5  Is Latitude Associated with Chronotype?

The following study was designed for publication in the journal Scientific Reports (IF 2023: 3.8/JCR | CAPES: A1/2017–2020) and structured in accordance with the journal’s submission guidelines.

5.1 Abstract

Chronotypes are temporal phenotypes that reflect our internal temporal organization, a product of evolutionary pressures enabling organisms to anticipate events. These intrinsic rhythms are entrain by zeitgebers — periodical environmental stimuli with the ability to regulate biological rhythmic expression, with light exposure being the primary mechanism. Given light’s role in these systems, previous research hypothesized that latitude might significantly influence chronotypes, suggesting that populations near the equator would exhibit more morning-leaning characteristics due to more consistent light/dark cycles, while populations near the poles might display more evening-leaning tendencies with a potentially freer expression of intrinsic rhythms. To test this hypothesis, we analyzed chronotype data from a large sample of \(65,824\) subjects across diverse latitudes in Brazil. Our results revealed a negligeble effect size of latitude on chronotype (\(f^2 = 0.0030818, 95\% \ \text{CI}[0, 0.0121371]\)), indicating that the entrainment phenomenon is far more complex than previously conceived. These findings challenge simplified environmental models of biological timing and underscore the need for more nuanced investigations into the mechanisms underlying temporal phenotypes, opening new avenues for understanding the intricate relationship between environmental cues and individual circadian rhythms.

5.2 Introduction

Humans exhibit a variety of observable traits, such as eye or hair color, which are referred to as phenotypes. These phenotypes also manifest in the way our bodies function.

A chronotype is a temporal phenotype (Ehret, 1974; Pittendrigh, 1993), typically used to refer to endogenous circadian rhythms — biological rhythms with periods close to 24 hours. Chronobiology, the science that studies biological rhythms, suggests that the evolution of these internal oscillators is closely linked to our environment, particularly the day/night cycle. This cycle, alongside human evolution, created environmental pressures that led to the development of temporal organization within organisms (Aschoff, 1989b; Paranjpe & Sharma, 2005; Pittendrigh, 1981). Such organization allowed organisms to predict events and better manage their needs, such as storing food for winter (Aschoff, 1989a).

For a temporal system to be useful, it must be capable of adapting to environmental changes. Environmental signals capable of regulating biological rhythms are known as zeitgebers (from the German zeit, meaning time, and geber, meaning donor (Cambridge University Press, n.d.)). These zeitgebers provide inputs that can shift and synchronize biological rhythms in a process called entrainment (Khalsa et al., 2003; Minors et al., 1991).

The primary zeitgeber influencing biological rhythms is the light/dark cycle, or, simply, light exposure (Aschoff, 1960; Pittendrigh, 1960; Roenneberg & Merrow, 2016). Given its significant role in entraining the biological clock, several studies have hypothesized that the latitudinal shift of the sun, due to the Earth’s axial tilt, might lead to different temporal traits in populations near the equator compared to those closer to the poles (Bohlen & Simpson, 1973; Horzum et al., 2015; Leocadio-Miguel et al., 2014, 2017; Randler, 2008). This is based on the idea that populations at low or higher latitudes experience greater fluctuations in sunlight and a weaker overall solar zeitgeber. This concept is known as the latitude hypothesis, or the environmental hypothesis of circadian rhythm regulation.

Several studies have claimed to find this association in humans, but the evidence they provide is of very low quality or is misleading (Horzum et al., 2015; Leocadio-Miguel et al., 2014, 2017; Randler, 2008; Wang et al., 2023). A notable attempt was made by Leocadio-Miguel et al. (2017), who measured the chronotype of \(12,884\) Brazilian subjects across a wide latitudinal range using the Horne-Östberg (HO) Morningness–Eveningness questionnaire (Horne & Östberg, 1976). Although the authors concluded that there was a significant association between latitude and chronotype, their results were too small to be considered practically significant (even by lenient standards), with latitude explaining only approximately \(0.388\%\) of the variance in chronotype (Cohen’s \(f^2 = 0.004143174\)). One possible explanation for this result is that the HO measures psychological traits rather than the biological states of circadian rhythms themselves (Roenneberg, Pilz, et al., 2019), suggesting it may not be the most suitable tool for testing the hypothesis (Leocadio-Miguel et al., 2014).

Building on Leocadio-Miguel et al. (2017), this study offers a novel attempt to test the latitude hypothesis by employing a biological approach through the Munich ChronoType Questionnaire (MCTQ) (Roenneberg et al., 2003) and an enhanced statistical methodology. Additionally, it utilizes the largest dataset on chronotype from a single country, as far as the existing literature suggests, comprising \(65,824\) respondents, all residing within the same timezone in Brazil and completing the survey within a one-week window (Figure 5.1).

Code
weighted_data |> 
  plot_brazil_municipality(
    transform = "log10",
    direction = -1,
    alpha = 0.75,
    breaks = c(10, 500, 1000, 5000, 7500),
    point = TRUE
  )
Figure 5.1: Geographical distribution of the sample used in the analysis (\(n = 65,824\)).
Each point represents a municipality, with its size proportional to the number of participants and its color indicating participant density. The sample includes Brazilian individuals aged 18 or older, residing in the UTC-3 timezone, who completed the survey between October 15th and 21st, 2017 (one-week window). The size and color scale are logarithmic (\(\log_{10}\)).

Source: Created by the author.

5.3 Results

The Munich Chronotype Questionnaire (MCTQ) uses the midpoint between sleep onset (SO) and sleep end (SE) on work-free days (MSFsc), with a sleep correction (sc) applied if sleep debt is detected, as a proxy for chronotype (Roenneberg et al., 2003). For example, if an individual sleeps from \(\text{00:00}\) to \(\text{08:00}\), the midpoint would be \(\text{04:00}\). This measure is based on the current understanding of sleep regulation, which comprises a homeostatic/sleep-dependent process (\(\text{S}\) process) and a circadian process (\(\text{C}\) process) (Borbély, 1982; Borbély et al., 2016). The midpoint of sleep on free days offers a way to observe unrestrained sleep behavior, thereby minimizing the influence of the \(\text{S}\) process and providing a better approximation of the circadian phenotype (i.e., the \(\text{C}\) process).

Our analysis revealed an overall mean MSFsc of \(\text{04:28:41}\), with an standard deviation of \(\text{01:26:13}\). The distribution is shown in Figure 5.2.

Code
weighted_data |> plot_chronotype()
Figure 5.2: Distribution of the local time of the sleep-corrected midpoint between sleep onset and sleep end on work-free days (MSFsc), a proxy for chronotype.
Chronotypes are categorized into quantiles, ranging from extremely early (\(0 |- 0.11\)) to extremely late (\(0.88 - 1\)).

Source: Created by the author, based on a data visualization from
Roenneberg, Wirz-Justice, et al. (2019, fig. 1, right part).

This represents the midsleep point for Brazilian subjects living in the UTC-3 timezone, with an intermediate or average chronotype. Considering the \(7–9\) hours of sleep recommended for healthy adults by the American Academy of Sleep Medicine (AASM) (Watson et al., 2015), one might infer that this average individual, in the absence of social constraints, would typically wake up at approximately \(\text{08:28:41}\).

The study hypothesis was tested using nested multiple linear regressions. The core idea of nested models is to evaluate the effect of including one or more predictors on the model’s variance explanation (\(\text{R}^2\)) (Maxwell et al., 2018). This is achieved by comparing a restricted model (without the latitude) with a full model (with the latitude). Cell weights, based on sex, age group, and state of residence, were applied to account for sample imbalances.

To ensure practical significance, the hypothesis test incorporated a minimum effect size (MES) criterion, aligning with the original Neyman-Pearson framework for data testing (Neyman & Pearson, 1928a, 1928b; Perezgonzalez, 2015). The MES was set at a Cohen’s \(f^2\) of \(0.02\) (equivalent to an \(\text{R}^2\) of \(0.01960784\)), a lenient threshold (Cohen, 1988, 1992). Given the well-established influence of the solar zeitgeber on biological rhythm entrainment, it is unlikely that the latitude hypothesis could be meaningfully supported without demonstrating at least a non-trivial effect.

Two tests were conducted, both starting with the same restricted model, which included age, sex, longitude, and the average monthly Global Horizontal Irradiance (GHI) at the time of questionnaire completion as predictors (\(\text{R}^2_{\text{adj}} = 0.0851096\), \(\text{F}(4, 65818) = 1531.808\), \(p\text{-value} < 1e-05\)). The first full model (A) added the average annual GHI and daylight duration for the nearest March equinox, as well as the June and December solstices, as proxies for latitude, following the methods of Leocadio-Miguel et al. (2017) (\(\text{R}^2_{\text{adj}} = 0.0879205\), \(\text{F}(8, 65814) = 794.12\), \(p\text{-value} < 1e-05\)). The second full model (B) added only latitude as a predictor (\(\text{R}^2_{\text{adj}} = 0.0856143\), \(\text{F}(5, 65817) = 1233.588\), \(p\text{-value} < 1e-05\)). All coefficients were statistically different from zero (\(\text{p-value} < 0.05\)). Assumption checking and residual diagnostics primarily relied on visual inspection, as formal assumption tests (e.g., Anderson-Darling) are often not recommended for large samples (Shatz, 2024). All validity assumptions were met, and no serious multicollinearity was found among the predictor variables.

Sunrise times for the nearest March and September equinoxes, as well as the June and December solstices, were excluded due to high multicollinearity. Daylight duration for the September equinox was excluded for its multicollinearity with daylight duration during the March equinox.

An ANOVA for nested models revealed a significant reduction in the residual sum of squares in both tests (A \(\text{F}(4, 65814) = 51.71\), \(p\text{-value} < 1e-05\)) (B \(\text{F}(1, 65817) = 37.325\), \(p\text{-value} < 1e-05\)). However, similarly to Leocadio-Miguel et al. (2017), when estimating Cohen’s \(f^2\) effect size, the results were below the MES (i.e., negligible) (A \(f^2 = 0.0030818, 95\% \ \text{CI}[0, 0.0121371]\)) (B \(f^2 = 0.0005519, 95\% \ \text{CI}[0, 0.0095239]\)).

5.4 Discussion

We emphasize that the assumption of a causal, linear relationship between latitude and chronotype constitutes an a priori hypothesis, which this study seeks to falsify.

Despite a broad latitudinal range (\(33.85026\) degrees) and a large, balanced sample, our results indicate that the effect of latitude on chronotype is negligible. Indeed, despite suggestions of a potential link in several studies, robust empirical evidence supporting this claim in humans is lacking.

Our results align with those of Leocadio-Miguel et al. (2017), who reported a similar effect size (Cohen’s \(f^2 = 0.004143174\)). However, their analysis did not incorporate a minimum effect size criterion, leading to misleading interpretations. The small and inconsistent nature of the latitude effect is illustrated in Figure 5.3, while Figure 5.4 displays the mean chronotype by Brazilian state.

Code
weighted_data |> plot_latitude_series()
Figure 5.3: Boxplots of mean MSFsc values aggregated by \(1°\) latitude intervals, illustrating the relationship between latitude and chronotype.
MSFsc represents the local time of the sleep-corrected midpoint between sleep onset and sleep end on work-free days, a proxy for chronotype. Higher MSFsc values indicate later chronotypes. The × symbol points to the mean. The orange line represents a linear regression. The differences in mean/median values across latitudes are minimal relative to the Munich ChronoType Questionnaire (MCTQ) scale.

Source: Created by the author.

The absence of a clear relationship between latitude and chronotype can be attributed to multiple factors. As Jürgen Aschoff might have put it, this may reflect a lack of “ecological significance” (Aschoff et al., 1972). Even if latitude does influence circadian rhythms, the effect could be too minor to detect or might be overshadowed by other, more prominent factors like social behaviors, work hours, or the widespread use of artificial lighting. Furthermore, the variations in sunlight exposure between latitudes may not be substantial enough to meaningfully impact the circadian system, which is highly responsive to light. Given that even minor light fluctuations can lead to measurable physiological changes (Khalsa et al., 2003; Minors et al., 1991), latitude alone may not be a decisive factor in determining chronotype.

Code
limits <- # Interquartile range (IQR): Q3 - Q1
  c(
    weighted_data |>
      dplyr::pull(msf_sc) |>
      transform_time() |>
      stats::quantile(0.25, na.rm = TRUE),
    weighted_data |>
      dplyr::pull(msf_sc) |>
      transform_time() |>
      stats::quantile(0.75, na.rm = TRUE)
  )

weighted_data |> 
  dplyr::mutate(
    msf_sc = 
      msf_sc |>
      lubritime:::link_to_timeline(
        threshold = hms::parse_hms("12:00:00")
      ) |>
      as.numeric()
  ) |>
  plot_brazil_state(
    col_fill = "msf_sc",
    transform = "identity",
    binned = FALSE,
    breaks = seq(limits[1], limits[2], length.out = 6) |> rm_caps(),
    labels = labels_hms,
    limits = limits, # !!!
    quiet = TRUE
  )
Figure 5.4: Geographical distribution of MSFsc values by Brazilian state, illustrating how chronotype varies with latitude in Brazil.
MSFsc is a proxy for chronotype, representing the midpoint of sleep on work-free days, adjusted for sleep debt. Higher MSFsc values correspond to later chronotypes. The color scale was not transformed and it has as limits the first and third quartile (interquartile range). Differences in mean MSFsc values across states are small and fall within a narrow range relative to the scale of the Munich ChronoType Questionnaire (MCTQ), limiting the significance of these variations.

Source: Created by the author.

This study suggests a more complex relationship between latitude and the circadian system than originally expected. The perceived link between these variables may be a consequence of prioritizing statistical rituals over statistical thinking and a tendency toward confirmation bias, rather than rigorous and unbiased data analysis.

5.5 Methods

5.5.1 Measurement Instrument

Chronotypes were assessed using a sleep log based on the core version of the Munich ChronoType Questionnaire (MCTQ) (Roenneberg et al., 2003), a well-validated and widely used self-report tool for measuring sleep-wake behavior and determining chronotype (Roenneberg, Pilz, et al., 2019). The MCTQ derives chronotype from the sleep-corrected midpoint of sleep on free days (MSFsc), which compensates for sleep debt incurred during the workweek (Roenneberg, 2012). Figure 5.5 illustrates the variables collected by the MCTQ.

Figure 5.5: Variables measured by the Munich Chronotype Questionnaire (MCTQ). In its standard version, these variables are collected in the context of workdays and work-free days.
BT = Local time of going to bed. SPrep = Local time of preparing to sleep. SLat = Sleep latency (Duration. Time to fall asleep after preparing to sleep). SO = Local time of sleep onset. SD = Sleep duration. MS = Local time of mid-sleep. SE = Local time of sleep end. Alarm = Indicates whether the respondent uses an alarm clock. SI = “Sleep inertia” (Duration. Despite the name, this variable represents the time the respondent takes to get up after sleep end). GU = Local time of getting out of bed. TBT = Total time in bed.

Source: Created by the author.

Participants completed an online questionnaire, which included the sleep log as well as sociodemographic (e.g., age, sex), geographic (e.g., full residential address), anthropometric (e.g., weight, height), and data on work and study routines. A version of the questionnaire, stored independently by the Internet Archive organization, can be viewed at https://web.archive.org/web/20171018043514/each.usp.br/gipso/mctq.

5.5.2 Geographic Parameters

We obtained latitude and longitude data by geocoding participants’ residential addresses using two main resources:

To ensure consistency, we randomly compared results from QualoCEP and Google Geocoding API. This can be seen in the supplementary materials.

5.5.3 Solar Irradiance Data

The solar irradiance data came from the 2017 Solar Energy Atlas of Brazil’s National Institute for Space Research (INPE) (E. B. Pereira et al., 2017). We used the Global Horizontal Irradiance (GHI) data, representing the total amount of irradiance received from above by a surface horizontal to the ground.

5.5.4 Astronomical Calculations

The suntools R package (Bivand & Luque, n.d.) was employed to calculate sunrise, sunset times, and daylight duration for each participant’s location. These calculations are based on equations provided by Meeus (1991) and the National Oceanic and Atmospheric Administration (NOAA).

The dates and times of equinoxes and solstices were acquired from the Time and Date AS service (Time and Date AS, n.d.). To verify accuracy, we compared this data with the equations from Meeus (1991) and the results from the National Aeronautics and Space Administration (NASA) ModelE AR5 Simulations (National Aeronautics and Space Administration & Goddard Institute for Space Studies, n.d.).

5.5.5 Sample Characteristics

The analysis dataset consisted of \(65,824\) participants aged 18 or older residing in the UTC-3 timezone. These individuals completed the survey during a one-week period from October 15th to 21st, 2017, providing a snapshot of the population at that specific time.

The unfiltered valid sample included \(115,166\) participants from all Brazilian states. The raw dataset contained \(120,265\) individuals, with \(98.173\%\) of the responses collected between October 15th and 21st, 2017. This data collection period coincided with the promotion of the online questionnaire via a broadcast on a nationally televised Sunday show in Brazil (Rede Globo, 2017).

Based on 2017 data from the Brazilian Institute of Geography and Statistics’s (IBGE) Continuous National Household Sample Survey (PNAD Contínua) (Instituto Brasileiro de Geografia e Estatística, n.d.), Brazil had \(51.919\%\) of females and \(48.081\%\) of males with an age equal to or greater than 18 years old. The sample is skewed for female subjects, with \(66.433\%\) of females and \(33.567\%\) of male subjects. The mean age was \(32.109\) (\(\text{SD} = 32.109\)), ranging from \(18\) to \(58.95\) years.

To balance the sample, weights were incorporated into the models. These weights were calculated through cell weighting, using sex, age group, and state of residence as references, based on population estimates from IBGE for the same year as the sample.

A survey conducted in 2019 by IBGE (2021) found that \(82.17\%\) of Brazilian households had access to an internet connection. Therefore, this sample is likely to have a good representation of Brazil’s population.

The sample latitudinal range is \(33.85026\) decimal degrees (\(\text{Min.} = -33.52156\), \(\text{Max.} = 0.32869\)) with a longitudinal span of \(22.74063\) (\(\text{Min.} = -57.5531\), \(\text{Max.} = -34.81247\)). For comparison, Brazil has a latitudinal range of \(39.02299\) decimal degrees (\(\text{Min.} = -33.75115\); \(\text{Max.} = 5.27184\)) and a longitudinal span of \(45.15451\) (\(\text{Min.} = -73.99045\); \(\text{Max.} = -28.83594\)), according to data from IBGE collected via the geobr R package (R. H. M. Pereira & Goncalves, n.d.).

Additional details about the sample are available in the supplementary materials.

5.5.6 Power Analysis

To assess the adequacy of the sample size for detecting effects reaching the Minimum Effect Size (MES) threshold (\(f^2 = 0.02\)), we conducted an a posteriori power analysis using the pwrss R package (Bulus, n.d.). This analysis revealed a minimum sample size of \(1,895\) observations per variable to achieve a power (\(1 - \beta\)) of \(0.99\) with a significance level (\(\alpha\)) of \(0.01\). Our large sample size (\(n = 65,824\)) comfortably surpasses this threshold, ensuring adequate power.

5.5.7 Data Wrangling

Data wrangling and analysis followed the data science framework proposed by Hadley Wickham and Garrett Grolemund (Wickham et al., 2023). All processes were conducted using the R programming language (R Core Team, n.d.), the RStudio IDE (Posit Team, n.d.), and several R packages. The tidyverse and rOpenSci peer-reviewed package ecosystem and other R packages adherents of the tidy tools manifesto (Wickham, 2023) were prioritized.

The MCTQ data was analyzed using the mctq R package (Vartanian, n.d.), which is part of the rOpenSci peer-reviewed ecosystem. The data pipeline was built using the rOpenSci peer-reviewed targets R package (Landau, 2021), which provides a reproducible and efficient workflow for data analysis.

All processes were designed to ensure result reproducibility and adherence to the FAIR principles (Findability, Accessibility, Interoperability, and Reusability) (Wilkinson et al., 2016). All analyses are fully reproducible and were conducted using Quarto computational notebooks. The renv R package (Ushey & Wickham, n.d.) was employed to ensure that the R analysis environment can be reliably restored.

5.5.8 Hypothesis Test

To test the study hypothesis, nested multiple linear regression models were compared: a restricted model (excluding latitude) and a full model (including latitude). To ensure practical significance, a Minimum Effect Size (MES) criterion was applied, in line with the original Neyman-Pearson framework for hypothesis testing (Neyman & Pearson, 1928a, 1928b; Perezgonzalez, 2015). The MES was set at a Cohen’s threshold for small effects (\(f^2 = 0.02\), equivalent to \(\text{R}^2 = 0.01960784\)) (“just barely escaping triviality” (Cohen, 1988, p. 413)). Consequently, latitude was considered significant only if its inclusion explained at least \(1.960784\%\) of the variance in the dependent variable.

The hypothesis test was structured as follows:

  • Null hypothesis (\(\text{H}_{0}\)): Adding latitude does not meaningfully improve the model’s fit, indicated by a negligible change in adjusted \(\text{R}^{2}\) or a non-significant F-test (with a Type I error probability (\(\alpha\)) of \(0.05\)).

  • Alternative Hypothesis (\(\text{H}_{a}\)): Adding latitude meaningfully improves the model’s fit, indicated by an increase in adjusted \(\text{R}^{2}\) exceeding the MES and a significant F-test (with \(\alpha < 0.05\)).

Formally:

\[ \begin{cases} \text{H}_{0}: \Delta \ \text{Adjusted} \ \text{R}^{2} \leq \text{MES} \quad \text{or} \quad \text{F-test is not significant} \ (\alpha \geq 0.05) \\ \text{H}_{a}: \Delta \ \text{Adjusted} \ \text{R}^{2} > \text{MES} \quad \text{and} \quad \text{F-test is significant} \ (\alpha < 0.05) \end{cases} \]

Where:

\[ \Delta \ \text{Adjusted} \ \text{R}^{2} = \text{Adjusted} \ \text{R}^{2}_{\text{full}} - \text{Adjusted} \ \text{R}^{2}_{\text{restricted}} \]

5.6 Data Availability

Some restrictions apply to the availability of the main research data, which contain personal and sensitive information. As a result, this data cannot be publicly shared. Data are, however, available from the author upon reasonable request.

The code repository is available on GitHub at https://github.com/danielvartan/mastersthesis, and the research compendium can be accessed via The Open Science Framework at the following link: https://doi.org/10.17605/OSF.IO/YGKTS.

5.7 Acknowledgments

This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001, Grant number 88887.703720/2022-00.

5.8 Ethics Declarations

The author declares that the study was carried out without any commercial or financial connections that could be seen as a possible competing interest.

5.9 Additional Information

See the supplementary material for more information.

Correspondence can be sent to Daniel Vartanian ().

5.10 Rights and Permissions

This article is released under the Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution, and reproduction in any medium or format, as long as be given appropriate credit to the original author and the source, provide a link to the Creative Commons license, and indicate if changes were made.