Donald Trump’s support in recent Republican primary victories spanned the demographic board.

However, polling shows a key weak spot: church attenders. People who attend church frequently remain much less likely to back Trump.

Despite Trump’s struggles with active church attenders, media pundits have given much attention to the evangelical support for the thrice-married casino mogul with a reputation for womanizing and profanity.

Although Texas Sen. Ted Cruz topped Trump among evangelicals in Iowa, Trump won the evangelical vote in New Hampshire, South Carolina and Nevada.

In South Carolina – the most evangelical state so far where self-identified born-agains accounted for 72 percent of the turnout – Trump actually did better among evangelicals than among non-evangelicals.

He also narrowly won among voters in the Palmetto State who said it mattered to them that a candidate shares their religious beliefs.

While Trump dominates among most voters, a soft spot in his support emerges when comparing voters by church attendance rates.

Polling shows that church attendance rates indicate more significant changes in support for Trump than whether or not one identifies as an evangelical.

Looking at a three-way match-up between Trump, Cruz and Florida Sen. Marco Rubio, Trump easily leads in Reuters polling overall with 39 percent of the vote. Filtering the results based on church attendance reveals a significant turning point.

Trump overwhelmingly wins those who infrequently, rarely or never attend church. Forty-three percent of those who attend less than monthly back Trump, compared to just 15 percent each for Cruz and Rubio.

Among those who attend church at least monthly, the race tightens with Trump at 36 percent, Cruz at 27 and Rubio at 25.

For those who attend more than once a week, the numbers shift dramatically as Cruz leads at 34 percent, followed by Rubio at 29 and Trump far behind at 19.

Using church attendance as a filter – instead of self-identification as an evangelical – may suggest a critical divide between those who merely feel a cultural affinity with a religious identity from those who actively embrace the religious life.

The commitment to engage in the life of a local church likely also means one spends more time in prayer, Bible reading and reflection on religious teachings and values.

Bob Perry, strategy development team leader for Churchnet and a longtime church health consultant and author, told EthicsDaily.com that in “the highly politicized environment of an election year, words often are redefined to carry heavy political implications.” He fears the term “evangelical” suffered this fate.

“It is now a political word being thrown around by politicians and pundits who have no clear idea of the origin of the word or an accurate definition of it,” he explained. “One can now see ‘evangelicals’ whose values and behavior show no semblance to the gospel and no conformity to the teachings of Jesus.”

“There have long been ‘cultural evangelicals’ who identified themselves with the religious traditions of their parents or grandparents,” Perry added. “Now we have ‘political evangelicals’ for whom arrogance is preferred over humility, material success is preferred over generosity of spirit, hateful speech is preferred over kindness, and spectacle is preferred over dignity. God forgive us.”

Despite this significant dividing point in the polling data, media coverage remains focused on evangelicals since entrance and exit polls thus far have focused almost exclusively on that question.

In Iowa, New Hampshire and Nevada, evangelical identification was the only question pollsters asked, while in South Carolina they added the question about importance of shared religious beliefs.

The exclusive focus on evangelicals occurred despite the fact evangelicals are a small slice of the vote in New Hampshire and Nevada.

In general elections, other religious exit poll questions explore broader religious affiliation (like Protestant, Catholic, other or none) and church attendance rates.

In the 2012 Republican primaries, the exit polls included religious affiliation questions in New Hampshire, South Carolina and Nevada, but those questions were dropped this year.

For Democrats in South Carolina on Saturday, exit polls asked about church attendance – the first religion question in four states for Democratic voters this year.

While frontrunner Hillary Clinton did better among those who attend weekly or more (54 percent of Democratic voters), Vermont Sen. Bernie Sanders improved among those who rarely attend.

In addition to exit poll data, media coverage also adds to the narrow focus on evangelicals as the measure for religiosity and presidential support.

Media coverage often appears to use the word “evangelical” as a generic replacement for “Christian.”

Many of Trump’s most prominent religious supporters – like Liberty University President Jerry Falwell Jr. and First Baptist Church of Dallas senior pastor Robert Jeffress – could be labeled “fundamentalist” instead of “evangelical.”

Additionally, when Trump attended a church service shortly before the Iowa caucus, several prominent media outlets called it part of his outreach to evangelicals even though the church he attended is part of Presbyterian Church, USA, which is a mainline – not evangelical – denomination.

If Trump garners the Republican nomination, he will need to convince regular church attenders to back him.

Republican presidential candidates traditionally win more support from regular attenders, while Democrats gain more support from infrequent attenders.

Some pundits and scholars refer to this contrast as the “god gap,” although Democrats perform better on other religious indicators.

Brian Kaylor is a contributing editor for EthicsDaily.com. You can follow him on Twitter @BrianKaylor.