Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 7.
Published in final edited form as: J Res Adolesc. 2018 Mar;28(1):56–69. doi: 10.1111/jora.12344

Neural Correlates of Risky Sex and Response Inhibition in High-Risk Adolescents

Natasha S Hansen 1, Rachel E Thayer 2, Sarah W Feldstein Ewing 3, Amithrupa Sabbineni 4, Angela D Bryan 5
PMCID: PMC8900213  NIHMSID: NIHMS1780962  PMID: 29460357

Abstract

Adolescence is a neurodevelopmental period of heightened sexual risk taking. Neuroimaging can help elucidate crucial neurocognitive mechanisms underlying adolescent sexual risk behavior, yet few empirical studies have investigated this neural link. To address this gap in the literature, we examined the association between neurocognitive function during response inhibition—a known correlate of risk behaviors—and frequency of intercourse without a condom among adolescents. We examined the correlation between condom use and fMRI-based Stroop response in a large ethnically diverse sample of high-risk adolescents (n = 171). Partially replicating previous literature, sexual risk was positively correlated with blood-oxygen-level-dependent (BOLD) activation in the middle frontal gyrus during response inhibition, highlighting the relevance of this region during risky sexual decision making within this age group.

Adolescence is a neurodevelopmental period often characterized by increased sexual risk taking (Centers for Disease Control and Prevention [CDC], 2016). More than any other sexual behavior, penetrative sex without a condom puts people at risk of sexually transmitted infections (STIs), and unplanned pregnancy (CDC, 2016). Adolescents contribute disproportionately to the negative consequences of sexual risk behaviors, including unprotected intercourse; of the 20 million new cases of STIs in the United States each year, approximately half occur among 15–24-year-olds (CDC, 2016). Unplanned pregnancies are also disproportionately common among adolescents (CDC, 2015). Despite recent declines, the U.S. teen pregnancy rate remains one of the highest in the Western world (Sedgh, Finer, Bankole, Eilers, & Singh, 2015), and an estimated 82% of these pregnancies among young women aged 15–19 years are unintended, the highest rate of any age group (Finer & Zolna, 2014). Teen pregnancies cost U.S. taxpayers an estimated $9.4 billion in increased foster care and health care, as well as lost tax revenue from teen mothers’ lower education attainment and diminished income potential (CDC, 2017). The negative consequences of adolescent pregnancies are also more likely to impact racial/ethnic minorities, contributing to the perpetuation of health and resource disparities among minority populations in the United States (CDC, 2015; Martin, Hamilton, Osterman, Driscoll, & Mathews, 2017).

The numerous negative health impacts of STIs and unplanned pregnancies make understanding the factors that contribute to the high rates of unprotected sex among adolescents a pressing public health concern. One possibility is that adolescents in the United States simply do not yet have the neurodevelopmental processes in place to fully unravel the risks associated with unprotected sex, or are unaware of how to mitigate those risks through condom use. While it has been argued that many areas of the United States could benefit from more practical sexual education programs shown to be effective in increasing teen condom use (see Bryan, Gillman, & Hansen, 2016), absence of condom education alone is not enough to fully explain the high rates of unprotected sex among US adolescents. For example, an ethnically diverse study of over 700 adolescents found that most adolescents successfully identify the dangers of unprotected sex, yet the report suggests their sexual decision making is driven more by perceived benefits of forgoing condom use (e.g., greater spontaneity and intimacy with the sexual partner) than by estimation of the potential costs (Parsons, Halkitis, Bimbi, & Borkowski, 2000). Adolescence is widely viewed as the prolonged transition period between puberty and the presumption of adult roles, and data continue to reflect that the adolescent brain is undergoing extensive changes from ages 14 to 25, at which point, for many youth, the brain has fully matured (Casey, Galvan, Hare, Friederici, & Ungerleider, 2005; Spear, 2000).

A rich empirical literature suggests that successful response inhibition—the ability to inhibit a prepotent drive toward a desirable outcome, even in the face of a potential negative consequence—is associated with making safer sexual decisions. Research with populations ranging from adult substance users (Golub, Starks, Kowalczyk, Thompson, & Parsons, 2012; Nydegger, Ames, Stacy, & Grenard, 2014) to emerging adults (Derefinko et al., 2014) and high-risk adolescents (Robbins & Bryan, 2004), demonstrates associations between stronger self-reported response inhibition abilities and safer sexual behaviors. Recent advances in neuroimaging have allowed more precise investigation of the neurocognitive mechanisms underlying this association between response inhibition and adolescent risky decision making in the domain of sexual behavior. In an experimental behavioral study, adult participants were assessed for response inhibition prior to being randomized to observe a sexual or neutral video before performing a go/no-go task with sexual or neutral stimuli (Macapagal, Janssen, Fridberg, Finn, & Heiman, 2011). Results indicated no relationship between response inhibition and behavioral task performance in the neutral conditions, but a stronger relationship between response inhibition and performance for sexual stimuli and after the sexual video. Although conducted with adults, these data suggest that even individuals who have good capacity for response inhibition in low-valence or “cold” conditions, may experience interruption in their ability to successfully engage and follow through with response inhibition in emotionally laden and highly valent “hot” cognitive conditions. These results suggest that adolescents, who are in the midst of developing successful cognitive control centers, when presented with these “hot” cognitive conditions, are particularly likely to be unable to inhibit impulses in a sexual context (Casey, Jones, & Hare, 2008).

To the best of our knowledge, only two empirical studies have investigated neurocognitive mechanisms underlying response inhibition in adolescent risky sex. The first used a letter-based go/no-go task (all letters except the letter “X” = go, the letter “X” = no-go) in a block design fMRI paradigm to examine the neural correlates of contraceptive use in a small sample of sexually active adolescents (n = 20), ages 15–17 years (Goldenberg, Telzer, Lieberman, Fuligni, & Galván, 2013). Contraceptive use was measured on a 5-point scale (1 = condom and birth control, 2 = only condom, 3 = only birth control, 4 = withdrawal, 5 = none), where higher numbers indicated greater sexual risk. Results indicated there was no relationship between task performance and sexual risk behavior; however, there was a significant negative correlation between sexual risk and blood-oxygen-level-dependent (BOLD) response such that more sexual risk (less contraceptive use) was associated with less activation in the right inferior frontal gyrus (rIFG) and insula during response inhibition (operationalized as successful no-go vs. go trials).

The only other published empirical study with adolescents examining the neural correlates of response inhibition in the context of risky sex also used a letter-based go/no-go task (the letter “X” = go, the letter “K” = no-go), but with an event design, a significantly larger sample (n = 95), and a slightly broader age range (i.e., ages 14–18 years (Feldstein Ewing, Houck, & Bryan, 2015). Here, risky sex was defined as days of sexual intercourse without a condom over the past month. Like the Goldenberg et al. (2013) study, results indicated a significant relationship between risky sex and BOLD response during response inhibition, again operationalized as no-go trials without a response compared to go trials with a response. Feldstein Ewing and colleagues (2015) found a significant association between BOLD signal in the rIFG and condomless intercourse days. However, the correlation was in the opposite direction: greater BOLD activation in the rIFG was associated with more condomless intercourse days. Taken together, these two studies suggest the relevance of the right prefrontal cortex, specifically, the rIFG, in response inhibition and risky sex among adolescents. These results are highly consistent with broader neuroscience studies; meta-analyses of evidence spanning human fMRI and brain lesion studies suggest the rIFG and associated neural networks characterize a “brake” system, and thus play a critical role in response inhibition and other forms of executive control, relevant across a wide range of behaviors (Aron, Robbins, & Poldrack, 2004, 2014).

Neural correlates of response inhibition have similarly been investigated in the context of other important adolescent risk behaviors such as alcohol and cannabis use using the Stroop task, another widely studied and empirically robust measure of response inhibition (Banich et al., 2007; Hatchard, Fried, Hogan, Cameron, & Smith, 2014; Hatchard et al., 2015). Results of these studies have shown a pattern whereby adolescent substance users and nonusers achieved equivalent Stroop performance, but differed significantly with respect to their neural activation as they performed the task. Specifically, adolescent substance users showed increased activation compared to nonusers across a variety of brain regions including the thalamus, right medial prefrontal areas, bilateral parahippocampal regions, cerebellum, and caudate. These outcomes have been interpreted as potentially reflecting a compensatory strategy among adolescent substance users, in which they may require greater brain activation to achieve the same inhibitory performance as nonusers (Hatchard et al., 2014). To the best of our knowledge, no study has yet used the Stroop task to investigate the underlying neurocognitive correlates of response inhibition in the context of adolescent condom use, but results from the two adolescent risky sex fMRI studies using the go/no-go task (Feldstein Ewing et al., 2015; Goldenberg et al., 2013) suggest a common neural signature in the rIFG across adolescent sexual risk behaviors.

In summary, the high rate of sexual risk behaviors among adolescents is a pressing public health concern. In particular, sexual intercourse without a condom carries a heightened risk of STIs and unplanned pregnancy, both of which can have widespread and devastating health consequences. fMRI-based evaluations of response inhibition could provide the missing link regarding why many adolescents may be aware of the risks associated with unprotected sex and the protective efficacy of condoms, yet often do not translate this knowledge into prudent sexual decision making. The purpose of this study was thus to provide data to extend the nascent field of response inhibition in the context of adolescent risky sex to a large ethnically diverse sample of high-risk adolescents using a different empirically validated response inhibition task, the Stroop (Andrews-Hanna et al., 2011), which has been used in multiple fMRI studies of other adolescent risk behaviors (Banich et al., 2007; Hatchard et al., 2014, 2015; Thayer et al., 2015). Selection of the Stroop allowed for the expansion of construct validity for the measurement of response inhibition beyond that established by the two previous studies, which both used a similar go/no-go task. Based on the mixed findings of the two existing risky sex studies (Feldstein Ewing et al., 2015; Goldenberg et al., 2013), we did not have an a priori directional hypothesis, but instead posited that we would find a significant association between adolescent condom use and BOLD response during response inhibition in the rIFG, the region found to be significantly related to sexual risk taking in both studies.

METHODS

Participants

Participants were adolescents recruited from juvenile justice partner programs in the southwest United States as part of a large longitudinal intervention study (Feldstein Ewing et al., in press). All data included in this article were collected at a baseline assessment, prior to randomization to intervention condition. Research staff members informed adolescents that study participation was entirely voluntary and separate from their juvenile justice involvement. All youth provided written assent prior to the start of the study. For adolescents under age 18, informed consent was also obtained via phone from a parent or legal guardian prior to study participation (e.g., Schmiege, Broaddus, Levin, & Bryan, 2009). Audio consents were recorded and archived to provide proof of consent. For inclusion in the study, participants were required to be 14–18 years old and fluent in English. Exclusion criteria included current use of anti-convulsant or antipsychotic medications; use of illicit substances (excluding tobacco and cannabis) more than three times in the previous month; loss of consciousness for more than five minutes in the 6 months prior to screening; and MRI contraindications (e.g., claustrophobia, current pregnancy, or irremovable metal piercings or implants). There were 250 participants who completed the baseline assessments and the Stroop fMRI task. After relevant exclusions for the study questions examined herein, the sample was predominantly male (70.8%), Hispanic (55.6%), and/or bi/multi-racial (24%). Ages ranged from 14 to 18 years old (M = 16.2; SD = 1). The average participant had completed 9.2 years of education at the time of enrollment (SD = 1.2). Additional demographic details are presented in Table 1. All study procedures were performed with approval of the participating institutional review board, and under a federal certificate of confidentiality.

TABLE 1.

Sample Characteristics (n = 171)

Mean (SD) Range
Gender Male = 70.8%
Female = 29.2%
Age 16.2 (1) 14–18
Years of education (Last grade completed) 9.2 (1.2) 5–12
Race/ethnicity Mixed race/ethnicity = 24%
Hispanic = 55.6%
Caucasian = 10.5%
African American = 3.5%
Native American = 3.5%
Other = 2.9%
Condom usea at last intercourse (n = 171) Yes = 49.7%
No = 50.3%
Risky sexb (n = 126) 10.8 (7.7) 1–30
Open in a new tab
a

Condom use = condom used during most recent penetrative sex act: Yes or No (higher numbers indicate less risk).

b

Risky sex = frequency of penetrative sex acts without a condom in the past 3 months (higher numbers indicate more risk).

Measures

The measures included: (1) a self-report demographics questionnaire, (2) a self-report questionnaire designed to assess risky sexual behavior, and (3) an fMRI Stroop paradigm. The demographics questionnaire queried participants’ self-reported gender, age, race/ethnicity, and number of years of education completed at enrollment. Two variables widely used in other studies of adolescents were used to measure risky sexual behavior (Bryan, Schmiege, & Broaddus, 2009; Schmiege et al., 2009): Condom Use and Risky Sex. Participants were first asked whether they had ever had sexual intercourse, defined as penile-vaginal or penile-anal penetration. Those who answered “yes” then answered the main sexual risk variables. Condom Use was defined as whether or not a condom was used during the most recent penetrative sex act. Participants were asked to respond to the question, “The last time you had sex, did you use a condom?” (0 = no; 1 = yes). Thus, for Condom Use, a higher number indicates less sexual risk. Risky Sex was defined as frequency of sexual intercourse without a condom in the past 3 months. Frequency of sexual intercourse was assessed with, “On average, in the past 3 months only, how often have you had sexual intercourse?” (1 = never; 2 = once a month; 3 = once a week; 4 = 2–3 times a week; 5 = 4–5 times a week; 6 = almost every day). Frequency of condom use was assessed with, “In the past three months, how much of the time did you use condoms when you had sexual intercourse?” (1 = never; 2 = almost never; 3 = sometimes; 4 = almost always). Consistent with prior work in this area (Schmiege et al., 2009), a Risky Sex index was calculated by multiplying frequency of sexual intercourse by reverse-coded condom use, with a higher number indicating more sexual risk. It is expected that these two sexual risk variables would be significantly correlated but not entirely redundant with one another. Risky Sex captures behavior over a 3-month time frame, and thus provides an assessment of typical sexual risk behavior. However, like all retrospective assessments, it is vulnerable to both memory errors and response biases. By focusing only on the most recent intercourse, Condom Use minimizes the risk of memory errors and response biases, but is in turn vulnerable in that the most recent sexual event may not correspond to a person’s typical pattern of behavior. Thus, the two measures were selected to complement one another’s strengths and weaknesses, and allow the study to triangulate the construct of sexual risk with multiple assessments.

Functional MRI Paradigm: Stroop Task

Consistent with other publications on this task (Thayer et al., 2015), we utilized the functional Color–Word Stroop task developed and empirically tested by Andrews-Hanna et al. (2011) to examine response inhibition. Participants were presented with a series of words printed in different colors and were instructed to indicate the color in which each word was printed, regardless of the word’s meaning, using one of four buttons on hand-held button boxes. The task included three trial types: congruent, incongruent, and neutral. In congruent trials, a color word would appear in a font that matched the word’s meaning (e.g., the word “blue” printed in blue font). In incongruent trials, a color word would appear in a font that did not match the word’s meaning (e.g., the word “blue” printed in red font). In neutral trials, a non-color word would appear in a random font color (e.g., the word “bond” printed in blue font). Neutral non-color words were matched for word length with congruent and incongruent color words. Participants were not required to correct any errors before subsequent trials proceeded.

Stroop trials were clustered into congruent, incongruent, and neutral trials for a block design fMRI analysis. Each block was comprised of 12 trials: six trials with stimuli specific to that block (e.g., congruent, incongruent, or neutral), and the other six trials with neutral stimuli that were consistent across all block types (see Figure 1). Thus, within the congruent blocks, participants completed six block-specific congruent trials, mixed in with six block-general neutral trials; within incongruent blocks, participants completed six block-specific incongruent trials, mixed in with six block-general neutral trials; and within the neutral blocks, participants completed six block-specific neutral trials, mixed in with six block-general neutral trials. The neutral stimuli that appeared across all three block types were included to reduce potential habituation effects within each block, and to help prevent participants from relying on a strategy (e.g., merely reading the words during the congruent blocks). Within blocks, trials were pseudo-randomly ordered to ensure no more than two of the same trial type appeared in a row. During the fMRI scanning session, each participant completed two Stroop runs, each with nine congruent, incongruent, or neutral blocks, for a total of 216 trials. In each trial, the word was displayed for 1,500 ms, separated by a 500 ms fixation cross. Thus, each block of 12 trials interleaved with 12 fixation crosses lasted 24 s, and each run of 9 blocks lasted 216 s, for a total Stroop task time of 432 s, or approximately 7.2 min. Consistent with other publications in this area (Andrews-Hanna et al., 2011; Thayer et al., 2015) the contrast of interest was activation during incongruent blocks minus activation during congruent blocks (incongruent>congruent).

FIGURE 1.

FIGURE 1

Open in a new tab

Schematic of fMRI Stroop paradigm.

Image Acquisition and Spatial Preprocessing

Participants were scanned using a 3T Siemens Trio (Erlangen, Germany) whole-body MRI scanner with a 12-channel head coil. High-resolution structural brain images were acquired with a multi-echo magnetization-prepared rapid gradient-echo (MPRAGE) sequence (TE = 1.64, 3.50, 5.36, 7.22, and 9.08 ms, TR = 2.53 s, TI = 1.20 s, flip angle = 7°, NEX = 1, slice thickness = 1 mm, 33 slices, FOV [field of view] = 256 mm, and in-plane resolution = 256 × 256). Structural images were acquired oblique to the anterior-posterior commissure line to diminish susceptibility artifacts. Functional images (BOLD) during the Stroop task were acquired using a single-shot, gradient-echo echoplanar pulse sequence (TR = 2,000 ms; TE = 29 ms; flip angle = 75°; FOV = 240 mm; matrix size = 64 × 64). All functional data were processed using SPM8 software (http://www.fil.ion.ucl.ac.uk/spm/software/spm8/). fMRI images were motion-corrected, spatially normalized to the standardized space established by the Montreal Neurologic Institute (MNI; www.bic.mni.mcgill.ca), resampled to 2 mm3 voxels, and smoothed with a three-dimensional Gaussian kernel of 8 mm width (FWHM).

Analytic Strategy

After preprocessing, a whole-brain voxel-wise first-level analysis was conducted to obtain brain regions that showed an increase in BOLD signal for the contrast of interest (incongruent>congruent). The SPM8 general linear model (GLM) was applied to the functional time series, convolved with the canonical hemodynamic response function (HRF; Friston, Frith, Turner, & Frackowiak, 1995), and a 128 s high-pass filter. For each subject, condition effects were then modeled with box-car regressors representing the occurrence of each block type (incongruent and congruent). For each subject, condition effects were estimated at each voxel, and contrast images were produced for the contrast (incongruent>congruent). To estimate conditions on the group level, these contrast images from the individual subjects’ analysis were entered into a one-sample t-test to examine task main effect in the incongruent>congruent contrast. For this whole-brain analysis, we report activations/deactivations with a p < .001 and a threshold of ≥100 contiguous voxels.

The association between risky sexual behavior and response inhibition was evaluated using two multiple regression analyses in SPM8: (1) to examine the association between Condom Use and BOLD activation in the Stroop incongruent>congruent contrast, and (2) to examine the association between Risky Sex and BOLD activation in the Stroop incongruent>congruent contrast. In each analysis, the Stroop SPM contrast image (incongruent>congruent) was the dependent variable, predicted by Condom Use (regression 1) or Risky Sex (regression 2). In both regressions, age and gender were included as covariates.

For this analysis, based on the two published studies in this area (Feldstein Ewing et al., 2015; Goldenberg et al., 2013), we defined an a priori spheric search territory of 1.5 cm radius in the right prefrontal cortex (rIFG). Within this a priori specified search territory, we considered those locations as significant that surpassed a p ≤ .001 and a threshold of ≥100 contiguous voxels (see also below: Results, Neural Correlates of Sexual Risk). This approach is advantageous over using existing Type I error correction methods because, in light of the two existing findings used to guide our a priori search territory, those approaches would be overly conservative by unnecessarily increasing the risk of Type II error (i.e., missing a true effect). To help ensure the validity of our findings, we set a conservative alpha (p < .001) and a threshold of ≥100 contiguous voxels. In addition, we ran partial correlations to determine whether Condom Use or Risky Sex were associated with Stroop behavioral task performance (error rate and reaction time overall and within each trial type). All partial correlation tests included age and gender as covariates.

RESULTS

Missing Data, Movement in the Scanner, and Stroop Task Accuracy

The original study involved 250 sexually active youth who completed the Stroop fMRI paradigm. Of those, some participants were missing the Condom Use variable (n = 39), and an additional subset were missing the Risky Sex variable (n = 45). Exclusions for data quality followed previous work (Thayer et al., 2015), including mean FD > .5 (n = 22) and Stroop task accuracy ≤66% correct responses across all trials (n = 18), thus leaving a final sample of 171 participants with the Condom Use variable, and a subset of 126 with the Risky Sex variable. In the final sample, average framewise displacement was .25 degrees (SD = .1), and average Stroop task accuracy was 89% correct (SD = 7%).

Within this sample of sexually active youth, for Condom Use (n = 171) 49.7% reported they had used a condom during their most recent penetrative sex act, and 50.3% reported they had not used a condom the last time they had sex. For Risky Sex (n = 126), frequency of penetrative sex acts without a condom in the past 3 months ranged from 1 to 30 (M = 10.8; SD = 7.7; Table 1). Descriptively, participants scored an average of 3.4 (SD = 1.4) on the frequency of sex scale, which corresponds to having sex on average once a week in the past 3 months. The past 3 months average frequency of sex scale percentage breakdown was as follows: never = 2.3%; once a month = 22.8%; once a week = 15.8%; 2–3 times a week = 19.3%; 4–5 times a week = 5.8%; almost every day = 8.8%. Participants scored an average of 3 (SD = 1.6) on the past 3 months condom use frequency scale, which corresponds to using condoms on average “sometimes” during sex in the past 3 months, with 21.1% reporting they never used a condom, 25.4% reporting they always used a condom, and the remaining 46.9% reporting they only inconsistently used a condom during sex in the past 3 months. As expected, there was a strong negative correlation (Cohen, 1988) between responses to the two sexual risk variables, Risky Sex and Condom Use, in our sample, r(124) = −.65, p < .001.

Association Between Sexual Risk and Stroop Task Performance

Neither Condom Use nor Risky Sex was significantly correlated with Stroop task performance as measured by either error rate or reaction time overall or within any individual trial type (incongruent, congruent, or neutral).

Stroop Task fMRI Main Effects

A one-sample t-test was used to examine task main effects. Results indicated multiple areas of significant activation (incongruent>congruent; uncorrected p ≤ .001; extent threshold ≥100 voxels). Areas of positive BOLD response included the left inferior frontal gyrus, left precuneus, right insula, left thalamus, and left temporal lobe. Areas of negative response (incongruent<congruent; uncorrected p ≤ .001; extent threshold ≥100 voxels) included the right and left insula, right middle frontal gyrus, and left parahippocampal gyrus (see Table 2).

TABLE 2.

Main Effects of Response Inhibition Stroop Task (Incongruent vs. Congruent)

Region BA KE t X Y Z
Activation (Incongruent>Congruent)
 L Frontal Lobe, Inferior Frontal Gyrus 9 69,478 8.69 −38 4 30
 L Parietal Lobe, Precuneus 19 24,200 7.66 −29 −61 39
 R Sub-lobar, Insula 13 16,629 6.55 34 22 12
 L Sub-lobar, Thalamus NA 13,892 6.12 −9 −11 11
 R Sub-lobar, Insula 13 173,430 5.77 35 21 13
 L Temporal Lobe, Sub-Gyral 37 1,700 5.85 −44 −49 −7
 R Cerebellum, Culmen NA 6,985 5.73 31 −61 −23
 R Cerebellum, Pyramis NA 4,262 5.09 7 −66 −23
 R Parietal Lobe, Superior Parietal Lobule 7 4,633 4.39 28 −61 44
 R Frontal Lobe, Inferior Frontal Gyrus 9 1,668 4.38 39 5 28
 L Cerebellum, Culmen NA 1,117 4.12 −38 −48 −21
 R Limbic Lobe, Parahippocampal Gyrus 28 275 4.02 21 −18 −25
 R Occipital Lobe, Middle Occipital Gyrus 18 315 3.94 40 −87 10
 R Frontal Lobe, Inferior Frontal Gyrus 10 1,120 3.86 52 41 −1
 R Temporal Lobe, Superior Temporal Gyrus 38 170 3.7 44 17 −38
 L Occipital Lobe, Lingual Gyrus 17 439 3.65 −17 −103 −9
 R Frontal Lobe, Sub-Gyral 6 107 3.39 24 1 52
Deactivation (Incongruent<Congruent)
 R Sub-lobar, Insula 13 26,653 6.86 40 −11 17
 Left Cerebrum, Sub-lobar, Insula 13 33,686 6.31 −37 −12 20
 R Frontal Lobe, Medial Frontal Gyrus 10 6,159 5.92 8 46 −3
 R Limbic Lobe, Parahippocampal Gyrus 34 1,959 4.74 20 −11 −14
 R Occipital Lobe, Lingual Gyrus 18 1,307 4.67 14 −68 −4
 R Limbic Lobe, Cingulate Gyrus 31 2,521 4.60 6 −44 31
 L Occipital Lobe, Cuneus 18 5,851 4.50 −8 −88 20
 L Occipital Lobe, Lingual Gyrus 18 1,294 4.41 −15 − 74 −5
 R Temporal Lobe, Superior Temporal Gyrus 38 1,661 4.30 57 4 −9
 R Limbic Lobe, Anterior Cingulate 25 785 4.21 1 8 −7
 R Parietal Lobe, Postcentral Gyrus 3 3,206 4.10 22 −33 68
 R Limbic Lobe, Parahippocampal Gyrus 30 1,109 3.99 15 −44 1
 L Parietal Lobe, Precuneus 7 384 3.92 −16 −50 56
 R Temporal Lobe, Middle Temporal Gyrus 37 585 3.71 46 −58 1
 L Frontal Lobe, Paracentral Lobule 31 607 3.69 −8 −12 46
 L Frontal Lobe, Medial Frontal Gyrus 10 124 3.59 −11 56 13
 R Frontal Lobe, Middle Frontal Gyrus 8 142 3.55 19 29 43
Open in a new tab

Note. Uncorrected p < .001; extent threshold ≥100 voxels. Each cluster is listed with the corresponding peak voxels of activation (BA = Brodmann area; KE = cluster size; R = right; L = left).

Neural Correlates of Sexual Risk

For Condom Use, results indicated a significant negative correlation between use of a condom during most recent penetrative sex act and BOLD response in the right dorsolateral prefrontal cortex (dlPFC), middle frontal gyrus (rMFG; Brodmann area 9; max voxel: X = 40, Y = 26, Z = 34; uncorrected p < .001; extent threshold ≥100 voxels; see Table 3 and Figures 2 and 3), such that more BOLD response was associated with less condom use/more sexual risk. For Risky Sex, we found a significant positive correlation between frequency of penetrative sex acts without a condom in the past 3 months and BOLD response in the same regions (right dlPFC, rMFG, Brodmann area 9; max voxel: X = 49, Y = 30, Z = 36; uncorrected p < .001; extent threshold ≥100 voxels; see Table 3 and Figures 2 and 3), such that more BOLD response was associated with more sexual risk. There were no significant correlations between sexual risk and deactivation in BOLD signal in the aforementioned a priori region.

TABLE 3.

Neural Correlates of Condom Use and Risky Sexual Behavior in the Stroop Incongruent > Congruent Contrast.

Region BA KE t X Y Z
Condom Usea
 R Occipital Lobe, Fusiform Gyrus 18 333 3.77 20 −89 −18
 R Frontal Lobe, Middle Frontal Gyrusc 9 605 3.63 40 26 34
 L Frontal Lobe, Superior Frontal Gyrus 6 357 3.58 −6 15 61
 L Frontal Lobe, Precentral Gyrus 9 242 3.49 −31 9 32
 L Frontal Lobe, Middle Frontal Gyrus 6 100 3.35 −32 8 51
Risky Sexb
 R Occipital Lobe, Superior Occipital Gyrus 19 460 4.17 36 −83 32
 R Frontal Lobe, Middle Frontal Gyrusc 9 234 3.61 49 30 36
 L Frontal Lobe, Middle Frontal Gyrus 6 216 3.55 −35 5 55
 L Frontal Lobe, Superior Frontal Gyrus 6 199 3.5 −3 22 57
Open in a new tab

Note. The correlation with activations presented above are negative for Condom Use (higher numbers indicate less risk), and positive for Risky Sex (higher numbers mean more risk) in the incongruent>congruent Stroop contrast. Both regressions include age and gender as covariates of noninterest. Uncorrected p < .001; extent threshold ≥100 voxels. Each cluster is listed with the corresponding peak voxels of activation (BA = Brodmann area; KE = cluster size; R = right; L = left).

a

Condom Use = condom used during most recent penetrative sex act: Yes or No.

b

Risky Sex = frequency of penetrative sex acts without a condom in the past 3 months.

c

Indicates a priori specified region.

FIGURE 2.

FIGURE 2

Open in a new tab

(a) Negative correlation between Condom Use and BOLD activation in the Middle Frontal Gyrus (MFG; Brodmann Area 9) in the incongruent>congruent Stroop contrast. Condom Use = condom used during most recent penetrative sex act: Yes or No (higher numbers indicate less risk). (b) Positive correlation between Risky Sex and BOLD activation in the Middle Frontal Gyrus (MFG; Brodmann Area 9) in the incongruent>congruent Stroop contrast. Risky Sex = frequency of penetrative sex acts without a condom in the past 3 months (higher numbers mean more risk). Uncorrected p < .001; extent threshold ≥100 voxels.

FIGURE 3.

FIGURE 3

Open in a new tab

(Left) Negative correlation between Condom Use and BOLD activation in the right Middle Frontal Gyrus (5 mm sphere surrounding MFG max voxel: X = 40, Y = 26, Z = 34) in the incongruent>congruent Stroop contrast, r(169) = −.25, p < .001. Condom Use = condom used during most recent penetrative sex act: Yes or No (higher numbers indicate less risk). (Right) Positive correlation between Risky Sex and BOLD activation in the Middle Frontal Gyrus (5 mm sphere surrounding MFG max voxel: X = 49, Y = 30, Z = 36) in the incongruent>congruent Stroop contrast, r(124) = .27, p < .01. Risky Sex = frequency of penetrative sex acts without a condom in the past 3 months (higher numbers mean more risk). Uncorrected p < .001; extent threshold ≥100 voxels.

A post hoc analysis also revealed a significant positive correlation between sexual risk and BOLD activation in the right occipital lobe for both sexual risk variables (incongruent>congruent; Condom Use: Brodmann area 18; Risky Sex: Brodmann area 19; uncorrected p < .001; extent threshold ≥100 voxels) and a significant positive correlation between sexual risk and BOLD activation in Brodmann area 6 for both sexual risk variables (incongruent>congruent; uncorrected p < .001; extent threshold ≥100 voxels) (see Table 3).

DISCUSSION

This study examined associations between risky sexual behavior and neural activation during an empirically validated and widely utilized response inhibition task in a large ethnically diverse sample of sexually active adolescents. Compellingly, in the two published fMRI studies investigating the neurocognitive mechanisms of response inhibition associated with risky sex in adolescents (Feldstein Ewing et al., 2015; Goldenberg et al., 2013), there has been evidence for an association between adolescent risky sex and BOLD response in the inferior frontal gyrus, although, the direction of this relationship has been inconsistent across studies. Thus, we anticipated replicating and extending that work using a different operationalization of the response inhibition task, and hoped to provide some clarity as to the direction of the association.

In support of our hypothesis, we found a significant association between risky sexual behavior and BOLD signal during response inhibition in our a priori specified region of the right prefrontal cortex, although our voxel of peak activation was fractionally more dorsal in the rMFG. Specifically, for both Condom Use and Risky Sex, we found a positive correlation between BOLD response in rMFG during response inhibition and risky sex. There were no significant associations between either sexual risk variable and Stroop task performance (error rate, or reaction time overall, or on any individual trial type).

Our results replicate the findings reported by Feldstein Ewing et al. (2015) in both region, size of the region of response, and direction of the correlation in a completely different sample of adolescents, with a different response inhibition task, and with a different operationalization of risky sexual behavior. Both our study and the Feldstein Ewing et al. (2015) study showed that more sexual risk was associated with greater BOLD activation during response inhibition in the right prefrontal cortex. In fact, the max activation voxel in the rIFG reported by Feldstein Ewing et al. (2015; nogo>go; max voxel: X = 48, Y = 27, Z = 24; uncorrected p < .001), was only a few voxels away from our max activation voxels in the same overall region, although ours were marginally more dorsal on the rMFG (Stroop incongruent>congruent; Condom Use max voxel: X = 40, Y = 26, Z = 34; Risky Sex max voxel: X = 49, Y = 30, Z = 36; uncorrected p < .001). Our findings only partially replicate Goldenberg et al. (2013), which found an association between BOLD activation in a similar region of the right prefrontal cortex (rIFG), but in the opposite direction (more sexual risk was associated with less activation). It should be noted that the Goldenberg et al. (2013) study included a much smaller sample (n = 20), and a somewhat coarse ordinal measure of sexual risk that included ineffective “contraception” (i.e., withdrawal) and did not include any assessment of frequency of intercourse.

Recent reviews (e.g., Feldstein Ewing et al., 2016a) continue to highlight the multilevel nature of sexual risk among adolescents. Things that make sexual decision making challenging in this age group include the inconsistent availability of sexual opportunities (i.e., such opportunities are not always available or predictable during adolescence), as well as the requirement to engage in likely novel negotiations with a partner around sexual decision making. From there, a decision must be made about the desirability/salience of the activity, whether safer behavior is even possible (i.e., availability of a condom), and if that safer behavior is not possible, whether to default to the prepotent response of unprotected sex or risk losing the sexual opportunity.

Thus, response inhibition is an important factor underlying adolescent sexual risk behaviors, including unprotected intercourse (Hoyle, Feifar, & Miller, 2000). Our study investigated the neural mechanisms underlying response inhibition and condom use among adolescents, and found that sexually active adolescents exhibited no significant difference in behavioral performance on our response inhibition task compared to their less risky peers, but showed greater BOLD activation during response inhibition in the right dorsolateral prefrontal cortex (rDLPFC), specifically the rMFG, Brodmann area 9. Previous research has shown this area to be implicated in response inhibition (Horn, Dolan, Elliott, Deakin, & Woodruff, 2003), executive control over automatic behavior (Kübler, Dixon, & Garavan, 2006), reward-driven learning (Ridderinkhof, van den Wildenberg, Segalowitz, & Carter, 2004), and cognitive processing in emotional contexts (Kerestes et al., 2012)—all processes that may factor into engagement in an adolescent’s sexual decision making, including the decision to use a condom.

A review of the fMRI literature on neural correlates of response inhibition using different versions of the Stroop task found that more difficult versions of the task required greater recruitment of areas involved in response inhibition, including the DLPFC, than did less difficult versions of the task, resulting in a positive correlation between Stroop task difficulty and BOLD activation across multiple studies (Mitchell, 2005). Our findings also replicate the results of previous literature that used the Stroop task to investigate neural correlates of response inhibition in the context of other adolescent risk behaviors and similarly found that greater risk behavior was associated with increased BOLD activation in the absence of differences in Stroop task performance (Banich et al., 2007; Hatchard et al., 2014, 2015). As in these previous studies, our results suggest sexually risky adolescents may have more difficulty inhibiting their prepotent responses, and may therefore require greater compensatory recruitment of inhibitory brain regions to achieve the same inhibitory performance as their less sexually risky peers.

Our post hoc analyses also revealed a significant positive correlation between BOLD response and sexual risk (across both variables) in the lMFG and lSFG (Brodmann area 6; uncorrected p < .001; extent threshold ≥100 voxels), and in the right occipital lobe (Brodmann area 18 and 19; uncorrected p < .001; extent threshold ≥100 voxels). Feldstein Ewing et al. (2015) also reported finding a positive correlation between sexual riskiness and response inhibition-related BOLD activation in the occipital lobe, although on the opposite side. Brodmann area 6 has traditionally been shown to be involved in motor control (Tanaka, Honda, & Sadato, 2005), and the occipital lobe is known for its prominent role in visual processing (Grill-Spector & Malach, 2004). These results likely reflect the noise of task processing, as occipital lobe and motor region activation fluctuations are often observed in fMRI studies that use visual and motor-based tasks. Our analyses did not suggest any significant association between sexual risk and response inhibition-related BOLD signal in the insula, as reported by Goldenberg et al. (2013).

Limitations and Future Directions

Although our study has several noteworthy strengths, including a large and ethnically diverse sexually active sample to investigate an important and relatively underexplored topic, it also has limitations. Our study only investigated one important adolescent risk behavior. Future studies could directly compare the neurocognitive mechanisms involved in risky sex to those underlying other important high-risk behaviors among adolescents, such as illicit substance use, as the limited existing literature on this topic suggests there may be differential neural correlates involved in these different risk behaviors (Feldstein Ewing et al., 2015; Thayer et al., 2015). We conducted our work with adolescents involved in the juvenile justice system, who some argue may differ qualitatively from their non–justice involved peers. However, such adolescents are merely at the higher end of the normal distribution in terms of risk behavior (Skeem, Scott, & Mulvey, 2014), and work such as ours that focuses on those with higher levels of risk is perhaps most critical for the overall reduction in negative outcomes of risk behavior at the societal level. Further, the presence of many of these young people in the justice system is likely the result of circumstances in their communities (e.g., greater police presence) rather than severity of conduct disorder per se. Our version of the Stroop used only neutral stimuli, although some past research has suggested there may be something unique about response inhibition in the presence of sexually charged stimuli (Macapagal et al., 2011), another topic to be explored in future research. The block design analysis used in our study offers a more robust fMRI signal than rapid event-related paradigms, but also poses the disadvantage that activations cannot be traced to individual task trials. Future research in this area may therefore consider replicating these findings using a rapid event-related fMRI Stroop paradigm. Further, response inhibition is only one important neurocognitive factor involved in risky sexual decision making (Hoyle et al., 2000). A comprehensive understanding of how brain development and function may drive adolescent sexual risk behaviors would require future research to elucidate the neural mechanisms underlying other known predictors of sexual risk in adolescents, including peer influence (Reyna & Farley, 2006), emotion dysregulation (Messman-Moore, Walsh, & DiLillo, 2010), and potentially emotional urgency (Zapolski, Cyders, & Smith, 2009). Finally, our study is limited by its cross-sectional design and narrow age range. Future work by our team and others is focusing on using longitudinal designs, to evaluate the temporal relationships between neural differences and sexual risk behavior, and how the interaction of these variables may shift as the brain matures over the course of adolescence.

Despite its limitations, our study provides a valuable contribution by replicating and expanding the existing literature and furthering the scientific understanding of an important and still largely unexplored topic: the neural mechanisms underlying risky sex among sexually active adolescents. MRI and other in-vivo neuroimaging techniques provide an invaluable tool for investigating the inner workings of the living adolescent brain. By integrating neuroscience findings, the scientific study of adolescent risk behaviors can expand beyond the limitations of simple self-report and behavioral assessments to gain key insights about underlying physiological differences in brain function. For example, although response inhibition has been shown to be an important factor contributing to many adolescent risk behaviors (Nigg et al., 2006), the behavioral tasks of response inhibition alone were not sufficient to capture any significant differences between risky and safe adolescents in our study, nor in either of the two previous studies of risky sex (Feldstein Ewing et al., 2015; Goldenberg et al., 2013), nor in studies of other adolescent risk behaviors (Banich et al., 2007; Hatchard et al., 2014, 2015). In all these studies, it was only with the addition of fMRI that meaningful underlying differences emerged in how risky teens use their brains during the process of response inhibition.

Beyond the contribution to the basic science of how brain function is associated with risk behavior, there is the potential for application of our findings. There is tremendous interest in the extent to which patterns of neural response might help us to understand both how our behavioral interventions work and for whom they work (Erickson, Creswell, Verstynen, & Gianaros, 2014; Feldstein Ewing, Tapert, & Molina, 2016b). For example, differences in fMRI brain activation patterns during response inhibition tasks are already successfully being used to predict treatment response in posttraumatic stress disorder (Falconer, Allen, Felmingham, Williams, & Bryant, 2013). It is our hope that our study, and future work in this area, may similarly be applied, either by using neural correlates of response inhibition to predict psychosocial intervention response (Erickson et al., 2014), or by training more effective response inhibition skills with risk behavior–relevant stimuli (Houben, Nederkoorn, Wiers, & Jansen, 2011). Our research also speaks to the importance of delivering preventive interventions that take into consideration variability in brain development and functioning, particularly in the area of executive control. Many of our interventions to reduce sexual risk focus on cold cognitions involving perceived benefits of condom use, building perceptions of self-efficacy for condom use, and increasing intentions to use condoms. Our data support the idea that we also need to address the reality that protective behaviors in the context of sexual activity sometimes occur in circumstances that involve rather more impulsive decision making in the heat of the moment. Developing behavioral intervention content that facilitates young people’s abilities to stop, take a step back, and consider the pros and cons of risky versus safer behavior is critical to protecting them from the deleterious health impacts of high-risk sexual behaviors, and insuring they mature into safe and healthy adults.

Acknowledgments

This research was supported by 1R01NR013332-01 to SWFE and ADB. Authors thank Andrew B. Dodd and Andrew R. Mayer for their work in designing the fMRI data preprocessing pipeline and first-level analyses. The authors declare no conflicts of interest. SWFE and ADB designed research; NSH, RET, and ARM designed and performed analyses; NSH wrote the paper; and RET, SWFE, and ADB edited the paper.

Contributor Information

Natasha S. Hansen, University of Colorado Boulder

Rachel E. Thayer, University of Colorado Boulder

Sarah W. Feldstein Ewing, Oregon Health & Science University

Amithrupa Sabbineni, University of Colorado Boulder.

Angela D. Bryan, University of Colorado Boulder

REFERENCES

  1. Andrews-Hanna JR, Mackiewicz Seghete KL, Claus ED, Burgess GC, Ruzic L, & Banich MT (2011). Cognitive control in adolescence: Neural underpinnings and relation to self-report behaviors. PLoS ONE, 6, e21598. 10.1371/journal.pone.0021598 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aron AR, Robbins TW, & Poldrack RA (2004). Inhibition and the right inferior frontal cortex. Trends in Cognitive Sciences, 8, 170–177. 10.1016/j.tics.2004.02.010 [DOI] [PubMed] [Google Scholar]
  3. Aron AR, Robbins TW, & Poldrack RA (2014). Inhibition and the right inferior frontal cortex: One decade on. Trends in Cognitive Sciences, 18, 177–185. 10.1016/j.tics.2013.12.003 [DOI] [PubMed] [Google Scholar]
  4. Banich MT, Crowley TJ, Thompson LL, Jacobson BL, Liu X, Raymond KM, & Claus ED (2007). Brain activation during the Stroop task in adolescents with severe substance and conduct problems: A pilot study. Drug and Alcohol Dependence, 90, 175–182. 10.1016/j.drugalcdep.2007.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bryan AD, Gillman AS, & Hansen NS (2016). Changing the context is important and necessary, but not sufficient, for reducing adolescent risky sexual behavior: A reply to Steinberg (2015). Perspectives on Psychological Science, 11, 535–538. 10.1177/1745691616638093 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bryan AD, Schmiege SJ, & Broaddus MR (2009). HIV risk reduction among detained adolescents: A randomized, controlled trial. Pediatrics, 124, e1180–e1188. 10.1542/peds.2009-0679 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Casey B, Galvan A, Hare TA, Friederici AD, & Ungerleider LG (2005). Changes in cerebral functional organization during cognitive development. Current Opinion in Neurobiology, 15, 239–244. 10.1016/j.conb.2005.03.012 [DOI] [PubMed] [Google Scholar]
  8. Casey BJ, Jones RM, & Hare TA (2008). The adolescent brain. Annals of the New York Academy of Sciences, 1124, 111–126. 10.1196/annals.1440.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Centers for Disease Control and Prevention. (2015). Unintended pregnancy prevention. Retrieved May 14, 2017, from https://www.cdc.gov/reproductivehealth/unintendedpregnancy/
  10. Centers for Disease Control and Prevention. (2016). CDC fact sheet: Information for teens and young adults: Staying healthy and preventing STDs. Retrieved May 14, 2017, from https://www.cdc.gov/std/life-stages-populations/stdfact-teens.htm
  11. Centers for Disease Control and Prevention. (2017). Teen pregnancy in the United States. Retrieved May 16, 2017, from https://www.cdc.gov/teenpregnancy/about/index.htm
  12. Cohen J (1988). Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum. [Google Scholar]
  13. Derefinko KJ, Peters JR, Eisenlohr-Moul TA, Walsh EC, Adams ZW, & Lynam DR (2014). Relations between trait impulsivity, behavioral impulsivity, physiological arousal, and risky sexual behavior among young men. Archives of Sexual Behavior, 43, 1149–1158. 10.1007/s10508-014-0327-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Erickson KI, Creswell JD, Verstynen TD, & Gianaros PJ (2014). Health neuroscience: Defining a new field. Current Directions in Psychological Science, 23, 446–453. 10.1177/0963721414549350 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Falconer E, Allen A, Felmingham KL, Williams LM, & Bryant RA (2013). Inhibitory neural activity predicts response to cognitive-behavioral therapy for posttraumatic stress disorder. Journal of Clinical Psychiatry, 74, 895–901. 10.4088/JCP.12m08020 [DOI] [PubMed] [Google Scholar]
  16. Feldstein Ewing SW, Houck JM, & Bryan AD (2015). Neural activation during response inhibition is associated with adolescents’ frequency of risky sex and substance use. Addictive Behaviors, 44, 80–87. 10.1016/j.addbeh.2014.12.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Feldstein Ewing SW, Hudson KA, Caouette J, Mayer AD, Thayer RE, Ryman SG, & Bryan AD (in press). Sexual risk-taking and subcortical brain volume in adolescence. Annals of Behavioral Medicine. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Feldstein Ewing SW, Ryman SG, Gillman AS, Weiland BJ, Thayer RE, & Bryan AD (2016a). Developmental cognitive neuroscience of adolescent sexual risk and alcohol use. AIDS and Behavior, 20, 97–108. 10.1007/s10461-015-1155-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Feldstein Ewing SW, Tapert SF, & Molina BSG (2016b). Uniting adolescent neuroimaging and treatment research:Recommendations in pursuit of improved integration. Neuroscience and Biobehavioral Reviews, 62, 109–114. 10.1016/j.neubiorev.2015.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Finer LB, & Zolna MR (2014). Shifts in intended and unintended pregnancies in the United States, 2001–2008. American Journal of Public Health, 104 (Suppl. 1), S43–S48. 10.2105/ajph.2013.301416 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Friston KJ, Frith CD, Turner R, & Frackowiak RS (1995). Characterizing evoked hemodynamics with fMRI. Neuroimage, 2(2), 157–165. 10.1006/nimg.1995.1018 [DOI] [PubMed] [Google Scholar]
  22. Goldenberg D, Telzer EH, Lieberman MD, Fuligni A, & Galván A (2013). Neural mechanisms of impulse control in sexually risky adolescents. Developmental Cognitive Neuroscience, 6, 23–29. 10.1016/j.dcn.2013.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Golub SA, Starks TJ, Kowalczyk WJ, Thompson LI, & Parsons JT (2012). Profiles of executive functioning: Associations with substance dependence and risky sexual behavior. Psychology of Addictive Behaviors : Journal of the Society of Psychologists in Addictive Behaviors, 26, 895–905. 10.1037/a0029034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Grill-Spector K, & Malach R (2004). The human visual cortex. Annual Review of Neuroscience, 27, 649–677. 10.1146/annurev.neuro.27.070203.144220 [DOI] [PubMed] [Google Scholar]
  25. Hatchard T, Fried PA, Hogan MJ, Cameron I, & Smith AM (2014). Marijuana use impacts cognitive interference: An fMRI investigation in young adults performing the counting stroop task. Journal of Addiction Research and Therapy, 5, 1–7. 10.4172/2155-6105.100019725927012 [DOI] [Google Scholar]
  26. Hatchard T, Smith AM, Halchuk RE, Longo CA, Fried PA, Hogan MJ, & Cameron I (2015). Effects of low-level alcohol use on cognitive interference: An fMRI study in young adults. Alcohol, 49, 7–13. 10.1016/j.alcohol.2014.07.020 [DOI] [PubMed] [Google Scholar]
  27. Horn NR, Dolan M, Elliott R, Deakin JFW, & Woodruff PWR (2003). Response inhibition and impulsivity: An fMRI study. Neuropsychologia, 41, 1959–1966. 10.1016/S0028-3932(03)00077-0 [DOI] [PubMed] [Google Scholar]
  28. Houben K, Nederkoorn C, Wiers RW, & Jansen A (2011). Resisting temptation: Decreasing alcohol-related affect and drinking behavior by training response inhibition. Drug and Alcohol Dependence, 116, 132–136. 10.1016/j.drugalcdep.2010.12.011 [DOI] [PubMed] [Google Scholar]
  29. Hoyle RH, Feifar MC, & Miller JD (2000). Personality and sexual risk taking: A quantitative review. Journal of Personality, 68, 1203–1231. 10.1111/1467-6494.00132 [DOI] [PubMed] [Google Scholar]
  30. Kerestes R, Ladouceur CD, Meda S, Nathan PJ, Blumberg HP, Maloney K, … Phillips ML (2012). Abnormal prefrontal activity subserving attentional control of emotion in remitted depressed patients during a working memory task with emotional distracters. Psychological Medicine, 42, 29–40. 10.1017/s0033291711001097 [DOI] [PubMed] [Google Scholar]
  31. Kübler A, Dixon V, & Garavan H (2006). Automaticity and reestablishment of executive control: An fMRI study. Journal of Cognitive Neuroscience, 18, 1331–1342. 10.1162/jocn.2006.18.8.1331 [DOI] [PubMed] [Google Scholar]
  32. Macapagal KR, Janssen E, Fridberg DJ, Finn PR, & Heiman JR (2011). The effects of impulsivity, sexual arousability, and abstract intellectual ability on men’s and women’s go/no-go task performance. Archives of Sexual Behavior, 40, 995–1006. 10.1007/s10508-010-9676-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Martin JA, Hamilton BE, Osterman MJKS, Driscoll AK, & Mathews TJ (2017). National vital statistics reports, 66(1). Atlanta, GA: Centers for Disease Control and Prevention; Retrieved from http://www.cdc.gov/nchs/data/nvsr/nvsr66/nvsr66_01.pdf [PubMed] [Google Scholar]
  34. Messman-Moore TL, Walsh KL, & DiLillo D (2010). Emotion dysregulation and risky sexual behavior in revictimization. Child Abuse and Neglect, 34, 967–976. 10.1016/j.chiabu.2010.06.004 [DOI] [PubMed] [Google Scholar]
  35. Mitchell RLC (2005). The BOLD response during Stroop task-like inhibition paradigms: Effects of task difficulty and task-relevant modality. Brain and Cognition, 59, 23–37. 10.1016/j.bandc.2005.04.001 [DOI] [PubMed] [Google Scholar]
  36. Nigg JT, Wong MM, Martel MM, Jester JM, Puttler LI, Glass JM, … Zucker RA (2006). Poor response inhibition as a predictor of problem drinking and illicit drug use in adolescents at risk for alcoholism and other substance use disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 45, 468–475. 10.1097/01.chi.0000199028.76452.a9 [DOI] [PubMed] [Google Scholar]
  37. Nydegger LA, Ames SL, Stacy AW, & Grenard JL (2014). Response inhibition moderates the association between drug use and risky sexual behavior. Substance Use and Misuse, 49, 1457–1464. 10.3109/10826084.2014.912230 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Parsons JT, Halkitis PN, Bimbi D, & Borkowski T (2000). Perceptions of the benefits and costs associated with condom use and unprotected sex among late adolescent college students. Journal of Adolescence, 23, 377–391. 10.1006/jado.2000.0326 [DOI] [PubMed] [Google Scholar]
  39. Reyna VF, & Farley F (2006). Risk and rationality in adolescent decision making: Implications for theory, practice, and public policy. Psychological Science in the Public Interest : A Journal of the American Psychological Society, 7, 1–44. 10.1111/j.1529-1006.2006.00026.x [DOI] [PubMed] [Google Scholar]
  40. Ridderinkhof KR, van den Wildenberg WPM, Segalowitz SJ, & Carter CS (2004). Neurocognitive mechanisms of cognitive control: The role of prefrontal cortex in action selection, response inhibition, performance monitoring, and reward-based learning. Brain and Cognition, 56, 129–140. 10.1016/j.bandc.2004.09.016 [DOI] [PubMed] [Google Scholar]
  41. Robbins RN, & Bryan A (2004). Relationships between future orientation, impulsive sensation seeking, and risk behavior among adjudicated adolescents. Journal of Adolescent Research, 19, 428–445. 10.1177/0743558403258860 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Schmiege SJ, Broaddus MR, Levin M, & Bryan AD (2009). Randomized trial of group interventions to reduce HIV/STD risk and change theoretical mediators among detained adolescents. Journal of Consulting and Clinical Psychology, 77, 38–50. 10.1037/a0014513 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sedgh G, Finer LB, Bankole A, Eilers MA, & Singh S (2015). Adolescent pregnancy, birth, and abortion rates across countries: Levels and recent trends. The Journal of Adolescent Health, 56, 223–230. 10.1016/j.jadohealth.2014.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Skeem JL, Scott E, & Mulvey EP (2014). Justice policy reform for high-risk juveniles: Using science to achieve large-scale crime reduction. Annual Review of Clinical Psychology, 10, 709–739. 10.1146/annurev-clinpsy-032813-153707 [DOI] [PubMed] [Google Scholar]
  45. Spear LP (2000). The adolescent brain and age-related behavioral manifestations. Neuroscience and Biobehavioral Reviews, 24, 417–463. 10.1016/S0149-7634(00)00014-2 [DOI] [PubMed] [Google Scholar]
  46. Tanaka S, Honda M, & Sadato N (2005). Modality-specific cognitive function of medial and lateral human Brodmann area 6. The Journal of Neuroscience, 25, 496–501. 10.1523/JNEUROSCI.4324-04.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Thayer RE, Feldstein Ewing SW, Dodd AB, Hansen NS, Mayer AR, Ling JM, & Bryan AD (2015). Functional activation during the Stroop is associated with recent alcohol but not marijuana use among high-risk youth. Psychiatry Research - Neuroimaging, 234, 130–136. 10.1016/j.pscychresns.2015.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zapolski TCB, Cyders MA, & Smith GT (2009). Positive urgency predicts illegal drug use and risky sexual behavior. Psychology of Addictive Behaviors, 23, 348–354. 10.1037/a0014684 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES