The POSIT reportedly is a good measure for discriminating between groups of adolescents known to have problems (in-treatment) and groups of adolescents without problems (NIDA: National Institute on Drug Abuse, 1987). Subjects scoring above the clinical cut-off or having a red flag are administered a series of additional scales for more extensive assessment of the problem areas (CAB: Comprehensive Assessment Battery).

RESULTS

Group Differences on the POSIT

Chi-square analyses and t tests revealed no significant differences between the two groups (depressed versus nondepressed) on any of the sociodemographic variables except for level of education and caffeine use. Depressed mothers had half a year less education than did nondepressed mothers, probably explained by the nondepressed mothers being almost 6 months older. Mothers who were depressed were more likely to have used caffeine throughout their pregnancy ([[Chi].sup.2] = 8.2, df = 1, p = .005).

Multivariate analysis of variance (MANOVA) was performed on the 10 POSIT subscales by group (depressed, nondepressed), yielding a significant main effect of group, F(10, 142) = 6.56, p [is less than] .001. Univariate follow-up tests were conducted on each POSIT subscale. As shown in Table 2, depressed mothers had higher mean scores for several POSIT categories. Depressed mothers received higher (less optimal) total scores than did nondepressed mothers, t(152) = 5.13, p [is less than] .001. In addition, several subscales of the POSIT significantly differentiated the depressed from the nondepressed mothers. Depressed mothers had higher scores on Physical Health, t(152) = 3.60, p [is less than] .001; Mental Health t(152) = 5.67, p [is less than] .001; Family Relations t(152) = 4.22, p [is less than] .001; Peer Relations t(152) = 3.35, p [is less than] .05; Educational Status t(152) = 4.71, p [is less than] .001, Social Skills t(152) = 4.32, p [is less than] .001, and Aggressive Behavior/Delinquency t(152) = 2.05, p [is less than] .05.

POSIT Subscales as Predictors of Maternal Depression

Stepwise multiple regression analysis was performed in order to ascertain the relative contribution of each of the POSIT subscales to maternal depression. The regression analysis yielded a three-variable model, including Mental Health, Family Relations, and Social Skills, accounting for 25% of the variance.

A discriminant function analysis was then conducted to determine how well the POSIT classified depressed and nondepressed mothers. Based on the discriminant function analysis using the POSIT subscales as predictors, 70% of the subjects were correctly classified. (The chance rate of correct classification in a two-category system would, of course, be 50%.) The best classification rate was found for the nondepressed group; 78% of those mothers were correctly classified. The analysis generated one discriminant function [[Chi].sup.2](4, N = 152) = 43.43, p [is less than] .001, which maximally separated the depressed from the nondepressed group. Variables contributing to this dimension included Mental Health, Family Relations, Social Skills, and Physical Health. This function loaded most heavily on Mental Health.

Table 2

Means and standard deviations for depressed and nondepressed mothers on
categories of the POSIT

 Group

Measures Depressed Nondepressed
 (N=105) (N=49)
 Mean SD Mean SD p

POSIT Total Score 46.7 (14.1) 32.3 (13.5) .001
Substance Use .1 (.4) .1 (.4) NS
Physical Health 4.2 (1.5) 3.3 (1.4) .001
Mental Health 8.7 (4.4) 4.7 (3.4) .001
Family Relations 4.0 (2.4) 1.4 (1.2) .001
Peer Relations 2.9 (1.8) 1.9 (1.5) .001
Educational Status 8.5 (3.9) 5.5 (3.0) .001
Vocational Status 5.2 (2.6) 4.4 (2.9) .NS
Social Skills 3.4 (1.6) 2.2 (1.5) .001
Leisure Activity 4.7 (1.8) 4.3 (1.8) NS
Aggressive Behavior 4.2 (2.6) 3.3 (2.8) .05

Notes: p levels are based on t-tests group comparisons, and lower POSIT scores
are optimal
Table 3

Stepwise regression analyses for predicting depression and classification

Variable Multiple R R2 R2 change F P

Mental Health .42 .17 .17 32.17 .000
Family Relations .47 .22 .05 22.70 .002
Social Skills .50 .25 .03 16.83 .030

Classification summary

Actual Group N Predicted Group
 1 2

1 Depressed 105 77 (67%) 38 (33%)
2 Nondepressed 49 11 (22%) 40 (78%)

Relationships between POSIT Subscales and other Measures of Psychosocial Functioning

As was noted in Table 1, the groups differed on the BDI and the PBF. As expected, depressed mothers had higher BDI scores (M = 18.0) than did nondepressed mothers (M = 6.4). In addition, depressed mothers scored significantly higher on the PBF.

The functional areas of the POSIT were highly correlated with each other and with the total score. Among the highest correlations were Social Skills and Leisure Activity (r = .73, p [is less than] .01), Mental Health and Educational Status (r = .75, p [is less than] .01), Leisure Activity and Aggressive/Delinquent Behavior (r = .66, p [is less than] .01), Mental Health and Physical Health problems (r = .53, p [is less than] .01), Mental Health and Aggressive/Delinquent Behavior (r = .56, p [is less than] .01), and Mental Health and Peer Relations problems (r = .54, p [is less than] .01). Peer Relations were also correlated with Educational Status (r = .69, p [is less than] .01) and Aggressive/Delinquent Behavior (r = .64, p [is less than] .01).

The correlations between the POSIT total score and the 8 subscales with the BDI, PBF, and Maternal Stress Interview are presented in Table 5. The POSIT total score, Physical Health, Mental Health, Educational Status, and Vocational Status categories were each significantly correlated with the BDI and PBF. Depressed mothers functioned more poorly on these POSIT categories as did mothers with high scores on the PBF. Several of the POSIT scores were also correlated with the Maternal Stress Interview total score particularly the own childhood stress subscale.

DISCUSSION

In the present study depressed adolescent mothers experienced more psychosocial stressors than did nondepressed mothers. The psychosocial stressors were highly related to each other, suggesting that several factors are likely to affect the well-being of adolescent mothers. Studies of adolescent mothers have reported increased risk of psychosocial problems associated with early pregnancy (Abrahams, Morrison, & Waite, 1988; Jessor, Chase, & Donovan, 1980, Miller, & Simon, 1974). In light of the association between adolescent mothering and psychological problems, several researchers have attempted to delineate the factors that place adolescent mothers at risk for poor adjustment. Many of these studies have found a relationship between sociodemographic variables (race and socioeconomic status) and adjustment (Mott & Heurin, 1988; Jessor & Jessor, 1975). The importance of such findings underscores the need for closer assessments of specific populations.

TABULAR DATA OMITTED

TABULAR DATA OMITTED

By targeting a specific population of adolescent mothers who are similar in socioeconomic status, education, and age, we were able to address other factors (i.e., family relations, peer relations, and social skills) that are likely to be equally important as sociodemographic factors. Moreover, by isolating this group we were able to provide a better understanding of the stressors associated with a specific population. As noted by Ensminger (1990), this is clearly an essential task since researchers often cite differences between social classes, but rarely investigate the differences within social classes. Thus, one of the caveats in interpreting these results is that the relationships found between psychosocial stressors in this sample may not necessarily be as strong (or weak in some cases) as they may be in other samples of adolescent mothers (middle or upper-middle class) because of such factors as restriction of range.

Unlike other studies, mothers were classified as depressed and nondepressed, making it possible to determine which factors place some mothers at greater risk for maladjustment as well as those which enhance adjustment for some adolescent mothers. That depressed adolescent mothers reported poorer functioning in the areas of physical health, mental health, peer relations, family relations, vocational status, and social skills clearly indicates the need for continued efforts in the screening of qualitative factors that affect them. Perceived lack of social skills and poor peer and family relations tend to isolate young pregnant mothers. Such isolation during pregnancy is likely to negatively affect mental health. Although the well-being of adolescent mothers appears to be affected by a variety of psychosocial stressors, the stage of adolescence itself is accompanied by a variety of social pressures, such as establishing autonomy, friendship and familial relations, and occupational goals (Erickson, 1968). This highlights the importance of accounting for the "typical" stressors that specifically affect adolescents who are also young mothers.

One of the major problems in studying adolescent stressors is in achieving an adequate assessment. Fortunately, the screening tool used here permitted a comprehensive assessment. The POSIT was a reliable screening instrument that aided in the examination of the diverse psychosocial factors that affect adolescent mothers' functioning. By identifying specific problems on the POSIT, individualized treatment plans can be designed for the adolescents.

REFERENCES

Adams, M., & Adam, J. (1991). Life events, depression, and perceived problem-solving alternatives in adolescents. Journal of Child Psychology and Psychiatry, 32, 811-820.

Baldwin, W. H., & Cain, V. S. (1980). The children of teenage parents. Family Planning Perspectives, 12, 34-43.

Barnett, J. K., Papini, D. R., & Gbur, E. (1991). Familial correlates of sexually active pregnant and nonpregnant adolescents. Adolescence, 26, 457-472.

Beck, A. T., Rush A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford Press.

Bolton, F. G. (1980). The pregnant adolescent: Problems of premature parenthood. Beverly Hills, CA: Sage.

Colleta, N. D., & Gregg, C. H. (1981). Adolescent mothers' vulnerability to stress. Journal of Nervous and Mental Disease, 169, 50-54.

Coyne, J. C., & Gotlib, I. H. (1983). The role of cognition in depression: A critical appraisal. Psychological Bulletin, 94, 472-505.

Crnic, K. A., Greenberg, M. T., & Slough, N. M. (1986). Early stress and social support influences on mothers and premature and full-term infants. Infant Mental Health Journal, 7, 19-33.

Crockenberg, S. (1986). Professional support for adolescent mothers: Who gives it, how adolescent mothers evaluate it, and what they would prefer. Infant Mental Health Journal, 7, 49-58.

Donovan, J. E., & Jessor, R. L. (1985). Structure of problem behavior in adolescence and young adulthood. Journal of Consulting and Clinical Psychology, 53, 890-904.

Dunst, C. J., Vance, S. D., Cooper, C. S. (1986). A social systems perspective of adolescent pregnancy: Determinants of parent and parent-child behavior. Infant Mental Health Journal, 7, 34-48.

Elster, A. B., Lamb, M. E., Peters, L., Kahn, J., & Tavare, J. (1987). Judicial involvement and conduct problems of fathers of infants born to adolescent mothers. Pediatrics. 79, 932-936.

Ensminger, M. E. (1990). Sexual activity and problem behaviors among black, urban adolescents. Child Development, 61, 2032-2046.

Erikson, E. (1968). Identity, youth and crisis. New York: Norton.

Field, T., Healy, B., Goldstein, S., Perry, D., Bendell, D., Schanberg, S., Zimmerman, S., & Kuhn, C. (1988). Infants of depressed mothers show "depressed" behavior even with nondepressed infants. Child Development, 59, 1569-1579.

Field, T., Morrow, C., Healy, B., Foster, T., Adelstein, D., Goldstein, S. (1991). Mothers with zero Beck depression scores act more depressed with their infants. Development and Psychopathology, 3, 253-262.

Field, T., Widmayer, S. M., Stringer, S., & Ignatoff, E. (1980). Teenage, lower-class, black mothers and their preterm infants: An intervention and developmental follow-up. Child Development, 51, 426-436.

Furstenberg, F. F. (1971). Birth control experience among pregnant adolescents: The process of unplanned parenthood. Social Problems, 19, 199-203.

Furstenberg, F. F. Jr., Brooks-Gunn, J., & Chase-Lansdale, L. (1989). Teenage pregnancy and childbearing. American Psychologist, 44, 313-320.

Haggstrom, G. W., Kanouse, D. E., & Morrison, P. A. (1987). Accounting for the educational shortfalls of mothers. Journal of Marriage and the Family, 48, 175-186.

Hofferth, S. (1984). A comment on the "social determinants of age at first birth." Journal of Marriage and the Family, 46, 7-8.

Jessor, R. L., & Jessor, S. L. (1977). Problem behavior and psychological development: A longitudinal study of youth. New York: Academic Press.

Lewis, R. A, (1973). Parents and peers: Socialization agents in the coital behavior of young adults. Journal of Sex Research. 9, 156-170.

Levine, L., Garcia, C. T., & Oh, W. (1985). Determinants of mother-infant interaction in adolescent mothers. Pediatrics, 75, 23-29.

Millon, T. (1982). Millon Clinical Multiaxial Inventory Manual. Minneapolis: National Computer Systems.

Mott, F. L., & Heurin, R. J. (1988). Linkages between sexual activity and alcohol and drug use among American adolescents. Family Planning Perspectives, 20, 128-136.

National Institute on Drug Abuse (1987). The Adolescent Assessment/Referral System. Manuscript available from the U.S. Department of Health and Human Services, Alcohol, Drug Abuse and Mental Health Administration.

Shah, F., Zelnik, M., & Kanther, J. F. (1973). Unprotected intercourse among unwed teenagers. Family Planning Perspectives, 1, 39-44.

Zuckerman, B., Amaro, H., Bauchner, H., Cabral, H. (1989). Depressive symptoms during pregnancy: Relationship to poor health behaviors. American journal of Obstetric Gynecology, 160, 1107-1011.