statistics work need completed in 10 hours

Sleep Quantity, Family Income and Depression Outcomes

Student Name

Course Title Semester

Abstract

The current study’s objective was to test for the significance of the relationship between sleep quantity and mental health whilst controlling for family income. The sleep quantity, measured as the absolute deviation of sleep hours from the optimal sleep time (6 hours), and depression data were retrieved from the 2015 to 2016 NHANES survey. A simple regression model with sleep quantity as the sole independent variable was estimated prior to the estimation of the adjusted model with family income as a covariate. Both the adjusted and unadjusted beta coefficients for sleep quantity were insignificant.

Introduction

Sleep quantity has been shown to be a significant predictor of mental health outcomes in subpopulations that are homogenous in at least one respect for instance, in the population of elderly primary care patients (Reid et al., 2005). The relationship between sleep quantity and health outcomes arises from the role of sleep as a hormone release modulator making; hence, the quantity of sleep influence obesity, cardiovascular diseases, Type 2 diabetes, and hypertension outcomes (Lemola et al., 2009). This study aims at evaluating the nature and strength of the relationship between sleep quantity and depression scores in the general population. The study hypothesizes that while the quantity of sleep will be significantly correlated with the depression scores, the former will have little explanatory power on the variation in the latter’s scores since depression as an outcome arises from a multitude of variables other than sleep quantity. In order to develop an understanding of the unadjusted effect of sleep quantity on depression scores, it is necessary to control for potential confounders. Researchers have established a relationship between depression and income (Zimmerman & Katon, 2005) with individuals under pressure to make ends meet using meagre income exhibiting more depression symptoms. This study will, therefore, include family income as a covariate in order to assess the effect of sleep quantity while controlling for family income. In so doing, the effect of family income on depression is eliminated and the adjusted effect of sleep quantity of depression can be analyzed.

Methods

The current study is based on secondary data collected in the NHANES 2015 to 2018 survey using the mental health-depression screener, sleep disorders, and income questionnaires. The sleep disorder questionnaire was administered by trained interviewers in the home setting to participants aged 16 and above. The depression screener questionnaire was, on the other hand, administered to respondents Mobile Examination Center (MEC) participants above the age of 12 (CDC, 2017a). The responses provided by individuals aged below 18 years were, however, excluded from the dataset for sensitivity reasons. The depression screener questionnaire items are based on a validated depression screening instrument known as the Patient Health Questionnaire (CDC, 2017a; Kroenke & Spitzer, 2002). The PHQ collects information on the depression symptoms that patients exhibited over the past 2-week period (Kroenke & Spitzer, 2002). Participants’ depression scores were computed as the sum of the scores on the questionnaire items, each of which was positively coded. The scores on the nine questionnaire items ranged from 0; which corresponded to not at all, to 3; which corresponded to nearly every day. Subsequently, the possible depression scores range from 0 to 30.

Sleep quantity data was collected using NHANES sleep disorders questionnaire. The data was collected using a 7-item sleep disorder questionnaire that collected information on respondents’ sleep habits and disorders (CDC, 2017b). The data was collected from respondents aged 16 and higher. In terms of the optimal sleep quantity, there is evidence that both the lack of sleep and excessive sleep are associated with poor mental health outcomes (Kroenke & Spitzer, 2002). Research has also shown that adults who received 6 to 8 hours of sleep every night had a more optimistic outlook and were more satisfied with life compared to their counterparts who slept for less than six hours (Lemola et al., 2009). In response to the aforementioned factors, the sleep quantity variable was computed as a score that measures the absolute deviation of sleep hours from the arbitrarily selected optimal time of 6 hours. The formula: Sleep Quantity= Abs (Sleep Hours -6) was used. The ABS function ensures that both excessive sleep and the lack of sleep receive the same treatment in the regression analysis.

Monthly family income data was collected using NHANES’ income questionnaire (INQ). The variable measures the total income flowing to the household from all its economic activities within a range of values so as to minimize disclosure risks. The categories range from 1 to 12 with 1 representing the lowest monthly income ($0-$399) while 12 corresponds with the highest monthly income ($8,400 and over). The data will be analyzed using weighted and unweighted descriptive statistics, linear regression, and the general linear model in SPSS. A simple regression model with sleep quantity as the sole independent variable will be estimated prior to the estimation of the adjusted model with family income as a covariate.

Results

The descriptive statistics and frequency tables are presented in Tables 1 to 3 in the appendix. The average depression score was 5.05 (SD= 4.94) with a maximum of 25 and minimum of 0. Based on the PHQ cut-off scores, a total score ranging between 5 and 9 is suggestive of mild depression symptoms. The valid sample size used in the determination of unweighted sleep quantity and depression means was 3,211. Sleep Quantity had an unweighted mean of 1.18 (SD = 1.06) indicating that individuals sleep quantities deviated from the optimal value of 6 by approximately 1.184. The weighted mean sleep quantity (M= 1.176, SD = 1.05) and depression (M = 5.03, SD = 5.03), were more or less similar to the unweighted means. The population parameters and sample statistics are almost similar; i.e., no group is significantly overrepresented/ underrepresented in the sample. From Table 3, the distribution of responses across the 12 income categories was significantly spread out. A majority (12.5%) of the respondents reported monthly income values that ranged from $2,900 to $3,749 while an almost equally large proportion (12.1%) reported monthly family income of $8,400 and above. Of the 9,971 responses, 15.3% were invalid.

From the simple regression tables in the appendix, sleep quantity was an insignificant predictor of depression scores at an alpha level of 0.05 (F(1,3559) = 3.384, p = 0.066). From the adjusted R Squared value of 0.001, the model had low explanatory power. The variation in sleep quantity accounted for only 0.1% of the variation in the depression scores. The beta coefficient indicating the change in depression score arising from a unit change in sleep quantity was negative albeit insignificant (B = -0.143, p = 0.066); an increase in the deviation score was associated with a decrease in depression scores. The negative relationship was unexpected and might have resulted from the use of an absolute deviation score. The relationship between depression and sleep quantity when controlling for family income was analyzed using the Analysis of Covariance via the General Linear Model. After controlling for family income, the effect of sleep quantity on depression remained insignificant at an alpha level of 0.05 (F(14,3018)= 1.13, p = 0.324). Both the family income (F = 0.245, p = 0.620) and sleep quantity (F = 1.2, p = 0.272) variables were insignificant predictors of depression scores.

Discussion

A simple regression model with sleep quantity as the sole independent variable was estimated prior to the estimation of the adjusted model with family income as a covariate. The unadjusted coefficient of sleep quantity from the simple regression model was insignificant. The data, therefore, disproves the hypothesis that sleep quantity and depression scores are correlated. Given the insignificance of the relationship between sleep quantity and depression without controlling for family income as a covariate, it was unlikely that a significant relationship would be established in the regression model with a covariate. The ANCOVA model results were as expected: the adjusted coefficient of sleep quantity was insignificant. The coefficient of family income was also insignificant indicating the absence of a relationship between family income and depression scores.

The findings of the current study differ from the findings of Augner (2011) and Zimmerman and Katon (2005) where sleep quantity and family income significantly influenced the depression outcome. The difference in findings could stem from the method of data collection as well as the scoring of the variables. Augner (2011) and Zimmerman and Katon (2005) collected primary data using validated instruments. The data used by the researchers was therefore, more reliable than the data used in the current study. Additionally, Zimmerman and Katon (2005) used continuous rather than categorical income data making the linear regression model an appropriate analysis tool. The use of an absolute deviation score presented a shortcoming in this study as the negative sleep quantity coefficient suggests that the effect of excessive sleep and lack of sleep on depression scores might be dissimilar. Future studies should, therefore, contrast the effect of excessive sleep and lack of sleep on mental health outcomes. This can be done by formulating two separate linear regression models in the excessive sleep and lack of sleep populations.

References

Augner, C. (2011). Associations of subjective sleep quality with depression score, anxiety,

physical symptoms and sleep onset latency in students. Central European journal of public health, 19(2), 115.

CDC (2017a). National Health and Nutrition Examination Survey 2015-2016 data

documentation, codebook, and frequencies: Mental Health- Depression Screener (DPQ_I) Retrieved from: https://wwwn.cdc.gov/Nchs/Nhanes/2015-2016/DPQ_I.htm

CDC (2017b). National Health and Nutrition Examination Survey 2015-2016 data

documentation, codebook, and frequencies: Sleep Disorders (SLQ_I). Retrieved from: https://wwwn.cdc.gov/Nchs/Nhanes/2015-2016/SLQ_I.htm

CDC (2017c). National Health and Nutrition Examination Survey 2015-2016 data

documentation, codebook, and frequencies: Income (INQ_I). Retrieved from: https://wwwn.cdc.gov/Nchs/Nhanes/2015-2016/INQ_I.htm

Kroenke K, Spitzer RL. (2009) The PHQ-9: a new depression and diagnostic severity measure.

Psych Annals 32: 509-21.

Lemola, S., Räikkönen, K., Matthews, K. A., Scheier, M. F., Heinonen, K., Pesonen, A. K., ... &

Lahti, J. (2010). A new measure for dispositional optimism and pessimism in young children. European Journal of Personality: Published for the European Association of Personality Psychology, 24(1), 71-84.

Reid, K. J., Martinovich, Z., Finkel, S., Statsinger, J., Golden, R., Harter, K., & Zee, P. C.

(2006). Sleep: a marker of physical and mental health in the elderly. The American journal of geriatric psychiatry, 14(10), 860-866.

Zimmerman, F. J., & Katon, W. (2005). Socioeconomic status, depression disparities, and

financial strain: what lies behind the income‐depression relationship?. Health economics, 14(12), 1197-1215.

Appendix

Table 1

Unweighted Descriptive Statistics

Table 2

Weighted Descriptive Statistics

Table 3

Family Income Frequency Table

SIMPLE REGRESSION TABLES

SCATTER PLOT

GENERAL LINEAR MODEL TABLES

Univariate Analysis of Variance