The association between quality and quantity of carbohydrate with sleep, mood, anxiety, depression and stress among elderly

Abstract

Background

While previous studies have explored the relationship between carbohydrate intake, sleep quality, and mental health, further research is needed to better understand these associations. Our study is specifically examining carbohydrate quality and quantity in relation to sleep, mood, anxiety, stress, and depression to provide deeper insights into these complex interactions. Notably, there is a lack of previous studies focusing on these associations in the elderly population, highlighting the importance of our research.

Objectives

This study aimed to investigate the association between the quality and quantity of carbohydrate with sleep quality and mental health outcomes, including mood, anxiety, stress, and depression.

Methods

This cross-sectional study involved 398 Iranian elderly individuals aged 65–85 years, recruited from healthcare centers of Tehran University of Medical Sciences, Iran using simple random sampling. Dietary intake was assessed using a validated 168-item Food Frequency Questionnaire (FFQ). Psychological health was evaluated using the Depression, Anxiety, and Stress Scale (DASS-21). Sleep quality was measured with the Pittsburgh Sleep Quality Index (PSQI). Mood was assessed using the Profile of Mood States (POMS) questionnaire. Statistical analysis included logistic regression to examine associations between carbohydrate quality and quantity and sleep, mood, anxiety, stress, and depression.

Results

After adjusting for confounders, higher carbohydrate quantity was associated with increased odds of depression, with marginally significant higher adjusted odds ratios observed in the highest tertile (T3) compared to the lowest tertile (T1) (OR: 3.930, 95% CI: 1.2–12.867, P-trend = 0.04).Also, higher carbohydrate quality, assessed by fiber content, was associated with reduced odds of anxiety, and stress. (OR: 0.117, 95% CI: 0.075–0.419, P-trend < 0.001, OR: 0.307, 95% CI: 0.129–0.733, P-trend = 0.014) respectively. The glycemic load also showed a significant association with sleep quality (OR: 0.159, 95% CI: 0.054–0.468, P-trend < 0.001).

Conclusions

Our study demonstrates a strong link between carbohydrate quantity and quality (based on fiber content and glycemic load) with sleep and mental health outcomes. Higher carbohydrate quality intake as assessed by fiber content was associated with lower odds of anxiety and stress. Additionally, high carbohydrate quality intake as assessed by glycemic load was linked to improved sleep quality. These findings highlight the benefits of prioritizing high-quality carbohydrates in the diet for mental and sleep health, especially in the elderly population.

Introduction

With increasing age, individuals experience various physiological and psychological changes that may influence dietary needs and health outcomes [1]. Older adults are particularly vulnerable to mental health issues such as depression, anxiety, and sleep disturbances, which can significantly impact their quality of life and overall well-being [2]. Dietary factors, especially carbohydrate intake, have emerged as potential modifiable contributors to these outcomes [3]. Globally, an estimated 5% of adults suffer from depression, with anxiety disorders affecting approximately 4% of the population, translating to 301 million individuals worldwide [4]. Given the high prevalence of these mental health conditions, there is growing interest in understanding how dietary factors, such as carbohydrate consumption, influence mental and physical health [5–8]. While some studies have explored the effects of carbohydrates on mood and cognition [5, 7, 8], the relationship between carbohydrate intake and sleep, mood, depression, anxiety, and stress remain unclear, particularly in the elderly population. This paper investigates the association between the quality and quantity of carbohydrate intake and sleep quality, mood, depression, anxiety, and stress.

Some carbohydrate-rich foods have a smaller impact on increasing blood glucose than others [9]. The glycemic index (GI) measures how quickly a food raises blood sugar levels while the glycemic load (GL) considers the GI and the total amount of carbohydrates in a serving [10]. Research suggests that a diet rich in refined carbohydrates, high in starch and low in fiber, may have a negative impact on mental health [11] due to hyperglycemia and inflammation [3, 12].

Several studies suggest that a diet high in refined carbohydrates and added sugars is associated with an increased risk of depression [11]. Studies also show that high GI and sugar intake are associated with an increased risk of depression, while higher consumption of fiber, fruits, and vegetables is associated with a lower risk [6, 8, 13]. Some research suggests that high carbohydrate intake may also be associated with anxiety [7, 14].

The quality and quantity of carbohydrates may also play a role in sleep quality [15]. Emerging evidence suggests that both the type and amount of carbohydrates consumed can influence sleep duration, latency, and overall sleep quality. High-quality carbohydrates _such as whole grains, fruits, vegetables, and legumes_ have been associated with better sleep outcomes due to their favorable effects on glycemic response and nutrient content (e.g., fiber, vitamins, and tryptophan precursors) [16–19]. In contrast, low-quality carbohydrates including those high in added sugars andrefined grains [20] have been linked to disrupted sleep and increased risk of insomnia. For instance, a large prospective cohort study found that a diet high in added sugars and refined carbohydrates was associated with a higher risk of developing insomnia in postmenopausal women [18]. Another study demonstrated that meals with a high glycemic index reduced sleep latency when consumed 4 h before bedtime, suggesting a temporal and compositional effect of carbohydrates on sleep [21]. Furthermore, excessive carbohydrate intake, regardless of quality, may lead to fluctuations in blood glucose and insulin levels, potentially disturbing sleep continuity [22]. Conversely, moderate intake of complex carbohydrates may improve sleep by enhancing tryptophan availability, which supports melatonin and serotonin synthesis [23]. Thus, both the quality and quantity of carbohydrates are critical dietary factors that may influence sleep physiology and mental health.

While numerous studies have explored the relationship between carbohydrate intake and factors such as sleep, mood, anxiety, stress, and depression [7, 24], research specifically targeting the elderly population remains limited. The aging process is associated with unique metabolic, hormonal, and psychological changes that may influence how dietary carbohydrates impact these outcomes [25, 26]. As such, there is a critical need for targeted studies to better understand the role of carbohydrate quality and quantity in addressing these issues within this vulnerable demographic. Thus, this study aims to evaluate the association between carbohydrate intake—both in terms of quantity and quality—and mental health and sleep quality specifically in elderly individuals, a group for whom these outcomes are particularly prevalent and understudied.

Methods

This is a cross-sectional study on Iranian elderly people aged 65–85 years. Patients with sleep and mental health problems were recruited. A total of 398 participants were recruited using simple random sampling from healthcare centers affiliated with Tehran University of Medical Sciences between October 2021 and May 2022. Patients were recruited from among those who refer to hospitals or private clinics. Sleep problems are defined as issues with the quality, timing, and amount of sleep, which result in daytime distress and impairment in functioning. Mental health is defined as a state of mental well-being that enables people to cope with the stresses of life, realize their abilities, learn well and work well, and contribute to their community, including stress, anxiety, and depression. Eligibility criteria for participation included being aged 60 years or older (with the final included age range being 65–85 years), reporting no chronic illnesses, not using specific medications, and maintaining a usual diet without alterations due to disease or dietary recommendations. These criteria ensured the inclusion of relatively healthy elderly individuals to minimize confounding factors. Written informed consent was obtained from all subjects prior to participation.

Assessment of diet

Dietary intake data were collected using a validated semi-quantitative Food Frequency Questionnaire (FFQ) comprising 168 food items, designed to reflect the typical Iranian diet [27, 28]. trained interviewers conducted face-to-face interviews with participants, during which they reported the frequency and quantity of each food item consumed over the previous year. Standardized household sizes were used to aid accurate reporting, and participants could indicate consumption frequency as daily, weekly, monthly, or yearly. To estimate daily food intake, the reported frequencies were converted into average daily intake amounts. Portion sizes were translated into gram weights using household measurement conversion tables appropriate for Iranian foods [29]. The resulting daily intake (in grams per day) for each food item was then entered into Nutritionist IV software, a dietary analysis program that had been modified and updated to reflect the nutrient composition of Iranian foods [30, 31]. This software enabled the precise calculation of daily energy intake, as well as the intake of macronutrients (carbohydrates, proteins, fats) and a wide range of micronutrients (vitamins and minerals). All nutrient values were derived using the software’s comprehensive food composition database, ensuring standardization and reliability in the estimation process.

Estimation of fiber content

The fiber intake of participants was estimated using data obtained from a validated semi-quantitative 168-item Food Frequency Questionnaire (FFQ). During face-to-face interviews, participants reported the frequency and portion size of food items consumed over the previous year. These reports were converted into daily intake values (grams/day) using standard household measures. The daily intake data were then analyzed using Nutritionist IV software, which had been specifically modified to include the nutritional composition of Iranian foods, including dietary fiber content. The software provided fiber values for each food item, allowing for the estimation of total daily fiber intake for each participant.

Estimation of Glycemic Load (GL)

The glycemic load was calculated by integrating both the glycemic index (GI)nof each food item and the amount of carbohydrate consumed from that item. The formula used to calculate GL is:

$$\text{GL}=(\text{GI}\times \text{amount of carbohydrate per serving in grams})/100$$

Total dietary GL for each participant was computed by summing the GL values of all consumed food items. GI values were obtained from published tables relevant to the Iranian food composition or from international GI tables and were either incorporated into the modified Nutritionist IV software or used alongside it. This approach allowed for a culturally relevant and accurate estimation of the glycemic load of participants'diets, used as a key indicator of carbohydrate quality in the analysis.

Assessment of psychological health

We employed Depression, Anxiety, and Stress Scale − 21 Items (DASS-21) to assess symptoms of depression, anxiety, and stress. The DASS-21 consisted of separate sections for each condition, with seven items in each section [32]. Depressive symptoms were evaluated based on dysphoria, hopelessness, devaluation of life, self-deprecation, lack of interest, anhedonia, and inertia. The anxiety component covered autonomic arousal, skeletal muscular responses, and subjective feelings of anxiety. The stress component was sensitive to sustained non-specific arousal levels. In our study, we employed predefined cut-off scores to assess the severity of depression, anxiety, and stress. Specifically, a score of 8 or higher indicated anxiety, a score of 10 or higher indicated depression, and a score of 15 or higher was indicative of stress. These thresholds aligned with the recommendations from the DASS questionnaire and were utilized in prior research. The DASS-21 underwent reliability and validity assessments both locally and internationally and was available in 34 languages [33].

Assessment of sleep

At the beginning of the study, sleep quality was assessed using responses from the Pittsburgh Sleep Quality Index (PSQI) questionnaire [34]. The PSQI is a reliable and valid tool for identifying sleep disorders across various populations, including clinical and non-clinical settings [35]. The Persian version of the PSQI, which has been previously validated for use in Iranian populations, was used in this study. The psychometric properties of the Persian PSQI, including its reliability and validity, was previously confirmed [36].The PSQI evaluated sleep quality over the past month. It assessed various aspects of sleep quality over a one-month period through 19 items grouped into seven subscales: subjective sleep quality, sleep latency (time to fall asleep), sleep duration, habitual sleep efficiency (hours slept divided by hours in bed), sleep disturbances, use of sleep medications, and daytime dysfunction (difficulty staying awake). Each subscale received a score from 0 to 3, and the scores were combined to create a “global” score ranging from 0 to 21. Higher global scores indicated worse sleep quality. A global score of ≤ 5 suggested optimal sleep quality. Additionally, participants reporting poor sleep quality were further categorized based on their global PSQI score (> 5) as borderline (6–8) or poor (≥ 9) sleep quality [37]. Participants with a PSQI score of ≥ 5 were categorized as poor sleepers, while those with a score of < 5 were classified as good sleepers.

Assessment of mood

We utilized the Profile of Mood States (POMS) questionnaire to assess the severity of mood disorders in participants. The reliability and validity of this questionnaire had been previously evaluated in the Iranian population [38]. Widely recognized and commonly used in cognitive studies, the POMS long version consisted of 65 items. Participants were instructed to read each item and respond using a 5-point Likert scale ranging from 1 (not at all) to 5 (very high), based on their feelings during questionnaire completion.

The POMS covered six mood domains: Depression-Dejection, Vigor-Activity, Tension-Anxiety, Anger-Hostility, Confusion-Bewilderment, and Fatigue-Inertia. The Total Mood Disturbance (TMD) score was computed by adding the scores of the five negative domains and subtracting the vigor score. Higher scores indicated a greater degree of mood disorder. The overall mood scores ranged from − 28 to 134, with lower scores indicating a more positive mood. In the absence of an established cut-off for mood classification, the scores were divided into tertiles: scores above 40 represented a poor mood (third tertile), scores between 9 and 39 indicated an average mood (second tertile), and scores below 9 reflected a highly positive mood (first tertile).

Assessment of other variables

Participant height and body weight were assessed. Body weight was measured with participants barefoot and wearing lightweight clothing, using a Seca model digital scale with a precision of 0.1 kg. Height measurements were obtained as individuals stood without shoes, aligning their head, buttocks, and heels against the wall while looking straight ahead horizontally. A tape measure with an accuracy of 0.1 cm was used for height measurement. Waist circumference (WC) was measured at the midpoint between the lower rib margin and the iliac crest, using a non-stretchable measuring tape with a precision of 0.1 cm. Hip circumference was measured by placing a non-stretchable tape measure with a precision of 0.1 cm around the widest part of the hips.

Demographic details and lifestyle information were collected using structured questionnaires. Each participant was interviewed regarding their gender, age, marital status (single or married), educational attainment (below high school, high school, or higher), and income levels (≤ 20 million rials [≤ $476], 20–60 million rials [$476–$1,429], and ≥ 60 million rials [≥ $1,429]). Educational level was assessed based on the number of completed school years.

Lifestyle data included smoking habits, with participants asked about their smoking status (smoker or non-smoker). Physical activity levels among the elderly were evaluated using the International Physical Activity Questionnaire (IPAQ), which had undergone validation and reliability testing in 12 countries, including Iran, by the year 2000. The IPAQ consisted of seven questions assessing the frequency and duration of individuals'participation in various physical activities. Participants reported their activity levels over the past week in four categories: vigorous activity, moderate activity, walking, and sitting. To accurately calculate the Metabolic Equivalent of Task (MET) per minute, the duration of physical activities was expressed in minutes. The total weekly physical activity was determined by summing the MET per minute values for each individual across the week.

Statistical analysis

General characteristics of participants were assessed using ANOVA for continuous variables and chi-square tests for categorical variables. Dietary intakes across tertiles of carbohydrate quality as assessed by fiber content and glycemic load were assessed using ANCOVA. The association between carbohydrate quality (by fiber content/glycemic load), with sleep, mood, depression, anxiety, and stress was evaluated using conditional logistic regression in different models. In the first model, adjustments were made for age (continuous) and sex (male/female). In the second model, further adjustments included marital status (married/not married), SES (low/medium/high class), energy intake (kcal), omega-3 fatty acids (continuous), magnesium (continuous), cobalamin (continuous), total folate (continuous), pyridoxine (continuous), thyroid disease (yes/no), dyslipidemia (yes/no), and blood pressure (yes/no). The BMI was additionally adjusted in the third model. To determine the association, both crude and multivariable-adjusted models were constructed, controlling for the aforementioned covariates. All confounders were chosen based on previous publications. Given that multiple outcomes (n = 5) were tested for each exposure (fiber and glycemic load), we applied a Bonferroni correction to control for Type I error due to multiple comparisons. The corrected significance threshold was set at α = 0.01 (0.05 ÷ 5). P-values below this threshold were considered statistically significant. Model fit was assessed using the Hosmer–Lemeshow goodness-of-fit test and Nagelkerke’s R². All models demonstrated acceptable fit (p > 0.05). The statistical analyses were performed using IBM SPSS Statistics 27, with a significance level set at P < 0.05.

Sample size calculation

Abdominal obesity was selected as the primary dependent variable for calculating the sample size, as it required the largest sample size. A minimum of 400 individuals were targeted for recruitment, allowing for potential dropouts. The sample size was calculated using a standard formula based on the difference in proportions between groups, considering a 95% confidence level and 80% power. This method has been widely applied in epidemiological studies and allowed for an accurate estimation of the required sample size for the cross-sectional design.

Carbohydrate quality indicators

The quality of carbohydrates plays a pivotal role in influencing various health outcomes, particularly in the elderly, where nutritional needs and metabolic responses undergo significant changes. In this study, we selected fiber content and glycemic load (GL) as key indicators of carbohydrate quality due to their established relevance in the literature [39]. Fiber content reflects the structural integrity and nutrient density of carbohydrates, which are critical for maintaining gut health, modulating blood glucose levels, and promoting satiety [40]. Glycemic load, on the other hand, accounts for both the quantity and glycemic index of carbohydrates, providing a comprehensive measure of their impact on postprandial glucose and insulin responses [41]. These factors are particularly important in the elderly, as they are more vulnerable to metabolic dysregulation, mood disorders, and sleep disturbances [42]. By focusing on these two indicators, this study aims to provide robust insights into the relationship between carbohydrate quality and mental health outcomes in this population.

Results

Table 1: The demographic characteristics of the study population, based on the quality of macronutrient intake (fiber content of carbohydrate tertiles and glycemic load tertiles), are summarized in Table 1. Our study found several significant associations based on the quality of macronutrients. Higher fiber intake was associated with a higher percentage of males (P = 0.037) and higher socioeconomic status (SES) (P < 0.001) and showed significant differences in BMI (P = 0.015) and sleep quality (P = 0.032). Higher glycemic load was linked to a higher percentage of males (P < 0.001), higher SES (P = 0.002), more smokers (P = 0.014), and better sleep quality (P = 0.013). Higher fiber intake was associated with a lower prevalence of depression (P = 0.026), and both higher fiber and glycemic load were significantly associated with lower anxiety (P < 0.001 and P = 0.004, respectively) and lower stress levels (P < 0.001 for both).

Table 1.

General characteristics based on the quality of carbohydrate

	Fiber content of carbohydrate tertiles				Glycemic load tertiles
	T1	T2	T3	P-value	T1	T2	T3	P-value
Agea	63.701 ± 3.68	62.803 ± 3.13	63.393 ± 3.90	0.116	63.58 ± 3.55	63.02 ± 3.35	63.32 ± 3.88	0.45
Gender: Male	27.3 (54)	36.9 (73)	35.9 (71)	0.037	24.2 (48)	33.3(66)	42.4 (84)	< 0.001
Female	39.4 (78)	29.8 (59)	30.8 (61)	0.037	42.1 (83)	33.5 (66)	24.4 (48)	< 0.001
SESb: Low	45.2 (71)	28 (44)	26.8 (42)	< 0.001	43.3 (68)	31.2 (49)	25.5 (40)	0.002
Medium	28.4 (31)	37.6 (41)	33.9 (37)	< 0.001	30.3 (33)	35.8 (39)	33.9 (37)	0.002
High	21.9 (28)	36.7 (47)	41.4 (53)	< 0.001	22 (28)	34.6 (44)	43.3 (55)	0.002
Being married	32.8 (103)	33.8 (106)	33.4 (105)	0.945	32.6 (102)	34.5 (108)	32.9 (103)	0.707
BMI	29.371 ± 3.37	29.38 ± 2.44	28.44 ± 3.13	0.015	29.47 ± 3.14	29.01 ± 2.89	28.70 ± 3.06	0.120
Being a smoker	27.2 (22)	29.6 (24)	43.2 (35)	0.107	28.4 (23)	24.7 (20)	46.9 (38)	0.014
Sleep quality: Poor	32.9 (95)	36.7 (106)	30.4 (88)	0.032	31.3 (90)	37.8 (109)	30.9 (89)	0.013
Good	33.7 (35)	24 (25)	42.3 (44)	0.032	36.5 (38)	22.1 (23)	41.3 (43)	0.013
Having depression	40.6 (52)	34.4 (44)	25 (32)	0.026	37.5 (48)	35.2 (45)	27.3 (35)	0.203
Having anxiety	36.9 (130)	40.3 (131)	22.7 (131)	< 0.001	33.1 (58)	41.1 (72)	25.7 (45)	0.004
Having stress	42.6 (69)	36.4 (59)	21 (34)	< 0.001	39.5 (64)	39.5 (64)	21 (34)	< 0.001

SES Social economic status

P values were obtained from one-way ANOVA or χ2 test, where appropriate

^aMean ± SD

^bPercentages and (count)

Table 2: The dietary intake analysis based on carbohydrate quality indicators revealed several significant findings. Higher fiber content and glycemic load were associated with increased intake of protein, carbohydrates, fat, MUFA, DHA, iron, magnesium, and vitamin C (all P < 0.001). Significant differences in protein and fat intake were also observed across glycemic load tertiles (P = 0.001). Cholesterol and PUFA intake did not show significant differences across fiber content tertiles, while calcium intake showed significant differences across glycemic load tertiles (P < 0.001).

Table 2.

Dietary intake based on the quality of carbohydrate across carbohydrate quality indicators

	Glycemic load tertiles				Fiber content of carbohydrate tertiles
	T1 (n = 131)	T2 (n = 132)	T3 (n = 132)	P-value	T1	T2	T3	P-value
Protein (g/d)	51.31 ± 11.32	65.44 ± 10.44	87.76 ± 18.02	< 0.001	52.71 ± 11.84	66.43 ± 13.53	58.53 ± 19.19	< 0.001
Carbohydrate (g/d)	245.98 ± 49.35	333.72 ± 57.82	464.04 ± 92.31	< 0.001	257.50 ± 66.17	337.21 ± 68.18	450.07 ± 102.83	< 0.001
Fat (g/d)	49.31 ± 14.75	62.72 ± 22.63	83.38 ± 29.16	< 0.001	51.36 ± 16.45	63.48 ± 23.23	80.75 ± 30.34	0.001
Cholesterol (mg/d)	208.06 ± 109.57	219.28 ± 94.36	297.34 ± 148.22	0.133	205.32 ± 90.04	233.17 ± 94.17	285.54 ± 165.02	0.931
Saturated Fatty acid (g/d)	15.98 ± 5.30	19.78 ± 7.36	26.73 ± 11.14	0.058	17.06 ± 5.90	20.22 ± 8.81	25.26 ± 10.90	< 0.001
MUFA (g/d)	15.64 ± 5.94	19.85 ± 8.64	26.77 ± 10.81	< 0.001	16.48 ± 6.55	20.11 ± 8.66	25.74 ± 11.32	< 0.001
PUFA (g/d)	13.04 ± 4.87	16.11 ± 7.03	20.21 ± 7.58	< 0.001	12.86 ± 4.62	16.55 ± 6.11	19.99 ± 8.45	0.122
EPA (g/d)	0.0053 ± 0.0062	0.0066 ± 0.0074	0.0123 ± 0.016	0.005	0.005 ± 0.006	0.007 ± 0.008	0.012 ± 0.016	0.056
DHA (g/d)	0.018 ± 0.017	0.023 ± 0.024	0.039 ± 0.042	0.004	0.018 ± 0.018	0.024 ± 0.023	0.037 ± 0.042	0.040
Calcium (mg/d)	609.41 ± 188.30	758.20 ± 202.82	1002.61 ± 377.32	< 0.001	649.46 ± 219.25	765.18 ± 291.19	958.92 ± 340.95	0.179
Vitamin B6 (mg/d)	1.05 ± 0.37	1.41 ± 0.52	2.08 ± 1.02	0.127	1.01 ± 0.29	1.35 ± 0.39	2.17 ± 1.02	< 0.001
Folate (mcg/d)	218.66 ± 73.13	277.41 ± 92.36	401.09 ± 207.04	0.372	211.64 ± 55.56	267.60 ± 73.61	419.14 ± 206.02	< 0.001
Vitamin B12 (mcg/d)	2.37 ± 9.58	2.65 ± 1.09	3.74 ± 1..65	0.005	2.46 ± 1.08	2.84 ± 1.32	3.46 ± 1.56	0.977
Iron (mg/d)	10.92 ± 2.38	14.94 ± 3.22	20.97 ± 5.08	< 0.001	11.17 ± 2.57	14.84 ± 3.31	20.94 ± 5.16	< 0.001
Magnesium (mg/d)	181.39 ± 39.7	231.49 ± 48.84	310.59 ± 83.46	< 0.001	183.95 ± 44.76	230.91 ± 52.16	310.29 ± 81.84	< 0.001
Vitamin C (mg/d)	94.69 ± 39.57	131.42 ± 49.50	189.26 ± 79.07	< 0.001	90.02 ± 39.41	125.32 ± 38.31	200.89 ± 73.29	< 0.001

Obtained by ANCOVA test; adjusted for energy intake

EPA eicosapentaenoic acid, DHA docosahexaenoic acid

Table 3 represents the association between carbohydrate intake quantity and mental health and sleep quality among elderly individuals. Higher carbohydrate intake, categorized into tertiles (T1, T2, T3), showed varying associations with mental health and sleep quality outcomes. For anxiety, no significant associations were found across the models. In the case of depression, higher carbohydrate intake showed a marginally significant association in the fully adjusted model with the highest tertile (T3) having increased odds (OR: 3.930, 95% CI: 1.2–12.867, P-trend = 0.040) compared to the lowest tertile (T1). Stress was marginally significantly associated with carbohydrate intake in the crude model (OR: 0.597, 95% CI: 0.364–0.978, P-trend = 0.040), but the association was not significant after adjusting for confounders.. Sleep quality showed a marginally significant association in the crude model (OR: 1.691, 95% CI: 0.991–2.884, P-trend = 0.045), but the significance was lost after adjustment. For mood, higher carbohydrate intake was significantly associated with better outcomes in the crude model (OR: 0.510, 95% CI: 0.305–0.854, P-trend = 0.010) and approached significance in Model I (P-trend = 0.053), but no significant associations were observed in the fully adjusted models. These results suggest that carbohydrate intake has complex associations with mental health and sleep quality, with some effects being attenuated after adjusting for confounders.

Table 3.

The association between sleep quality, mood, depression, anxiety and stress based on carbohydrate amount (%)

	T1	T2	T3		P-trend
Subjects, n	133	133	133
Mean + SD	67 ± 38.53	200 ± 38.53	333 ± 38.53
Lower limit	60.39	193.39	326.39
Upper limit	73.61	206.61	339.61
Anxiety	OR (95% CI)	OR (95% CI)	OR (95% CI)
Crude	1.00	0.956 (0.589–1.553)	0.846 (0.520–1.376)		0.499
Model I	1.00	1.029 (0.627–1.688)	1.009 (0.606–1.679)		0.972
Model II	1.00	1.087 (0.586–2.014)	1.328 (0.467–3.780)		0.617
Model III	1.00	1.087 (0.586–2.014)	1.327 (0.466–3.778)		0.619
Depression
Crude	1.00	0.773 (0.459–1.301)	1.022 (0.615–1.699)		0.929
Model I	1.00	0.829 (0.488–1.410)	1.241 (0.728–2.115)		0.436
Model II	1.00	1.410 (0.701–2.835)	3.922 (1.196–12.864)		0.041
Model III	1.00	1.408 (0.701–2.831)	3.930 (1.2–12.867)		0.040
Stress
Crude	1.00	0.724 (0.444–1.180)	0.597 (0.364–0.978)		0.040
Model I	1.00	0.776 (0.471–1.280)	0.722 (0.431–1.210)		0.214
Model II	1.00	1.341 (0.696–2.583)	2.181 (0.703–6.765)		0.196
Model III	1.00	1.336 (0.694–2.572)	2.175 (0.703–6.734)		0.197
Sleep Quality
Crude	1.00	0.759 (0.423–1.363)	1.691 (0.991–2.884)		0.045
Model I	1.00	0.679 (0.373–1.235)	1.367 (0.782–2.388)		0.224
Model II	1.00	0.442 (0.210–0.931)	0.568 (0.173–1.863)		0.180
Model III	1.00	0.443 (0.210–0.935)	0.575 (0.176–1.883)		0.187
Mood
Crude	1.00	0.613 (0.370–1.014)	0.510 (0.305–0.854)	0.010
Model I	1.00	0.651 (0.390–1.087)	0.596 (0.350–1.017)		0.053
Model II	1.00	0.862 (0.424–1.753)	0.963 (0.282–3.286)		0.842
Model III	1.00	0.868 (0.427–1.767)	0.975 (0.285–0.335)		0.860

Model I: Adjusted for age and sex

Model II: Model 1 + smoking, marital status, SES status, energy (kcal), omega-3 fatty acids, magnesium, cobalamin, total folate, pyridoxine, thyroid disease, dyslipidemia and blood pressure

Model III: Model II + BMI

P-trend is obtained from logistic regression by considering tertiles of fiber content & GL as ordinal variable

A Bonferroni correction was applied to adjust for multiple comparisons across five outcomes. Statistical significance was considered at P < 0.01

Table 4 examined the association between carbohydrate quality (fiber content and glycemic load) and mental health and sleep quality. Higher fiber content was associated with reduced anxiety, depression, and stress in both crude and adjusted models. In the fully adjusted model, highest fiber intake tertile (T3) was significantly associated with lower odds of anxiety (OR: 0.177, 95% CI: 0.075–0.419, P-trend < 0.001), depression (OR: 0.509, 95% CI: 0.204–1.271, P-trend = 0.183), and stress (OR: 0.307, 95% CI: 0.129–0.733, P-trend = 0.014). Conversely, glycemic load showed a significant association with reduced stress and improved mood in the crude model but lost significance in the fully adjusted model. Sleep quality was significantly associated with glycemic load in the fully adjusted model (OR: 0.159, 95% CI: 0.054–0.468, P-trend < 0.001).

Table 4.

The association between Sleep quality, mood, depression, anxiety and stress based on carbohydrate quality

	Fiber content				Glycemic load
	T1	T2	T3	P-trend	T1	T2	T3		P-trend
Subjects, n	132	132	132		131	132	132
Mean + SD	66.5 ± 38.25	198.5 ± 38.25	330.5 ± 38.25		66 ± 37.96	197.5 ± 38.24	329.5 ± 38.24
Lower limit	59.91	191.91	323.91		59.44	190.91	322.91
Upper limit	73.08	205.08	337.08		72.56	204.08	336.08
Anxiety	OR (95% CI)	OR (95% CI)	OR (95% CI)		OR (95% CI)	OR (95% CI)	OR (95% CI)
Crude	1.00	1.183 (0.728–1.924)	0.440 (0.265–0.729)	0.002	1.00	1.515 (0.930–2.468)	0.657 (0.399–1.084)		0.106
Model I	1.00	1.333 (0.807–2.201)	0.462 (0.276–0.774)	0.003	1.00	1.689 (1.022–2.790)	0.761 (0.454–1.278)		0.306
Model II	1.00	0.987 (0.554–1.759)	0.180 (0.076–0.424)	< 0.001	1.00	1.467 (0.8005–2.676)	0.644 (0.258–1.604)		0.547
Model III	1.00	0.990 (0.555–1.765)	0.177 (0.075–0.419)	< 0.001	1.00	1.467 (0.804–2.675)	0.639 (0.256–1.595)		0.541
Depression
Crude	1.00	0.759 (0.458–1.256)	0.485 (0.285–0.825)	0.008	1.00	0.894 (0.538–1.484)	0.629 (0.3725–1.065)		0.087
Model I	1.00	0.822 (0.491–1.377)	0.509 (0.298–0.872)	0.014	1.00	0.966 (0.576–1.619)	0.721 (0.419–1.241)		0.244
Model II	1.00	0.883 (0.476–1.638)	0.520 (0.209–1.294)	0.191	1.00	1.171 (0.606–2.263)	1.183 (0.433–3.231)		0.724
Model III	1.00	0.887 (0.478–1.645)	0.509 (0.204–1.271)	0.183	1.00	1.171 (0.606–2.261)	1.193 (0.436–3.259)		0.715
Stress
Crude	1.00	0.724 (0.445–1.179)	0.310 (0.184–0.522)	< 0.001	1.00	0.985 (0.606–1.601)	0.365 (0.217–0.615)		< 0.001
Model I	1.00	0.765 (0.462–1.265)	0.319 (0.187–0.543)	< 0.001	1.00	1.049 (0.636–1.728)	0.417 (0.244–0.714)		0.002
Model II	1.00	0.813 (0.450–1.467)	0.326 (0.138–0.773)	0.018	1.00	1.279 (0.684–2.390)	0.594 (0.226–1.565)		0.456
Model III	1.00	0.818 (0.453–1.479)	0.307 (0.129–0.733)	0.014	1.00	1.280 (0.684–2.393)	0.602 (0.229–1.586)		0.471
Sleep quality
Crude	1.00	0.640 (0.357–1.147)	1.357 (0.799–2.306)	0.236	1.00	0.5 (0.278–0.9)	1.144 (0.676–1.936)		0.580
Model I	1.00	0.563 (0.308–1.029)	1.233 (0.716–2.124)	0.381	1.00	0.441 (0.241–0.808)	0.923 (0.532–1.602)		0.867
Model II	1.00	0.432 (0.216–0.864)	0.651 (0.265–1.599)	0.259	1.00	0.236 (0.111–0.502)	0.167 (0.057–0.489)		< 0.001
Model III	1.00	0.423 (0.210–0.850)	0.697 (0.282–1.724)	0.315	1.00	0.229 (0.107–0.489)	0.159 (0.054–0.468)		< 0.001
Mood
Crude	1.00	0.496 (0.301–0.818)	0.261 (0.151–0.451)	< 0.001	1.00	0.696 (0.424–1.143)	0.324 (0.188–0.559)	< 0.001
Model I	1.00	0.544 (0.326–0.908)	0.272 (0.156–0.473)	< 0.001	1.00	0.748 (0.452–1.240)	0.363 (0.208–0.636)	< 0.001
Model II	1.00	0.703 (0.372–1.332)	0.5 (0.189–1.324)	0.150	1.00	0.845 (0.426–1.677)	0.492 (0.166–1.460)	0.239
Model III	1.00	0.711 (0.375–1.347)	0.486 (0.183–1.293)	0.141	1.00	0.843 (0.425–1.674)	0.502 (0.169–1.494)	0.253

Model I: Adjusted for age and sex

Model II: Model 1 + smoking, marital status, SES status, energy (kcal), omega-3 fatty acids, magnesium, cobalamin, total folate, pyridoxine, thyroid disease, dyslipidemia and blood pressure

Model III: Model II + BMI

P-trend is obtained from logistic regression by considering tertiles of fiber content & GL as ordinal variable

A Bonferroni correction was applied to adjust for multiple comparisons across five outcomes. Statistical significance was considered at P < 0.01

Figure 1 illustrates the associations between carbohydrate quality—measured by fiber content and glycemic load—and mental health outcomes including anxiety, depression, stress, sleep quality, and mood in the fully adjusted model. Higher fiber intake was significantly associated with lower odds of anxiety, depression, stress, poor sleep quality, and poor mood in fully adjusted models. Similarly, lower glycemic load was linked to better sleep quality and mood. These associations remained significant after adjusting for potential confounders and correcting for multiple comparisons.

Fig. 1.

Forest plot of associations between carbohydrate quality (Fiber Content and Glycemic Load) and sleep and menta health outcomes

Discussion

The present study provides substantial evidence on the association between carbohydrate intake quantity and quality, particularly fiber content and glycemic load, and mental health outcomes in an elderly population. High carbohydrate quantity was also associated with increased odds of depression. Our findings also revealed that higher carbohydrate quality, as assessed by fiber content, was consistently associated with reduced anxiety and stress. Additionally, high carbohydrate quality, as assessed by glycemic load, was linked to lower odds of poor sleep quality.

Our study found that higher carbohydrate quality, indicated by dietary glycemic load (GL), was not associated with improved mental health. Previous studies have investigated the relationship between dietary GL and various mental health parameters, including depression, anxiety, and stress, with mixed results. Some found an inverse association between higher GL and stress reported by Amirinejad et al. [43] and the finding from a systematic review by Rahimlou et al. [44], who noted that higher dietary GL was associated with lower risk of depression. However, other studies have reported no significant associations between dietary GL and depression or anxiety [43]. For example, Amirinejad et al. found no significant associations between dietary GL and odds of depression or anxiety, and Haghighatdoost et al. observed that higher GL values were linked to lower odds for mental disorders, depression, and psychological distress.

The discrepancies across studies can be attributed to several factors. Study design, where cross-sectional studies like ours, as well as those by Amirinejad et al. and Haghighatdoost et al., capture data at a single point in time, making it difficult to establish causal relationships, whereas longitudinal studies like those reviewed by Rahimlou et al. show how dietary patterns over time affect mental health. Population differences, such as age and health status, also play a role. Our study focuses on an elderly cohort, while previous studies included a broader age range, which affects metabolic and psychological responses to diet. Dietary assessment methods vary, with studies using different Food Frequency Questionnaires (FFQs) that have varying accuracy in assessing dietary intake, and differences in the calculation of GI and GL. Other confounding factors include varying adjustments for confounders like age, sex, BMI, physical activity, smoking, and other dietary components, as well as interaction effects, where some studies noted that the interaction between GI and sex can affect mental health outcomes [45].

Our findings also indicate that high carbohydrate quality, as assessed by glycemic load, is associated with lower odds of poor sleep quality in the elderly population. This aligns with previous research suggesting that dietary patterns with a high glycemic load may positively influence sleep quality. For instance, a study by Farhadnejad et al. [46] found that higher adherence to dietary glycemic load was associated with reduced odds of insomnia in Iranian adults. Similarly, a systematic review by Vlahoyiannis et al. [47] highlighted that carbohydrate quality could impact sleep initiation and efficiency. However, it is important to note that some studies have reported contrasting results. For example, Mirzababaei et al. [48] found that higher adherence to dietary insulin index and load was associated with poor sleep quality among overweight and obese women. This discrepancy may be due to differences in study populations, dietary assessment methods, and confounding factors.

Our findings revealed that higher carbohydrate quality based on its fiber content was consistently associated with reduced odds of anxiety and stress. To our knowledge, no previous study has assessed this specific association. However, a study conducted by Aysun Yuksel et al. [49] examined carbohydrate quality using a broader set of criteria, including total fiber intake, the ratio of whole grains to total grains, the ratio of solid carbohydrates to total carbohydrates, and the glycemic index. Their results indicated that higher carbohydrate quality was associated with better sleep quality and lower body mass index, but no significant correlation was found with anxiety, depression, or stress. The differences in findings could be due to varying methods for assessing carbohydrate quality. Our study specifically examined fiber content of carbohydrate, potentially having a more direct impact on mood and mental health. In contrast, Aysun Yuksel et al. included factors like whole grain ratio and glycemic index, offering a broader view that may dilute the specific effects of fiber on mental health, possibly explaining the lack of significant associations with anxiety, depression, or stress in their study.

Also, our study found a significant association between carbohydrate quantity and higher odds of depression. However, Ebrahimpour-Koujan et al. [50] found no significant association between adherence to a low carbohydrate diet and the prevalence of psychological disorders, including depression, anxiety, and psychological distress. Similarly, a cross-sectional study conducted among Japanese men found no association between carbohydrate or fat intake and depressive symptoms [51]. In contrast, other studies aligns with our findings have indicated an inverse relationship between high carbohydrate and the prevalence of depression and anxiety [52] or depression [53].

The discrepancy between our findings and those of Ebrahimpour-Koujan et al. and Nanri et al. may be attributed to differences in study populations and methodologies. Our study focused on an elderly population, which may have different dietary and psychological responses compared to the broader adult population studied. Additionally, the assessment methods for dietary intake and psychological outcomes may vary between studies, potentially influencing the results.

In the current study, the consistent association between higher carbohydrate quality, as assessed by fiber content, and reduced odds of anxiety and stress can be explained through a combination of interconnected mechanisms that involve the gut-brain axis, blood sugar regulation, and reduced inflammation [54]. Firstly, dietary fiber plays a key role in promoting a healthy gut microbiome [55]. Fiber acts as a prebiotic, feeding beneficial bacteria in the colon that then produce short-chain fatty acids (SCFAs) [56]. These SCFAs, in turn, can modulate the immune system and have anti-inflammatory properties [57]. This is crucial because a perturbed gut microbiota, characterized by a lower abundance of anti-inflammatory bacteria and higher levels of systemic inflammation, has been linked to psychiatric disorders [58]. Secondly, the gut microbiota is involved in the synthesis of neurotransmitters such as serotonin, which plays a vital role in mood regulation [59]. By supporting a diverse and healthy gut environment, fiber may thus enhance serotonin production, leading to an improvement in mood and reduction in symptoms of depression and anxiety. Furthermore, fiber slows down the absorption of glucose, which leads to more stable blood sugar levels and reduces the need for large insulin secretions [55]. This can prevent mood fluctuations associated with rapid blood sugar and insulin spikes. Finally, many fiber-rich foods are also rich in essential vitamins, minerals, and antioxidants, which are beneficial for brain health and have anti-inflammatory properties [60]. In sum, higher fiber intake may improve mental health by promoting a healthy gut microbiome, stabilizing blood sugar levels, reducing inflammation, and increasing the availability of key nutrients.

The association between higher glycemic load diets and improved sleep quality may be explained through several physiological and hormonal mechanisms. High-GL foods cause a rapid spike in blood glucose levels, triggering insulin secretion [47]. Insulin, in turn, promotes the uptake of large neutral amino acids into muscle cells, except for tryptophan, thereby increasing its circulating levels [61]. As a precursor to serotonin, elevated tryptophan levels enhance serotonin production, which is subsequently converted into melatonin, a hormone crucial for regulating sleep–wake cycles [62]. This process suggests that consuming high-GL foods may facilitate sleep by promoting serotonin and melatonin synthesis, thereby improving sleep quality.

While certain mechanisms might suggest a protective role of carbohydrates against depression, their observed findings in this specific elderly Iranian population indicate the opposite. High-carbohydrate intake triggers insulin release, which facilitates the uptake of large neutral amino acids into muscle cells, leaving tryptophan more available for transport across the blood–brain barrier [63]. Increased tryptophan availability enhances serotonin synthesis, which has been linked to improved mood and lower depressive symptoms [64]. Additionally, carbohydrates help regulate the hypothalamic–pituitary–adrenal (HPA) axis, which controls cortisol secretion, thereby reducing stress-related neuroendocrine disruptions that contribute to depression [65]. Diets rich in complex carbohydrates also provide a steady glucose supply, supporting optimal brain function and neurotransmitter balance, both of which are essential for mental health [66]. These mechanisms collectively suggest that adequate carbohydrate consumption plays a crucial role in reducing the risk of depression.

This study addresses a research gap by focusing on the elderly population, providing valuable insights specific to this group. Psychological health outcomes (mood, anxiety, stress, and depression) were comprehensively assessed using validated scales like DASS-21 and POMS. The use of a validated 168-item Food Frequency Questionnaire ensures a thorough measure of carbohydrate quality and quantity. Logistic regression was employed for robust statistical analysis. The study highlights the impact of carbohydrate quality (fiber content and glycemic load) on sleep and mental health. However, this study has several limitations that should be considered when interpreting the findings. First, its cross-sectional design precludes any conclusions about causality between carbohydrate intake and sleep or mental health outcomes, allowing only for the identification of associations. Second, the study population consisted exclusively of elderly individuals residing in Tehran, which may limit the generalizability of the results to other age groups, regions, or populations with differing dietary patterns. Third, dietary intake was assessed using a self-reported Food Frequency Questionnaire (FFQ), which is subject to recall bias and potential inaccuracies. Additionally, while several confounding variables were adjusted for, the possibility of residual confounding by unmeasured factors such as smoking or physical activity remains. The study also assessed carbohydrate quality based solely on fiber content and glycemic load, potentially overlooking other relevant aspects. Furthermore, the unique physiological and psychological characteristics of the elderly population may influence how these findings apply to the general population. Additionally, the exclusion of participants with chronic illnesses was based on self-reported data, which may be subject to misclassification or underreporting and could potentially influence the findings. Lastly, reliance on self-reported data for both dietary intake and psychological measures could introduce bias and affect the accuracy of the results.

Conclusion

In conclusion, our study highlights the beneficial impact of high-quality carbohydrate intake, as measured by fiber content and glycemic load, on sleep and mental health outcomes among elderly individuals. These findings underscore the potential importance of dietary quality in promoting well-being in this population. However, due to the cross-sectional design of the study, causality cannot be established, and the results may not be generalizable beyond the specific elderly population in Tehran. Therefore, further investigation, particularly through longitudinal studies, is necessary to fully understand these associations and their implications for public health.

Abbreviations

FFQ

Food frequency questionnaire

DASS-21

Depression, Anxiety, and Stress Scale

PQSI

Pittsburgh Sleep Quality Index

GI

Glycemic index

GL

Glycemic load

IPAQ

International Physical Activity Questionnaire

MET

Metabolic Equivalent of Task

​SCFA

Short-chain fatty acids

ANOVA

One-way analysis of variance

ANCOVA

Analysis of covariance

BMI

Body mass index

CI

Confidence Interval

OR

Odd Ratio

MUFAs

Monounsaturated fatty acids

PUFAs

Polyunsaturated fatty acids

SD

Standard Deviation

SES

Social economic status

Authors’ contributions

BG contributed to manuscript drafting, statistical analyses and data interpretation. HA, MKD and PNH contributed to data collection. LA contributed to conception, design and supervised the study. All authors approved the final manuscript for submission.

Funding

Not applicable.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Tehran University of Medical Sciences, Tehran, Iran. The study adhered to the ethical principles outlined in the Declaration of Helsinki for research involving human participants. Informed consent was obtained from all participants prior to their inclusion in the study. The purpose of the study was clearly explained at the beginning of the questionnaire, and participation was entirely voluntary. Participants were assured of the anonymity and confidentiality of their responses.

Consent for publication

Verbal consent was attained.

Competing interests

The authors declare no competing interests.