Fatigue, anxiety and depression in patients with prediabetes: a controlled cross-sectional study

Abstract

Objective

We aimed to determine fatigue, depression, anxiety levels, and quality of life (QoL) in patients with prediabetes (PD) and to compare them with healthy subjects.

Materials and methods

A controlled, cross-sectional study was conducted. The patients, aged 18–65, were recruited from a tertiary care hospital. A total of 105 patients with newly diagnosed untreated PD and 48 normoglycemic subjects were included in the study. Participants did not know the diagnosis of PD and did not give psychological distress to the newly diagnosed. All participants were evaluated for body mass index (BMI), oral glucose tolerance test (OGTT), glycated hemoglobin (HbA1c), Fatigue Severity Scale (FSS), Hospital Anxiety and Depression Scale (HADS), Short Form-36 (SF-36).

Results

Age, gender, BMI were similar between groups. The scores of psychological measures were significantly worse in patients with PD compared with normoglycemic subjects (FSS median score: 4.33 vs. 2.22, HADS-anxiety mean score: 7.53 vs. 2.64, HADS-depression mean score: 6.33 vs. 2.79, SF-36 total median score: 52.81 vs. 79.89). The FSS, HADS and SF-36 scores showed a weak but statistically significant relationship with BMI. OGTT, HbA1c and HOMA-IR did not show any relationship with these scores.

Conclusion

Psychosocial problems may present in the prediabetic stage before the onset of diabetes.

Introduction

Prediabetes (PD) is characterized by elevated plasma glucose levels, which is under the limits in diabetes. It may also cause microvasculary and/or macrovasculary complications [1]. Although PD has no specific symptoms, the potential loss of physical health may be caused by risks for diabetic complications [1, 2]. PD, which prevalence is increasing day by day, is a serious risk factor for cardiovascular and renal diseases [1]. PD is a window of opportunity to prevent patients from the development of type 2 diabetes mellitus (T2DM) and its complications [2]. Recently, because of becoming the target point of struggle with T2DM, PD have been of great importance. In this context, the number of studies are increasing in prediabetic patients. PD is regarded as one of the illnesses, caused by psychosocial conditions, like diabetes and other chronic diseases [1–3].

Health is defined as “not only the absence of illness and disability but a state of complete physical, mental and social well-being” by the World Health Organization. While life expectancy increases, due to better living conditions and improvements in health, improvements in quality of life also gain momentum [4]. When planning treatments for diseases, the issue of improving the quality of life is considered more sensitive. In this respect, it is of particular interest that many chronic diseases affect the quality of life and the search for solutions [4, 5]. The identification of health problems with increasing prevalence in prediabetics is a part of the struggle with T2DM. Although lifestyle change is the main preventive measure for diabetes, conditions such as fatigue and anxiety remain a serious obstacle to this change. [6, 7]. Given the interchangeable state of depression, lifestyle change and the triad of hyperglycaemia, the vicious cycle that occurs when one of these is ignored causes progressive illness [7, 8].

Despite many studies on fatigue and psychological distress in people with diabetes, there are comparatively scarce data on PD. The aim of this study was to determine fatigue, depression, anxiety levels, and quality of life in prediabetics and to compare them with normoglysemic subjects.

Methods

Design

A controlled, cross-sectional study.

Subject’s characteristics and laboratory data

People, 18–65 years old, who were admitted to recruit voluntary healthy subjects in the tertiary hospital’s internal medicine outpatient clinic for routine health control and newly diagnosed untreated patients with prediabetes according to American Diabetes Association were recruited to the study as a prediabetic group, and who had normal glucose values were recruited to the study as the control group, consecutively [1]. Participants’ age, gender, occupation were recorded. Study was conducted between December 2018 and April 2019. The participants who had a prior diagnosis of thyroid disorders, rheumatic, malignant or psychiatric diseases, fibromyalgia, or pregnancy were excluded, because of their effect on fatigue, depression, anxiety levels and quality of life [9–12]. Also, the patients did not have any announcement or educational programs including diet restriction or regular exercise. Subjects, who were unaware of their glycemic status when answering the questionnaires, took the test on the same day. They did not give psychological distress to newly diagnosed. They were evaluated for level of fatigue and psycological status. All participants were informed about their OGTT results afterwards. In addition, they should be cared for prevention for the progression of DM.

Height and weight were measured and BMI calculated. BMI was categorized as normal (BMI < 30 kg/m²), and obese (BMI: 30 kg/m² and above) [13].

Zero and 2nd hour plasma glucose values were conducted by oral glucose tolerance test (OGTT), and glycated hemoglobin (HbA1c) levels were measured for all participants. PD was defined as 0 h plasma glucose value (OGTT-0th) is 100–125 mg/dL and/or 2-h plasma glucose value (OGTT-2nd) is 140 mg/dL to 199 mg/dL. HbA1c value of 5.7–6.4% is also considered to be PD [1]. A fasting venous blood sample was collected after an overnight fast of at least 12 h for biochemical investigations and samples were processed at the hospital laboratory on the same day. Glucose levels were estimated using a Roche Cobas 8000 immunoassay analyzer (Roche Diagnostics, USA). The level of HbA1c were estimated using a Adams A1c HA-8180 V automatic analyzer (Arkray Diagnostics, USA). All assays were performed with specific kits and calibrators supplied by the manufacturers.

12 h fasting blood samples were obtained for fasting plasma insulin (FPI) and fasting plasma glucose (FPG) determinations to calculate the homeostasis model assessment of insulin resistance (HOMA-IR). It was determined by the formula [14]:

HOMA-IR = FPI (mU/L) × FPG (mmol/L)/22.5. If the result is ≥ 2.5, it means there is an insulin resistance. The higher the score, the greater the insulin resistance is measured.

Psychological data

Turkish validated Fatigue Severity Scale (FSS), is a self-reported questionnaire with 9 items each of them ranges from 1 (indicates strong disagreement) to 7 points (indicates strong agreement) in a Likert scale, was used to assess the severity of fatigue in different situations (motivation, exercise, physical functioning, carrying out duties, interfering with work, family, or social life) during the past week [15]. The final score was found by calculating the average of the total 9 items’ score. The higher scores indicate more severe fatigue. FSS final score 4 was interpreted as fatigue [16, 17].

The Hospital Anxiety and Depression Scale (HADS) is a self-reported questionnaire to evaluate and measure the risk of depression and/or anxiety. HADS is a reliable and validated psychometric scale, which includes 14 questions; that half of them assess anxiety and the others assess depression with four possible answers (score 0–3). According to the Turkish validation study of HADS, if scores ≥ 11 accepted as anxiety, and if ≥ 8 accepted as depression in the current study [18].

Short Form-36 (SF-36) is a valid, and reliable questionnaire to assess both physical and mental components of HRQoL [19, 20]. SF-36 contains 36 items associated with 8 dimensions: physical functioning for the limitation in performing all physical activities, role-physical for problems with work or other daily activities, bodily pain, general health, vitality, social functioning, role-emotional, and mental health [19]. Component analyses showed that there are two distinct concepts measured by the SF-36: a physical dimension, represented by the Physical Component Summary (PCS), and a mental dimension, represented by the Mental Component Summary (MCS). The arithmetic average of the two parameters is described as Total Component Summary (TCS). SF-36 is also a valid and reliable questionnaire for Turkish people [20].

The patient's written informed constent to publish the clinical information and materials was obtained. Local Ethical Committee approval was received.

Statistical analysis

Statistical analyses were performed with the SPSS software version 22.0 (IBM Corp., Armonk, NY, USA). Parametric variables were presented as means and standard deviations, non-parametric variables were presented as medians and interquartile ranges (25–75th percentiles). Shapiro-Wilks test and histograms analyses were used to determine whether continuous variables were normally distributed. Two independent groups of parametric variables were compared using Student’s t test. For non-parametric variables Mann–Whitney U test was administered. Categorical data were analyzed by Chi-square or Fisher’s exact test, where appropriate. Relationship between non-parametric variables were analyzed by Spearman correlation tests. Pearson's correlation tests were used for parametric variables. Number of cases and percentages were used for categorical variables. Binary logistic regression was used to estimate odds ratios (ORs) and 95% confidence interval via multivariate analyses. The association of psychological scores (FSS, HADS, and SF-36) with the presence of prediabetes was evaluated by logistic regression analysis via multivariate analyses. A p value of < 0.05 was considered to indicate statistically significant differences.

Results

A total of 105 prediabetics and 48 control group participants were enrolled to the study. The mean age of 153 subjects was 48.37 ± 10.19 years old. Whereas 111 (72.5%) of the subjects were male, and 42 (27.5%) were female. Also the mean BMI of all participants was 33.67 ± 7.81 kg/m². Age, gender, BMI, occupational, marital and educational status were similar between groups (mentioned in Table 1). OGTT-0th and OGTT-2nd hour values and HbA1c levels were higher in PD group (p < 0.001).

Table 1.

Comparison of demographic, clinical data, and self-reported outcome measures of prediabetics with the control group

	Control (n = 48)	Prediabetes (n = 105)	p value
Gender (F/M), n (%)	32/16 (66.7/33.3)	79/26 (75.2/24.8)	0.27
Age (year), mean (SD)	47.75 (10.70)	48.65 (9.98)	0.624
Occupation, n (%)			0.184
Housewife	26 (54.2)	74 (70.5)
Officer	6 (12.5)	11 (10.5)
Worker	9 (18.8)	9 (8.6)
Retired	7 (14.6)	11 (10.5)
Marital status			0.816
Married	45 (93.8)	97 (92.4)
Divorced	2 (5.9)	7 (6.7)
Single	1 (2.1)	1 (1.0)
Education			0.225
Illiteracy	0 (0)	4 (3.8)
Primary school	32 (66.7)	79 (75.3)
High school	5 (10.4)	12 (11.4)
University	11 (22.9)	10 (9.6)
BMI (kg/m2), mean (SD)	32.69 (8.61)	31.65 (7.11)	0.421
Obesity ( +)ve, n (%)	27 (57.4)	73 (69.5)	0.147
OGTT-0 (mg/dL), mean (SD)	91.20 (9.17)	105.68 (7.38)	< 0.001
OGTT-2 (mg/dL), mean (SD)	109.80 (19.74)	138.41 (29.59)	< 0.001
HbA1c (%), median (IQR)	5.6 (5.3–5.6)	6.1 (5.7–6.2)	< 0.001
HOMA-IR, median (IQR)	2.13 (1.37–3.37)	2.56 (1.60–3.77)	0.39
Fasting plasma insulin (mU/L) (IQR)	11.37 (6.58–15.23)	10.10 (6.20–15.75)	0.87
Insulin resistance ( +)ve, n (%)	20 (47.6)	47 (48.5)	0.928
Fatigue ( +)ve, n (%)	1 (2.1)	15 (14.3)	0.024
FSS, median (IQR)	2.22 (1.78–3.55)	4.33 (3.11–5.44)	< 0.001
HADS-anxiety
Score, mean (SD)	2.64 (3.84)	7.53 (5.98)	< 0.001
Anxiety ( +)ve, n (%)	1 (2.1)	26 (24.8)	0.001
HADS-depression
Score, mean (SD)	2.79 (3.92)	6.33 (5.71)	< 0.001
Depression ( +)ve, n (%)	6 (12.8)	37 (35.2)	0.004
SF-36 dimensions, median (IQR)
Physical functioning	100.00 (90.00–100.00)	80.00 (50.00–100.00)	< 0.001
Role-physical	100.00 (25.00–100.00)	0.00 (0.00–100.00)	0.013
Bodily pain	77.50 (57.50–90.00)	57.50 (38.75–78.75)	0.003
Social functioning	87.50 (75.00–100.00)	62.50 (37.50–82.50)	< 0.001
Mental health	68.00 (60.00–76.00)	60.00 (44.00–72.00)	0.008
Role-emotional	100.00 (66.66–100.00)	33.33 (0–100.00)	0.002
Vitality	75.00 (60.00–85.00)	60.00 (42.50–75.00)	< 0.001
General health	80.00 (55.00–90.00)	55.00 (30.00–80.00)	< 0.001
SF-36/PCS	86.25 (60.00–92.18)	56.25 (32.18–84.06)	< 0.001
SF-36/MCS	78.75 (69.95–86.00)	53.87 (35.50–78.62)	< 0.001
SF-36/TS	79.89 (65.13–89.84)	52.81 (35.81–80.92)	< 0.001

The statistically significant bold values are p-values < 0.05

F female; M male; BMI body mass index; OGTT oral glucose tolerance test; HbA1c glycated hemoglobin; HOMA-IR homeostasis model assessment of insulin resistance; SF-36 short form-36; PCS physical component scale; MCS mental component scale; TS total component scale; HADS hospital anxiety and depression scale; FSS fatigue severity scale

Fatigue was diagnosed in 15 of 105 patients with PD (14.1%) and in 1 of 48 control group participants (2.1%) (p = 0.024). Level of fatigue, which was assessed by FSS, was worse in the PD group than the control group (p < 0.001). More frequent anxiety and depression risks were found in PD group also (p < 0.001). HRQoL, in all parameters of SF-36, were poorer in the PD group (p < 0.001). The scores of psychological measures were significantly worse in patients with PD compared with normoglycemic subjects (FSS median score: 4.33 vs. 2.22, HADS-anxiety mean score: 7.53 vs. 2.64, HADS-depression mean score: 6.33 vs. 2.79, SF-36 total median score: 52.81 vs. 79.89). Comparison of the level of fatigue, risks of anxiety and depression, and HRQoL parameters among prediabetic and control group were summarized in Table 1.

Age and glucose tolerance were not associated with any psychological factors, but BMI showed a statistically positive correlation with scores of FSS and HADS/anxiety (Summarized in Table 2). Further, whereas age and glucose tolerance were not associated with any SF-36 scores, BMI showed a statistically negative correlation with all SF-36 scores (Summarized in Table 3).

Table 2.

Correlation analysis of HADS and FSS scores

	HADS/anxiety		HADS/depression		FSS
	Rho	p	Rho	p	Rho	p
Age	–0.135	0.168	–0.110	0.265	0.057	0.562
BMI (kg/m2)	0.225	0.021	0.191	0.051	0.252	0.010
OGTT-0 (mg/dL)	0.112	0.257	0.019	0.845	0.069	0.484
OGTT-2 (mg/dL)	− 0.005	0.961	0.003	0.979	0.045	0.645
HbA1c (%)	0.047	0.638	0.050	0.614	0.121	0.222
HOMA-IR	0.148	0.147	0.111	0.278	0.060	0.556

The statistically significant bold values are p-values < 0.05

BMI body mass index; OGTT oral glucose tolerance test; HbA1c glycated hemoglobin; HOMA-IR homeostasis model assessment of insulin resistance; HADS hospital anxiety and depression scale; FSS fatigue severity scale

Table 3.

Correlation analysis of SF-36 scores

	SF-36/PCS		SF-36/MCS		SF-36/TCS
	Rho	p	Rho	p	Rho	p
Age	0.052	0.600	0.015	0.879	0.046	0.643
BMI (kg/m2)	− 0.370	< 0.001	− 0.255	0.009	− 0.330	0.001
OGTT-0 (mg/dL)	− 0.011	0.910	− 0.081	0.414	− 0.059	0.550
OGTT-2 (mg/dL)	− 0.028	0.780	− 0.067	0.496	− 0.046	0.642
HbA1c (%)	− 0.146	0.139	− 0.138	0.163	− 0.142	0.150
HOMA-IR	− 0.117	0.253	− 0.153	0.135	− 0.140	0.172

The statistically significant bold values are p-values < 0.05

BMI body mass index; OGTT oral glucose tolerance test; HbA1c glycated hemoglobin; HOMA-IR homeostasis model assessment of insulin resistance; SF-36 short form-36; PCS physical component scale; MCS mental component scale; TS total component scale

The presence of prediabetes in all participants is evaluated by logistic regression analysis via multivariate analyses. Psychological problems have been found to not be effective factors in the presence of prediabetes. Presence of fatigue with an odds ratio of 0.023 (95% CI 0.000–36.460; p = 0.315), HADS/anxiety with an odds ratio of 0.893 (95% CI 0.496–1.608; p = 0.706) and HADS/depression with an odds ratio of 0.919 (95% CI 0.538–1.571; p = 0.758) was determined.

Discussion

This study revealed that severer fatigue, anxiety, depression and lower QOL levels in patients with PD compared with the normoglycemic subjects who were not statistically different matched in mean age, gender, BMI and sociodemographic status. Further, the results of multivariate analyses showed that there seemed weak but statistically positive correlations between several psychological problems and BMI.

To the best of our knowledge, fatigue levels in newly diagnosed untreated patients with prediabetes were compared firstly in the literature, whereas there were many studies in diabetics on fatigue assessment [21–24]. Senefeld et al. reported that mean Fatigue Impact Scale scores were 12.5 and 12.6 in PD and normoglycemic control, which was contrary to our results that fatigue may be was significantly increased in PD compared to normoglycemic control [25]. This difference may also be related to a difference in age group, the newly diagnosed, and the larger population of our study. Previous study reported that fatigue was associated with low QOL and BMI was independently related to fatigue level in T2DM [23], which is consistent with our results in PD. Other previous study also reported that fatigue rate increased as BMI increases, but no relationship was found between fatigue and HbA1c [24]. In our study, which we evaluated fatigue with FSS, it was shown that fatigue via both scales increased significantly in prediabetic patients, compared with the control group. In our study, we found a positive correlation among BMI and fatigue, but not found any other relationship between fatigue and age, HbA1c, serum glucose, HOMA-IR. It is already known that fatigue causes weight gain with immobilization, and overweight people also have higher fatigue. However, the mean BMI (kg/m2 a) was 32.7 and 31.7 in PD and normoglycemic controls in this study. Compared to the control group who had similar BMI, fatigue severity was higher in the PD group. These results in people with glycemic abnormality may suggest the possibility that a lifestyle to gain weight may have related to increasing fatigue and decreasing physical and mental QOL. On the other hand, several diabetic complications [24], glycemic variability [21, 22], and excess cytokines inflammations [26, 27] were reported to result in fatigue and depression in DM. Although glycemic variability and cytokines were not measured in this study, these molecular physiological changes may start in PD and these physiological factors may also contribute to the results in higher fatigue level in PD.

In our study, we found that there were significant increases in anxiety and depression risks using HADS and a significant decrease in HRQoL using SF36. Despite of many studies evaluates psychological status in T2DM, there are scarce data in PD. However, a previous study reported that PD with elevated depression and anxiety showed an increased risk of developing diabetes [6]. In addition, a recent metareview suggested that depression is associated with not only the development of diabetes but also increasing risks of vascular complications, low HRQoL, and mortality in diabetes [28]. Our study showed decreased HRQoL as well as increased fatigue, depression, and anxiety. It could not be denied that lowered quality of life and physical disability, lack of exercise, and difficulty in life became a psychological burden in this study, but considering previous results assessing the association between psychological problems and HRQoL [28], increasing depression and anxiety may cause lower HRQoL in PD than normoglycemic subjects. Whereas the current literature has reported consistently that T2DM is associated with poorer HRQoL.

There are certain limitations to our study. First, patients diagnosed with diabetes mellitus were not included in the study. Because the main essence of the study is only to compare prediabetics with normoglycemics. Second, the subjects in our study were not in the form of community screening but were recruited from people who applied to the hospital. Therefore, the findings may not be generalizable to the whole population. Third, the association between psychological problem and glucose tolerance from bidirectional angles could not be evaluated enough due to the nature of the study, which is not a prospective but a cross-sectional study. However, our study has strengths. First, the diagnosis of prediabetes was made using the OGTT. Second, patients were included in the study without knowing the OGTT result so as not to mislead psychological tests. Finally, the relationship between the presence of prediabetes and fatigue parameters has been shown for the first time in the literature thanks to our work.

Conclusion

The psychosocial problems including fatigue, anxiety and depression and the decreased quality of life may present in the prediabetic stage before the onset of diabetes. Further studies are needed on the social-psychological burden and the onset of diabetes.

Funding

No financial support for the research and/or authorship of this paper.

Declarations

Conflict of interest

All of the authors declare that they have no conflict of interest.

Ethical approval

The approval for this study was obtained from Erciyes University Local Ethics Committee (Approval Number: 2019/141, Date: 20.02.2019). This trial was performed in accordance with the Declaration of Helsinki and Good Clinical Practice. All study participants gave written consent prior to any trial-related activities, and the investigator retained the consent forms. All patients were informed about the study protocol and gave their written informed consents.